PhD PROPOSAL - Kenyatta University Repository

16 downloads 0 Views 1MB Size Report
MBA (MARKETING) KENYA METHODIST UNIVERSITY (2010) ...... quantitative and qualitative methods in order to contribute to a rich and comprehensive study. ...... using empirical data that was collected from a sample of 342 responses.
CUSTOMERS’ PERCEPTIONS AND USAGE OF ONLINE RETAILING SERVICES IN NAIROBI COUNTY, KENYA

BY

PETER M. MWENCHA D86/CTY/21719/2010

A THESIS SUBMITTED TO THE SCHOOL OF BUSINESS IN FULFILLMENT OF THE REQUIREMENT FOR THE AWARD OF DOCTOR OF PHILOSOPHY DEGREE IN BUSINESS ADMINISTRATION OF KENYATTA UNIVERSITY

JULY, 2015

i

EFFECT OF CUSTOMER PERCEPTIONS ON THE USAGE OF ONLINE RETAILING SERVICES IN NAIROBI COUNTY, KENYA

BY

PETER M. MWENCHA BA (IR) UNITED STATES INTERNATIONAL UNIVERSITY (2008) MBA (MARKETING) KENYA METHODIST UNIVERSITY (2010)

A DISSERTATION SUBMITTED TO THE SCHOOL OF BUSINESS IN PARTIAL FULFILMENT OF THE REQUIREMENT FOR THE AWARD OF DOCTOR OF PHILOSOPHY DEGREE IN BUSINESS ADMINISTRATION OF KENYATTA UNIVERSITY

JUNE, 2013

i

DECLARATION

This thesis is my original work and has not been presented for any award in any other university. No part of this thesis should be reproduced without the authority of the author and/or Kenyatta University.

Signature: ________________________________ Date: __________________ Mwencha, Peter Misiani Business Administration Department

We confirm that the work reported in this research thesis was carried out by the candidate under our supervision as the appointed university supervisors.

Signature: ________________________________ Date: ___________________ Dr. Muathe, SMA (PhD) Department of Business Administration, School of Business, Kenyatta University

Signature: _______________________________ Date: ___________________ Prof. Kuria Thuo, J. (PhD) School of Business Gretsa University

ii

DEDICATION

To my parents, my father Samuel Mwencha and my late mother Mary Mwencha, My uncles Amb. Erastus Mwencha, Hon. Henry Obwocha and Stephen Mwencha, my Aunts Mary Mwencha, Dolline Obwocha, Victoria Sagero, Getrude Sagero and Millicent Bochere, my elder brother Mogaka Mwencha, my sister Evelyne Kerubo, my younger brothers Charles Mokaya and Henry Orenge, friends Peter Muendo, Belinda Maina, Sheridan Muruka, Anthony Riri, Evelyne Kamau, Davies and Milda Kinyua as well as my cousins Mogaka and Kinya Mwencha, Patrick, Norah and Maurice Masenge for their love, encouragement and unwavering support.

iii

ACKNOWLEDGEMENT I wish to acknowledge several individuals who contributed both directly and indirectly towards the completion of this study. It has been a long, difficult but rewarding endeavor which would not have been possible without their encouragement and support.

First and foremost, I am most grateful to my first supervisor, Dr. Muathe, SMA (Ph.D) who was of great assistance to me at all stages of this study. His guidance, encouragement as well as patience throughout the course of this study were indispensable. I would also like to thank my second supervisor, Prof. J. Kuria Thuo, for his insightful suggestions and meticulous supervision at each and every stage. I would also like to express my gratitude to the faculty at Kenyatta University School of Business, in particular, Dr. Ambrose Jagongo (Ph.D), Dr. David Nzuki (Ph.D), Dr. James Kilika (Ph.D) and Dr. Samuel Maina (Ph.D) for being accessible and helpful whenever I needed clarifications regarding the thesis.

Second, I also acknowledge the support of my colleagues in the PhD program, in particular, Dr. Stanley Karanja (Ph.D), Dr. Reuben Njuguna (Ph.D), Dr. Siaw Frimpong (Ph.D), Dr. Jedidah Muli (Ph.D) and Dr. Rebecca Mensah (Ph.D). Stanley was always keeping tabs on my progress, Njuguna was a good sounding board for ideas, Frimpong was welcome company during the long hours in the library, Jedidah provided me with numerous research tips and Rebecca was an excellent motivator.

Third, I would also like to thank my family and friends for their moral and in-kind support during the course of this study as well as for bearing the inconvenience that comes with such a winding undertaking. I am truly indebted to them.

Fourth, there are other people who contributed towards the completion of this study but could not be mentioned here. They know themselves. I am grateful for all their contributions and I sincerely thank them all for the part they played.

Last but not least, I give thanks to God almighty for giving me good health and for sustaining me throughout the entire period of working on this document.

iv

TABLE OF CONTENTS Page Title....................................................................................................................................... i Declaration..........................................................................................................................ii Dedication ..........................................................................................................................iii Acknowledgements .......................................................................................................... iv Table of contents ................................................................................................................ v List of tables....................................................................................................................... ix List of figures ....................................................................................................................xii Operational definition of terms .....................................................................................xiii Abbreviations and acronyms ......................................................................................... xvi Abstract ..........................................................................................................................xviii CHAPTER ONE: INTRODUCTION ............................................................................. 1 1.1

Background of the Study ..................................................................................... ..1

1.1.1 Usage of Online Retailing Services ........................................................................ 4 1.1.2 Customers‘ Perceptions of Online Retailing........................................................... 6 1.1.3 Online Retailing Services in Kenya ........................................................................ 8 1.2

Statement of the Problem ...................................................................................... 9

1.3

Research Objectives ............................................................................................ 11

1.4

Research Hypotheses........................................................................................... 12

1.5

Significance of the Study .................................................................................... 13

1.6

Scope of the Study .............................................................................................. 14

1.7

Limitations of the Study...................................................................................... 15

1.8

Organization of the Study ................................................................................... 16

CHAPTER TWO : LITERATURE REVIEW .............................................................. 18 2.1

Introduction ......................................................................................................... 18

v

Page

2.2

Theoretical Review of E-Commerce Usage ........................................................ 18

2.2.1

Behavioral Model of System Usage ................................................................ 18

2.2.2

Innovation Diffusion Theory ............................................................................ 21

2.2.3

Expectation-Confirmation Theory .................................................................... 23

2.2.4

Perceived Risk Theory ...................................................................................... 25

2.2.5

Theory of Consumption Values ......................................................................... 27

2.3

Empirical Literature Review ............................................................................... 29

2.3.1

Usage of Online Retailing Services ................................................................... 29

2.3.2

Antecedent Role of Customer Perceptions ........................................................ 32

2.3.3 Mediating Role of Customer Satisfaction .......................................................... 39 2.3.4

Moderating Effect of Demographic Factors ...................................................... 44

2.4

Summary of Empirical Literature and Research Gaps ....................................... 45

2.5

Conceptual Framework ....................................................................................... 47

CHAPTER THREE : RESEARCH METHODOLOGY ............................................. 51 3.1

Introduction ......................................................................................................... 51

3.2

Research Philosophy ........................................................................................... 51

3.3

Research Design .................................................................................................. 52

3.4

The Empirical Model .......................................................................................... 53

3.4.1 The Direct Effects Model .................................................................................... 53 3.4.2 The Mediated Effects Model ............................................................................... 56 3.4.3 The Interaction Effects Model............................................................................. 57 3.5

Operationalization and Measurement of Study Variables................................... 58

3.6

The Study Area.................................................................................................... 60

3.7

Target Population ................................................................................................ 61

3.8

Sampling Design and Procedure ......................................................................... 62 vi

Page

3.8.1 Sampling Technique ........................................................................................... 63 3.8.2 3.9

Sample Size Determination ............................................................................... 65 Data Collection Instrument ................................................................................. 66

3.9.1 Self Administered Questionnaire ....................................................................... 67 3.9.2

Key Informant Interview .................................................................................. 67

3.9.3 Validity of Data Collection Instruments ............................................................ 68 3.9.4

Reliability of Data Collection Instruments ........................................................ 70

3.10 Data Collection Procedures ................................................................................. 72 3.11 Data Analysis and Presentation ........................................................................... 76 3.12 Ethical Issues ....................................................................................................... 83 CHAPTER FOUR: RESEARCH FINDINGS AND DISCUSSIONS ......................... 85 4.1

Introduction ......................................................................................................... 85

4.2

Descriptive Data Analysis ................................................................................... 85

4.2.1

Response Rate.................................................................................................... 85

4.2.2

Sample Demographic Characteristics ................................................................ 86

4.2.3

Customer Perceptions of Online Retailing Users .............................................. 88

4.2.4 Usage of Online Retailing Services .................................................................... 90 4.2.5 4.3

Customer Satisfaction of Online Retailing Users .............................................. 90 Regression Analysis and Test of Hypotheses ..................................................... 91

4.3.1

Diagnostic Tests ................................................................................................ 92

4.3.2

Test of Hypotheses ............................................................................................ 95

4.3.2

Summary of Research Hypotheses .................................................................. 102

4.4

Content Analysis ............................................................................................... 103

4.4.1

Interview Participants ..................................................................................... 103

4.3.2

Key Themes ..................................................................................................... 104

vii

Page

CHAPTER FIVE: SUMMARY, CONCLUSIONS AND RECOMMENDATIONS110 5.1

Introduction ....................................................................................................... 110

5.2

Summary ........................................................................................................... 110

5.3

Conclusions ....................................................................................................... 113

5.4

Policy Implications ............................................................................................ 115

5.5

Contributions of the study to knowledge .......................................................... 117

5.6

Suggestions for further study ............................................................................ 119

REFERENCES ............................................................................................................... 120 APPENDICES ................................................................................................................ 145 Appendix 1: Supplementary Statistical Analyses ........................................................... 145 Appendix 2: List of Online Retailing Firms in Nairobi, Kenya ...................................... 154 Appendix 3: Data Collection Instruments........................................................................ 155 a).

Cover Letter....................................................................................................... 155

b).

Questionnaire .................................................................................................... 156

c).

Interview Guide ................................................................................................. 159

Appendix 4: Research Authorization ............................................................................... 162 a).

Clearance Letter ................................................................................................ 162

b).

Research Permit................................................................................................. 163

Appendix 5: Code Book ................................................................................................. .164 a).

Codebook for Quantitative Data Analysis ....................................................... 164

b).

Codebook for Qualitative Data Analysis .......................................................... 165

c).

Summary of Major Themes............................................................................... 166

Appendix 6: Summary of Empirical Review & Research Gaps ..................................... .168

viii

LIST OF TABLES Page Table 3.1: Operationalization and Measurement of Study Variables ................................ 59 Table 3.2: Distribution of Target Population ..................................................................... 62 Table 3.3: Sampling Frame & Sample Distribution Table ................................................ 66 Table 3.4: Reliability of Questionnaire Items ................................................................... 71 Table 3.5: Inferential Data Analysis Techniques ............................................................... 78 Table 4.1: Distribution of Responses ................................................................................. 86 Table 4.2: Demographic Characteristics of the Sample..................................................... 87 Table 4.3: Descriptive Statistics Results for Customers‘ Perceptions ............................... 88 Table 4.4: Descriptive Statistics Results for Individual Perceptual Indicators ................. 89 Table 4.5: Descriptive Statistics Results for Customer Satisfaction.................................. 90 Table 4.6: Descriptive Statistics Results for Usage of Online Retail Services .................. 90 Table 4.7: Results of Collinearity Statistics ....................................................................... 93 Table 4.8: Results of Kolmogorov-Smirnov Normality Test ........................................... 94 Table 4.9: Results of Logit Regression Analysis ............................................................... 95 Table 4.10: Results of of Simple Linear Regression Analysis ......................................... 97 Table 4.11: Summary of Hypotheses Test ...................................................................... 102 Table 4.12: Distribution of Interview Participants ......................................................... 103 Table A.1: Reliability Output - Usefulness .................................................................... 145 Table A.2: Reliability Output - Compatibility ............................................................... 145 Table A.3: Reliability Output - Ease-of-Use ................................................................. 145 Table A.4: Reliability Output - Financial Risk ............................................................... 145 Table A.5: Reliability Output - Performance Risk ........................................................ 145 Table A.6: Reliability Output - Personal Risk ............................................................... 146 Table A.7: Reliability Output - Monetary Value ............................................................ 146

ix

Page Table A.8: Reliability Output - Convenience Value ...................................................... 146 Table A.9: Reliability Output - Social Value .................................................................. 146 Table A.10: Reliability Output - Emotional Value ......................................................... 146 Table A.11: Reliability Output - Level of Satisfaction ................................................... 147 Table A.12: Correlation Matrix for the Three Predictor Variables ................................. 147 Table A.13: Collinearity Output showing Tolerance and VIF ........................................ 147 Table A.14: Hosmer and Lemeshow Test for the Main Effects Model ........................... 147 Table A.15: Normality Test for Predictor Variables ....................................................... 148 Table A.16: Dependent Variable Encoding for Predictor Variables ............................... 148 Table A.17: Model Summary for Logistic Regression of Direct Effects Model ............ 148 Table A.18: Classification Table for Direct Effects Model ............................................ 149 Table A.19: Case processing summary for Direct Effects Model .................................. 149 Table A.20: Logistic Regression Results for Direct Effects Model ............................... 149 Table A.21: Model Summary for Linear Regression Results for the Relationship between Customers‘ Perceptions and Customer Satisfaction ..................... 150 Table A.22: ANOVA (F-Test) for Linear Regression Results of the

Relationship

between Customers‘ Perceptions and Customer Satisfaction .................... 150 Table A.23: Linear Regression Results of Relationship between Customers‘ Perceptions and Customer Satisfaction ........................................................................... 150 Table A.24: Dependent Variable Encoding for Logistic Regression of Relationship between Customer Satisfaction and Usage ................................................. 151 Table A.25: Classification Table: Logistic Regression of Relationship between Customer Satisfaction and Usage ................................................................ 151 Table A.26: Case Processing Summary for Logistic Regression of the Relationship between Customer Satisfaction and Usage ................................................. 151 x

Page Table A.27: Model Summary for Logistic Regression of Relationship between Customer Satisfaction and Usage .............................................................. 152 Table A.28: Logistic Regression Results of the Relationship between Customer Satisfaction and Usage ............................................................................... 152 Table A.29: Case Processing Summary for Moderated Effects Model .......................... 152 Table A.30: Dependent Variable Encoding for Moderated Effects Model .................... 152 Table A.31: Moderated Effects Model Summary ........................................................... 153 Table A.32: Classification Table for Moderated Effects Regression ............................. 153 Table A.33: Logistic Regression Results for Moderated Effects Model ........................ 153

xi

LIST OF FIGURES Page Figure 2.1 : Behavioral Model of System Usage .............................................................. 19

Figure 2.2 : Diffusion of Innovation Theory ..................................................................... 21

Figure 2.3 : Expectations-Confirmation theory ................................................................. 24

Figure 2.4 : Perceived Risk Theory ................................................................................... 26

Figure 2.5 : Theory of Consumption Value ....................................................................... 28

Figure 2.6 : Conceptual Framework .................................................................................. 48

xii

OPERATIONAL DEFINITION OF TERMS Active users

Individuals that have used a particular online retailing service at least once in the last three months.

Adoption

The acceptance (initial use) of an online retailing service

B2C E-Commerce

The direct activity between businesses and consumers through which consumers fulfill their needs (e.g. information, products or services) using information and communications technology applications such as the internet.

C2C E-Commerce

The direct activity amongst consumers through which they fulfill their needs (e.g. information, products or services) using information and communications technology applications such as the internet.

Compatibility

The degree to which using a particular online retailing service is perceived as being consistent with the existing values, needs and past experiences of the potential user.

Constructs

Abstractions (theoretical concepts) that cannot be observed directly but are useful in interpreting empirical data and in theory building.

Customers’ Perceptions

The subjective opinions/beliefs/judgments of an individual vis-à-vis an online retailing service based on prior use experience.

Customer Satisfaction

The customer‘s overall positive evaluation of the online retailing service following initial usage or based on all prior interactions/encounters and experiences with the online retailing service.

Data triangulation

The collecting of study data over different times or from different sources.

Demographic Variables

Personal characteristics/attributes of a consumer that tend to remain static throughout an individual‘s life time, or evolve slowly over time. This includes age, gender, race, education, income, lifestyle, etc.

E-commerce

The use of electronic communications and digital information processing technology in business xiii

transactions for value creation between organizations, between organizations and individuals as well as between individuals. Inactive users

Individuals that have not used a particular online retailing service at least once in the last three months.

Innovation

An idea, practice or object that is perceived as new by an individual.

Market Development

Possible ways of increasing and sustaining the usage of online retailing services.

Market Prospects

Economic/business potential of the online retailing subsector.

Member Checking

Verifying the credibility of constructions of the key informant interview participants.

Methodical triangulation

The use of a combination of methods such as document analysis, interviews and surveys in a study.

Moderating variable

A third variable that modifies the strength or direction of a causal relationship.

Monetary value

The customer‘s evaluation of the total financial cost of the online retailing service (including the price paid) relative to the benefits received.

Null hypothesis

A statement that contradicts the assumed result/outcome of the study.

Online retailing

The selling and buying of goods/services through the internet.

Operational definition

The definition that ascribes meaning to a construct by specifying operations that the researcher must perform to measure or manipulate the construct.

Perception

A subjective opinion/belief/judgment of an individual vis-à-vis an online retailing service.

Perceived risk

The transaction-related risks that consumers face as a result of using online retailing services.

Perceived value

The consumer‘s evaluation of the benefits of online retailing usage.

Post-adoption behavior

The various adoption outcomes, use behaviors, and feature extension behaviors made by an individual after xiv

an online retailing service has been implemented, made accessible to the user, and applied by the user in accomplishing his/her online shopping activities. Prevailing Attitudes

Opinions, thoughts and feelings regarding online retailing services.

Psychographics

A customer‘s inner feelings and predisposition to behave in a certain way.

Pure play online firms

Firms that provide their services only via online/internet channels e.g. internet retailing.

Retailer

An entity that sells products and services directly to the final consumers for their personal use, be it online, offline or both.

Theory

A set of interrelated constructs/variables that present a systematic view of a phenomenon by specifying relationships among the variables, with the purpose of explaining the phenomenon.

Usage

The utilization of one or more features of an online retailing service by registered users within a certain timeframe.

Usage Diversity

The types/nature and extent of usage/utilization of online retailing services

Usage Drivers

Determinants of online retailing usage

Variable

A construct that can assume different values and scores.

Variable respecification

A procedure in which the existing data are modified to create new variables, or in which a large number of variables are collapsed into fewer variables in line with the study‘s objectives.

xv

ABBREVIATIONS AND ACRONYMS

ANOVA

Analysis of Variance

B2B

Business-to-Business

B2C

Business-to-Consumer

BMSU

Behavioural Model of System Usage

CA

Communications Authority

CCK

Communication Commission of Kenya

CS

Customer Satisfaction

COFEK

Consumer Federation of Kenya

CUI

Continued Usage Intention

C2C

Business-to-Consumer

DOI

Diffusion of Innovation

DV

Dependent Variable

E-COMMERCE

Electronic Commerce

ECM

Expectations Confirmation Model

ECT

Expectations (Dis) Confirmation Theory

IS

Information Systems

IT

Information Technology

IV

Independent Variable

ICT

Information and Communication Technology

IDT

Innovation Diffusion Theory

KS

Kolmogorov Smirnov Test

LOGIT

Logistic Regression

MIS

Management Information Systems

xvi

NACOSTI

National Commission for Science, Technology and Innovation

PCI

Perceived Characteristics of Using an Innovation

PEOU

Perceived Ease of Use

PERVAL

Perceived Value Scale

PR

Perceived Risk

PU

Perceived Usefulness

PV

Perceived Value

RR

Response Rate

SERVQUAL

Service Quality

SCT

Social Cognitive Theory

SEM

Structural Equation Modelling

SNS

Social Networking Sites

SPSS

Statistical Package for Social Sciences

TAM

Technology Adoption Model

TRA

Theory of Reasoned Action

TCV

Theory of Consumption Values

US

United States

UN-HABITAT

United Nations Human Settlements Programme

UTAUT

Unified Theory of Acceptance and Use of Technology

VIF

Variance Inflation Factor

xvii

ABSTRACT In spite of the huge increase in internet usage in Kenya over the past years, the usage of online retailing services in Kenya is still very low, thereby posing an existential threat to the service providers. To induce more initial users to continue using these services, there is need to establish what affects their continued usage. Individual factors, in particular customer perceptions, have been shown by both the information systems (IS) as well as the marketing fields to have significant effect on sustained use of online retailing services. This study therefore sought to establish the effect of customer perceptions on the usage of online retailing services in Nairobi County, Kenya. Its objectives were to establish whether there is a relationship between perceived attributes and usage of online retailing services, to determine whether there is a relationship between perceived risk and usage of online retailing services, to analyse whether there is a relationship between perceived value and usage of online retailing services, to evaluate whether customer satisfaction has a mediating effect on the relationship between customer perceptions and usage of online retailing services and to establish whether customer demographics have a moderating effect on the relationship between customer perceptions and usage of online retailing services. The study employed a descriptive, cross-sectional, survey design and explanatory design. The target population was 6 online retailing firms and the respondents for this study were the 18,147 registered users of these six online retailing firms in Nairobi County, Kenya. A sample of 391 respondents was selected using multi-stage sampling methods including purposive, stratified and simple random sampling. Primary data was collected using a self-administered structured questionnaire and an interview guide, while secondary data was collected through document review. Questionnaire responses were analyzed using descriptive and inferential statistics which involved both linear and logistic regression analysis. Figures and tables were used to present the data. Data from key informant interviews was analyzed using content analysis technique to complement the quantitative data. The results showed that consumer perceptions have a significant effect on the usage of online retailing services. The study also found that customer satisfaction does have a mediating effect on the customer perception – usage relationship. Furthermore, the research established that demographic factors do not have a significant moderating effect on the customer perception - usage relationship. The findings of this study underscore the importance of customer perceptions and customer satisfaction in enhancing the likelihood of success of online retailing services. Consequently, the study recommends that online retailers should enhance service features/attributes as a way of ensuring success of their services by taking into consideration customer-specific needs by personalizing the website to make it more useful, compatible with customer requirements and easy to use for users. In addition, online retailing service providers need to build trust amongst their users regarding online purchasing. Further, online retailers should design and deliver a unique value proposition that has both functional as well as hedonistic appeals. Online retailers should also have an effective customer satisfaction strategy for purposes of customer retention. Moreover, it is imperative for online retailing firms to have a good understanding of their target customers, since this will not only help in determining the appropriate customer engagement strategies but also how to enhance the long-term usage of their services. On the government‘s part, the study recommends the tackling the barriers to online shopping usage primarily through legislation. Since usage also hinges on trust, the government could license a suitable entity to oversee online consumer protection to address users‘ concerns. xviii

CHAPTER ONE INTRODUCTION

1.1

Background of the Study

The commercial use of the Internet has grown tremendously over the last two decades, and is characterized by a proliferation of various online-based electronic commerce (ecommerce) services. One of these services is online retailing, which has been described using a number of different terms (Mottner, Thelen & Karande, 2002). It has been referred to as internet retailing, e-retailing, or e-tailing (Anderson, 2000), as part of interactive home shopping (Alba, Lynch, Weitz & Janisqewski, 1997), and by the broader terms electronic commerce (Daniel & Klimis, 1999) and e-commerce (Boscheck, 1998).

According to the Australian Government Productivity Commission (AGPC), online retailing can take several forms: i) as ‗pure play‘ services in which businesses provide online-only services in particular retail categories, ii) as brick-and-click (multi-channel) establishments where online activities are combined with bricks-and-mortar operations, iii) as online marketplaces where buyers and sellers interact on an electronic trading platform provided by a third-party and iv) as manufacturer-owned websites where products are sold directly to customers, thus by-passing middlemen (2011).

The late 1990s heralded the coming of age of online retailing, with the unprecedented growth in reported sales surpassing triple digit growth (United States Census Bureau, 2004), though it slowed considerably due the failure of e-commerce firms in 2000 (Rohm & Swaminathan, 2004). Nonetheless, the U.S. Commerce Department has been reporting annual e-commerce statistics since 1999, signifying the importance of this subsector to the world‘s largest economy (Haynes & Taylor, 2006).

1

Due to its huge popularity, online retailing has had a significant impact on several market segments such as travel, consumer electronics, hobby goods, and media goods across the globe (Weltevrenden & Boschma, 2008). Consequently, online retailing has evolved into an established marketing channel in its own right within the consumer marketplace (Doherty & Ellis-Chadwick, 2010).

In terms of size, the U.S. is the largest market, and is expected to reach $278.9 billion in sales in 2015 (Forrester, 2011a). In Europe, the second largest market, the number of online buyers is expected to grow from 157 million to 205 million by 2015; total sales are forecast to reach 133.6 billion Euros (Forrester, 2011b). Africa is also gradually embracing online retailing, with countries like South Africa and Egypt ahead of the rest. In South Africa for instance, 51% of those with access to the internet are shopping online, according to a 2011 MasterCard Worldwide survey (Kermeliotis, 2011).

Kenya is showing strong growth potential, as it was the fastest growing Internet market in Africa in 2011 (yStats.com, 2012) with its internet population rising by about 19% to stand at 14.032 million users in 2012 from 12.5 million in 2011 (Communications Commission of Kenya (CCK), 2012). A recent survey of 1700 individuals found that 18% to 24% of the respondents purchase music, movies and e-books online, thus signaling the growth of online shopping in Kenya (Juma, 2010). This upward trend has been aided by the increasing number of young people who prefer to access information via their mobile phones, coupled with the declining prices of internet connectivity costs as well as the high uptake of mobile payment services. This has created an opportunity for online trading platforms such as N-Soko, OLX, Jumia and Rupu, among others (Okuttah, 2014).

2

Notwithstanding the significant growth in internet usage in Kenya over the past few years, the online shopping market is still quite small even by regional standards. Reports indicate that the Communications Authority (CA) estimates the value of e-commerce in Kenya at Sh4.3 billion, in comparison to South Africa‘s Sh54 billion, Egypt‘s Sh17 billion and Morocco‘s Sh. 9.6 billion (Okuttah, 2014).

This disproportionately low online retailing usage situation is corroborated by Nakumatt, the leading supermarket chain in Kenya and East Africa, which insists that it has felt no impact from online retailing. The company decries the fact that despite its presence on online retail services, it was seeing little benefit if any there from. According to the retail chain‘s analysis, ―online shopping is yet to take root in East Africa and still commands less than 1 percent of turnover. We therefore cannot attest to having been affected by online shopping‖ (Consumer Federation of Kenya (COFEK), 2013).

In the long run, this low usage of online retailing services poses a problem for investors on how to monetize and sustain their investments in these platforms (Okuttah, 2014), since low usage may result in the service provider incurring undesirable costs of maintaining the loss-making service. Continued lossmaking may eventually lead to closure of the service (Cooper & Zmud, 1990; Bhattacherjee, 2001a), as was evident during the closure of the online retail service provider Kalahari.co.ke in 2011, the Kenyan arm of the South African based firm Naspers. According to Naspers, ―the performance of the service had been below expectations since the launch of the service in 2009 and reaching profitability was not (Bizcommunity, 2011).

3

a reasonable near-term prospect‖

Past studies have established that the success of e-retailers hinges more profoundly on the continued use of the system than on initial adoption (Parthasarathy & Bhattacherjee, 1998; Shih & Venkatesh, 2004; Limayem, Hirt & Cheung, 2007). In view of the current state-of-affairs, and given the growing importance online retailing, this study sought to understand the predictors of online retailing service usage by consumers in Kenya.

1.1.1

Usage of Online Retailing Services

Usage behavior is an important concept in both the information systems (IS) and marketing fields, due to the necessity of service providers to increase the uptake of their services and sustain their usage at levels that are economically viable for them to continue providing the service. For this reason, both researchers as well as practitioners have sought to understand system usage behavior, in particular, what influences it (Lucas, 1975; Schewe, 1976; Swanson, 1988; Taylor & Todd, 1995b; Bhattacherjee, 2001a; Hong, Thong, & Tam, 2006).

However, a review of usage literature shows that it is a complex construct with multiple conceptualizations (Burton-Jones & Straub, 2006). For instance, Bhattacherjee (2001a) divided IS usage into two stages: initial adoption and continuous usage, while Liu (2007) postulated that the use of online shopping services is likely to move from partial usage (i.e. information search only) to full usage (i.e. completing the entire transaction process online) during post-adoption use. As such, the context within which it is used is important in determining its relevant measures and dimensions. In the context of this research, the usage of online retailing services is examined at the individual level of analysis (i.e. the consumer use context) as a post-adoption phenomenon that is also referred to as continued use. Therefore, usage serves as the dependent variable in the research model conceptualized for this study.

4

The usage construct in this study is derived from the Behavioral Model of System Usage (BMSU), an early IS model by Schewe (1976) that relates user attitudes to end-user system usage behavior as well as the Expectation-Confirmation Theory (ECT), a marketing theory by Oliver (1980; 2010) and its related Expectation (Dis) Confirmation Model (ECM) by Bhattacherjee (2001a) which adapted the ECT to IS context. ECT is the dominant theoretical lens used to explain IT continuance/discontinuance behaviors (Bhattacherjee & Barfar, 2011). Due to the similarity between re-purchasing products or services in a consumer context and the continued use of technology, the ECM posits an equivalent relationship in the continued technology usage context (Bhattacherjee, 2001a).

Given the empirical support for the impact of continued usage on the IT system success, establishing what that affect customers‘ usage behavior (either to continue or to discontinue usage of an IT) is of importance (Hong et al, 2006). Accordingly, research in IT continuance has examined different factors and/or processes that motivate continued usage or discontinuance of IT products or services, following their initial acceptance (Bhattacherjee & Barfar, 2011).

In essence, continued usage of IS can be influenced by individual/psychological, system/technical and organizational factors (Bajaja & Nidumolu, 1998). However, this study restricts itself to examining individual psychological factors, specifically the antecedent role of customers‘ perceptions on usage of online retailing services, coupled with the mediating role of satisfaction on the perception-usage relationship. This is in line with a study conducted by Lucas (1975), which ascertained that the use of an IS is dependent on user attitudes and perceptions.

5

It is important to note that as opposed to organizational IS usage, individuals use IS such as online retailing services not only for utilitarian purposes, but also for hedonic purposes (Monsuwé, Dellaert & De Ruyter, 2004; Bridges & Florsheim, 2008; Ozen & Kodaz, 2012). Therefore, the affective aspect of online shopping is just as important as the cognitive aspect and therefore needs to be taken into consideration when seeking to establish what affects the usage of online retailing services (Ozen & Kodaz, 2012).

1.1.2

Customers’ Perceptions of Online Retailing

Perceptions are essentially mental maps made by people to give them a meaningful picture of the world on which they can base their decisions (Berelson & Steiner, 1964). Perception occurs when stimuli are registered by one of the five human senses: vision, hearing, taste, smell and touch (Hoyer & MacInni, 2008) via a process of sensing, selecting, and interpreting stimuli in the external, physical world into the internal, mental world (Wilkie, 1994). This perceptual process leads to a response, which is either overt (actions) or covert (motivations, attitudes, and feelings) or both.

From a consumer behavior perspective, perceptions are an attempt by a consumer to obtain and process information about a market situation with a purpose to make himself aware of the market and market offerings – market events, marketers, products/services, advertisement, physical environment of the market outlet and so on (Sahaf, 2008). Consumers establish and continuously update their perceptions about the alternative products/services that they are considering and based on those perceptions, they determine their attitudes towards the products (preferences).

6

The selection and interpretation of stimuli is highly subjective and is based on what the consumer expects to see in light of previous experience, on the number of plausible explanations he or she can envision, on motives and interests at the time of perception, and on the clarity of the stimulus itself. Due to the way people view, experience and remember things, two people may have differing perceptions of the same stimulus item (e.g. a product or service). This makes perceptions a confusing and complex phenomenon (Schiffman & Kanuk, 2010).

Perception has strategy implications for marketers because consumers make decisions based on what they perceive rather than on the basis of objective reality. As a result, marketers have realised that understanding the perceptual process of consumers helps them to design better ways to help customers perceive their products and services favorably, especially since products and services that are perceived distinctly and favorably have a much better chance of being purchased than products or services with unclear or unfavorable images (Schiffman & Kanuk, 2010).

Similarly, researchers in IS have over the years sought to establish how potential users‘ perceptions of an IT innovation influences its adoption (Tornatzky & Klein, 1982; Moore & Benbasat, 1991) and continued usage (Lucas, 1975; Schewe, 1976; Parthasarathy & Bhattacherjee, 1998; Bhattacherjee, 2001b; Venkatesh, Morris, Davis & Davis, 2003).

For this reason, organizations interested in influencing usage of their services need to better understand their customers‘ perceptions (Schewe, 1976; Bhattacherjee, 2001a; Bhattacherjee & Barfar, 2011). Consequently, the customer perception construct serves as the independent variable in this study and has theoretical foundations from three perceptual constructs identified in literature as playing an antecedent role via-a-vis online retailing usage. These are: - perceived attributes, perceived risk and perceived value. 7

1.1.3

Online Retailing Services in Kenya

In the last decade, Kenya has undergone a transformation in its information and communication technology (ICT) sector which has also had significant impact on Kenya‘s social and economic structures (World Bank (WB), 2010). The ICT sector‘s growth has outperformed every other sector, expanding by 23 percent annually during this time and is now six times its size at the beginning of the decade (Ibid). This remarkable growth has been characterized by introduction of various e-commerce products and services into the market, which target Kenya‘s rapidly growing internet population that stood at 14.032 million users in 2012 (CCK, 2012).

One of these innovative services is online retailing (or e-tailing), a subset of e-commerce where firms provide a platform for the purchase and sale of goods between consumers and sellers via the Internet (AGPC, 2011). Online retailing in Kenya is a beneficiary of the growing use of the internet by both businesses and consumers. A survey of 1700 individuals found that 18% to 24% of the respondents purchase music, movies and ebooks online (Juma, 2010), while reports indicate that the Communications Authority (CA) estimates the value of e-commerce in Kenya at Sh4.3 billion (Okuttah, 2014).

There are currently several online retailing firms in Kenya, majority of which are located in Nairobi. They include Ravenzo, Jumia, Mzoori.com, Rupu.com, OLX.com, amongst others. However, the usage of these services by consumers has not been commensurate with business projections, resulting in the closure of some of them (e.g. Kalahari.co.ke). There is therefore need to gain an increased understanding of what affects the use of these online retailing services at the individual context. For this reason, the online consumer context and more specifically, their usage of online retailing services in Nairobi, Kenya, is presented as the main context of this study.

8

1.2

Statement of the Problem

The remarkable growth in the Kenyan ICT sector in the last decade has been characterized by a surge in e-commerce activities, with several online services and applications being introduced into the market (World Bank, 2010). However, while the adoption of these online services is generally high, the conversion rate of the initial adopters to long-term users is very low, as noted by Magutu, Mwangi, Nyaoga, Ondimu, Kagu, Mutai, Kilonzo & Nthenya (2011).

This low usage of these online services by consumers poses a financial sustainability problem for service providers, who may incur undesirable costs of maintaining the lossmaking online service. Consequently, continued loss-making may eventually lead to closure of the online service, resulting in waste of effort to develop the service (Bhattacherjee & Parthasarathy, 1998). This has been the case in Kenya, where the failure of several online retailing firms – the latest being Kalahari.co.ke - has been largely attributed to financial losses as a result of low usage by consumers.

The poor usage of online retailing services in Kenya is also attested to by Nakumatt, the leading supermarket chain in Kenya and East Africa, which claims that it has felt no impact from e-commerce. The company decries the fact that despite its presence on online retail services, it was seeing little benefit if any there from, maintaining that online shopping is yet to take root in East Africa going by current turnover (COFEK, 2013). This low uptake of online retailing services in Kenya therefore signifies the need to understand what affects consumers‘ sustained usage of online retail services as a way of increasing the chances of success of these services in Kenya.

9

Past studies (Bajaj & Nidumolu, 1998; Liu & Forsythe, 2009; Bhattacherjee & Barfar, 2011) have thus sought to establish the determinants of online retailing services usage. In these studies, individual psychological factors - in particular customer perceptions - have been shown to have a significant effect on the usage (Whyte, Bytheway & Edwards, 1997; DeLone & McLean, 2003; Venkatesh et al., 2003).

A review of empirical literature in this area reveals the different types of customer perceptions and their relationship with usage. For instance, a study by Smith, (2008) showed that perceived attributes do affect online retailing usage. On the other hand, perceived risk has been established as having a significant effect on usage of ecommerce (Yildirim & Cengel, 2012). Likewise, perceived value has been established as one of the key factors affecting repeat usage in the online retailing context (Chen & Dubinsky, 2003; Hu & Chuang, 2012). However, no prior study has combined these three perceptions in the online retailing usage context. This study therefore contributes to knowledge in this area by doing so.

Customer satisfaction with an online retailing service also has an effect on its subsequent usage, as customers may discontinue usage due to unsatisfactory trial outcomes or usage experiences (Rogers, 1995; 2010; Bhattacherjee, 2001a; 2001b). Due to the similarity between re-purchasing products/services in a consumer context and the continued use of technology, an equivalent relationship in the continued technology usage context is posited. Moreover, satisfaction has also been shown to be affected by customer perception of their initial use experiences with an IT (Bhattacherjee, 2001a; 2001b). In other words, customer satisfaction has a mediating role on the relationship between customer perception and usage. There‘s therefore need to understand this mediating role

10

of customer satisfaction, particularly in the online retailing context in Kenya. This study therefore sought to fill this gap in extant research.

Further, the profile of consumers has also been shown to be of importance in understanding their continued usage of online services (Bhattacherjee & Parthasarathy, 1998). However, most IS studies have concentrated on user psychographic factors, ignoring demographic factors. To enhance the explanatory capacity of the proposed model, the moderating role of customer demographic factors was incorporated. Therefore, this research empirically examined the moderating role of three demographic characteristics - age, income and education level - on the relationship between customer perceptions and usage of online retailing services in Kenya.

1.3

Research Objectives

1.3.1

General Objective

The main objective of this study was to investigate the relationship between consumers‘ perceptions and usage of online retailing services in Nairobi County, Kenya.

1.3.2 i.

Specific Objectives To establish the relationship between perceived attributes and usage of online retailing services in Nairobi County, Kenya.

ii.

To analyse the relationship between perceived risk and usage of online retailing services in Nairobi County, Kenya.

iii.

To determine the relationship between perceived value and usage of online retailing services in Nairobi County, Kenya.

11

iv.

To assess the relationship between customers‘ perceptions and customer satisfaction with online retailing services in Nairobi County, Kenya.

v.

To establish the relationship between customer satisfaction and usage of online retailing services in Nairobi County, Kenya.

vi.

To establish the moderating effect of customer demographics on the relationship between customer perceptions and usage of online retailing services in Nairobi County, Kenya.

1.4 Research Hypotheses i.

H01: There‘s no relationship between perceived attributes and usage of online retailing services in Nairobi County, Kenya.

ii.

H02: There‘s no relationship between perceived risk and usage of online retailing services in Nairobi County, Kenya.

iii.

H03: There‘s no relationship between perceived value and usage of online retailing services in Nairobi County, Kenya.

iv.

H04: There‘s no relationship between customers‘ perceptions and customer satisfaction with online retailing services in Nairobi County, Kenya.

v.

H05: There‘s no relationship between customer satisfaction and usage of online retailing services in Nairobi County, Kenya.

vi.

H06: Customer demographics have no moderating effect on the relationship between customer perceptions and usage of online retailing services in Nairobi County, Kenya.

12

1.5

Significance of the Study

This study shall be of significance in the following ways:

1.5.1

Policy Significance

The findings will assist e-commerce stakeholders including service providers and the government in designing policies that are geared towards enhancing sustained use of online retailing services in Kenya. This will reduce the possibility of failure arising from low usage as it will enable online retailing companies to provide their services sustainably.

1.5.2

Practical Significance

The outcome of this study shall also be of significance to e-commerce practitioners in particular the managers who are responsible for developing and implementing strategies aimed at achieving a viable customer base. It is therefore important for online retailing firms to have a good understanding of their target customers, since this will not only help in determining the appropriate customer engagement strategies but also how to increase the long-term usage of their services.

1.5.3

Theoretical Significance

The academic value of this study is two-fold: First, the conceptual and empirical insights stemming from this study can be used to develop new knowledge, thereby helping both broaden and deepen researchers‘ understanding of consumer technology usage behavior, in particular, with regards to the online retailing context. Second, the study provides researchers with a rigorous and methodologically sound way of how to integrate quantitative and qualitative methods in order to contribute to a rich and comprehensive study. 13

1.6 Scope of the Study In line with Mottner, Thelen and Karande (2002), this research focused on online services that sell products and services to the end user or consumer, also known as Business-to-Consumer (B2C) e-commerce. This differs significantly from Business-toBusiness (B2B) e-commerce, which was not addressed in this study.

Seeing that the study focused on online retailing service providers in Nairobi, Kenya, it therefore restricted itself to online retailing services that are provided exclusively via the online channel (i.e. pure play and e-marketplace services). Due to the unique Kenyan context, results of this study will only be inferred to Kenyan online retailers, thus limiting their generalizability to other sectors and countries.

Further, the study‘s sampling pool was restricted to registered users of online retailing services, majority of who were highly educated and more conversant with internet use for shopping than the wider population of Kenyans. Therefore, to generalize the results for the larger population, it is suggested that in future, research should expand the scope of the current study by using a sample of both users and non-users of online shopping services.

The current study only considered individual aspects (i.e. customer perceptions, satisfaction and demographics) that affect online retailing usage, despite there being other usage determinants such as organizational and environmental factors, not to mention other psychological factors like motivation and learning (Kotler & Armstrong, 2000). Due to the subjective nature of perceptions, the study should be repeated with a different sample to ensure the validity of its findings.

14

1.7 Limitations of the Study Several limitations in this mixed-methods study are worth noting: First, quantitative data for this study was collected using a survey questionnaire and therefore suffers from biases such as non-response inherent in most survey-based research. To address this concern, the quantitative survey method was supplemented by qualitative key informant interviews for integration/triangulation in line with Bryman (2006).

The second limitation had to do with the sampling frame which poses a major challenge to internet surveys (Simsek, Veiga & Lubatkin, 2005). According to Wilson (2006), it is unlikely that the sampling frame/list will match the population of interest exactly, which will result in sampling frame error. This is compounded by the fact that few master directories exist that lists individuals (and their email addresses) from a particular population that has access to the internet, and the few that do exist may be seriously flawed. To reduce sample frame error, Wilson (2006) recommends adding a number of lists together to create the sample frame. For the current study, several lists of online retailing service providers were amalgamated in order to form the study population from which final sampling frame was derived. They include the Kenya ICT Board Tandaa Grant Applicants list, the Kenya Postel Directories list of E-Commerce service providers amongst others.

The third limitation regards the lack of representativeness arising from non-response. Internet surveys studies suffer from a lack of representativeness, which has to do with the extent to which the sample represents the population from which it was drawn. Representativeness may prove difficult to achieve using internet surveys for some population, particularly those where a large percentage of its members dislike the experience of participating in electronic surveys for various reasons (Simsek et al., 2005)

15

resulting in non-response problem. One way of reducing the number of non-responses in online surveys is to repeat the contact one or more times (Kumar, Aaker & Day, 2002). In this study, respondents were sent two reminders via e-mail to solicit their participation, as recommended by Nulty (2008).

The fourth limitation concerned the nature of the study‘s respondents. The respondents of this study reside in Nairobi County, the capital city of Kenya, which is more cosmopolitan and urbanized in comparison to the rest of the country. This makes it rather difficult to generalize the results to the rest of the country. Overall however, these limitations are not believed to have necessarily compromised the eventual findings.

1.8

Organization of the Study

This thesis is divided into five chapters. Chapter one, the introduction, consists of the research problem, research questions and hypotheses. The significance, scope and limitations were also outlined, finishing with the organization of the study.

Chapter two builds a theoretical foundation upon which the research is based by reviewing the relevant literature. The theoretical frame of reference of the study and links to relevant empirical discussions are presented. Also, the key concepts used in this study, i.e. customers‘ perceptions, customer satisfaction, usage and customer characteristics are discussed and presented as a conceptual framework. Finally, gaps within the literature are identified and linked to the research problem of this study.

Chapter three details the research methodology employed in this study and outlines the empirical research methods used. It covers research issues and hypothesis development, research design, model specification, data collection, scale development, sample

16

selection and size, research timing, data analysis, research validity and reliability as well as ethical issues.

Chapter four presents the findings of the study along with their discussion. It comprises data collection details as captured using the research questionnaire and other sources of secondary data as well as the analysis of those findings which are presented in the form of tables, figures, charts and narratives.

Chapter five provides a summary and discussion of the main findings of the study. It also outlines the conclusions, recommendations of the study, its contribution to knowledge as well as suggestions for further study.

17

CHAPTER TWO LITERATURE REVIEW

2.1

Introduction

This chapter reviews the theoretical as well as empirical literature on online retailing usage by customers. It begins with a review of five theories and models popular in IS and consumer behaviour, followed by an empirical review of the literature regarding the main concepts used in this study, i.e. customers‘ perceptions, customer satisfaction, usage and customer characteristics, which are then presented as a conceptual framework. Finally, a summary of relevant literature is conducted and subsequently gaps within the literature are identified and linked to the research problem of this study.

2.2

Theoretical Review

This study is underpinned by four theories and one model commonly used in explaining technology adoption and usage behavior. These are (i) Behavioral model of system usage (BMSU), (ii) Innovation diffusion theory (IDT), (iii) Expectation-confirmation theory (ECT), (iv) Perceived risk theory (PRT) and (v) Theory of consumption values (TCV). Amongst these, the BMSU and ECT are the dominant theoretical lenses used in this study.

2.2.1

Behavioral Model of System Usage

The Behavioral model of system usage (BMSU) is an early IS theory that was advanced by Schewe (1976) to explain end-user system usage behavior in organizational contexts. The parsimonious model attempted to predict system usage from perceptual, attitudinal and exogenous variables. It was developed in response to the need to explore how individual psychological factors and other behavioral aspects of the system user affect

18

MIS usage. Till then, attention to computer usage in the literature had focused on technical aspects of computer based information systems.

Constraints Evaluative Process

Beliefs A. About MIS Dimensions B. About MIS related objects, atmosphere and significant others

Beliefs Perceptual Processes

A. About MIS Dimensions B. About MIS related objects, atmosphere, and significant others

Attitude towards use of the MIS

System Usage

Figure 2.1: Behavioral model of system usage Source: Schewe (1976).

The attitudinal model can be reduced to four sets of variables: (1) perceptions of the MIS (system dimensions), (2) exogenous variables outside the MIS such as the perceived support and influence of IT personnel that normally may affect an individual‘s attitudes toward the system and system usage, (3) attitudes toward the MIS and (4) system usage (Schewe, 1976). The model was used to explore the relationships between MIS users' perceptions of their computer system, perceived variables exogenous to the system, attitudes, and system usage. In this study, the online retailing usage construct as well as the customer satisfaction construct (attitudes toward the MIS) are drawn from the BMSU.

19

The model presumed that a favorable attitude toward the use of an information system is central to obtaining high system use. However, the study by Schewe (1976) revealed that attitudes do not appear to determine individual usage behavior. This can be attributed to the fact that the study was conducted in an organizational setting and it is possible that there were constraints which intervene to override the influence of attitudes on behavior. While it does appear to show a direct relationship between usage and perceptions, Schewe (1976) found no significant relationship between system use and user satisfaction (Robey, 1979). This is corroborated in a study Mawhinney (1990), which found no relationship between user satisfaction and system use. Similarly, Lawrence and Low (1993) did not find this relationship to be significant.

IS researchers have not employed the BMSU as widely as other models in studies of continued use of IS by end-users. As such, empirical studies based on the BMSU have been confined to its traditional organizational context, thus limiting its usefulness in explaining IS usage behaviors in the consumer context. Perhaps, its low explanatory powers and factor inconsistencies may be due to the exclusion of important moderating variables reflecting individual differences, as argued by Sun and Zhang (2006). Its parsimonious nature implies that other variables that may be influencing system usage were not included in the model.

The main limitation of the model is that it was developed to explain individual MIS usage in the organizational context, which is generally perceived as mandatory (Shewe, 1976). This study therefore sought to extend its use in an individual context. To address its other shortcoming, the model was enhanced by incorporating additional perceptual factors as antecedents to usage in this study. Moreover, customer demographics were also included in this study as moderating factors to account for the interaction effect.

20

2.2.2

Innovation Diffusion Theory

Grounded in sociology, the Innovation Diffusion Theory (IDT) by Rogers (1962; 1995; 2003) is one of the first models to be employed in technology adoption research. It has been used since the 1960s to study a variety of innovations, ranging from agricultural tools to organizational innovation (Tornatzky & Klein, 1982). IDT describes how innovations (ideas, practices and technology) are spread into a social system network resulting in institutionalization of the innovation by incorporating it in routine practice/ continued usage (Murray, 2009). Based on this approach, Internet shopping is regarded as an innovation, which like other innovations takes time to spread through the social system (Alba, Lynch, Weitz & Janisqewski, 1997; Verhoef & Langerak, 2001).

Figure 2.2: Diffusion of innovation Theory Source: Rogers (1995)

21

The IDT focuses on the utility of an innovation - conceptualized as its perceived characteristics (attributes) - and posits that the rate of adoption is partially determined by the perceived attributes (or characteristics) of the innovation, and proposes several attributes potentially important across diverse innovation adoption domains. According to Rogers (1962; 1995; 2003), these perceived attributes (or core constructs) of this model

include

relative

advantage,

compatibility,

complexity,

trialability and

observability.

These attributes were later refined by Moore and Benbasat (1991) in their perceived characteristics of using an innovation (PCI) model for the IS context to study individual technology acceptance into relative advantage, compatibility, ease of use (instead of complexity), image, result demonstrability and visibility (instead of observability), and voluntariness of use. Another related model is the technology adoption model (TAM), whose two constructs, perceived usefulness and perceived ease of use, are quite similar to the IDT constructs - perceived relative advantage and perceived complexity (Davis, 1989; Al-Gahtani, 2001). Consequently, in this study, the perceived attributes construct (perceived usefulness, perceived compatibility and perceived ease of use) is drawn from the IDT, the related PCI model and TAM.

Empirical MIS studies based on the IDT have largely supported its predictive power (Fichman & Kemerer, 1999; Chircu & Kaufmann, 2000). For instance, an online shopping study of Dutch households by Verhoef and Langerak (2001) which explored the impact of relative advantage, compatibility, and complexity on e-shopping found that consumers‘ perceptions of relative advantage and compatibility positively influenced their intention to adopt online grocery shopping. Also, results obtained by Hansen (2005)

22

suggest that perceived complexity, perceived compatibility, and perceived relative advantage highly influence consumers‘ adoption of online grocery buying.

However, the theory has its limitations, the major one being that while it explains the formation of a favorable attitude toward a particular innovation, it does not provide further analysis of the attitude evolving into the adoption behavior (Chen, Gillenson & Sherrell, 2002).

2.2.3

Expectation-Confirmation Theory

The Expectation-Confirmation Theory (ECT) or Expectations Disconfirmation Theory (EDT) is a marketing theory that was first developed by Oliver (1977; 1980; 2010) and later refined by Churchill and Suprenant (1982). The theory focuses on the post-purchase behavior of individuals. It is a widely used in the consumer behavior literature, particularly in explaining consumer satisfaction and repeat purchase. According to this theory, a customer‘s initial expectations, combined with perceived product/service performance (confirmation), lead to post-purchase satisfaction. In addition, the positive or negative (dis)confirmation between expectations and performance mediates the effect on satisfaction (Oliver 1977; 1980; Churchill & Suprenant, 1982; Oliver, 2010).

The ECT was adapted to the IS context by Bhattacherjee (2001a), who constructed and empirically validated the Expectation-Confirmation Model (ECM) based on the ECT to predict IS continued usage in a consumer context. It lays emphasis on a user‘s psychological motivations that materialize post initial adoption of IS. According to the model, users‘ intention to continue to use an IS are dependent on three antecedent constructs: user satisfaction, user confirmation, and post-adoption expectations (or perceived usefulness).

23

Expectations

Disconfirmation

Satisfaction

Repurchase Intention

Perceived Performance

Figure 2.3: Expectations-Confirmation Theory Source: Oliver (1977; 1980).

Bhattacherjee (2001a) describes the process by which IS users reach a continued use decision is as follows. First, users form a conception of perceived usefulness after using a particular IS for a period of time. Second, users compare the performance of the IS to the perception of usefulness (so as to determine to what extent their perception of usefulness about that IS has been confirmed). If the user finds that the product/service is as useful as he/she perceived, confirmation is formed and the user forms a notion of satisfaction. Finally, satisfied users are more likely to continue the usage of that IS whereas dissatisfied users intend to discontinue the service (Bhattacherjee (2001a, as cited in Hossain & Quaddus, 2012).

Notable use of the ECT has been made in an effort to better understand end-user satisfaction with IS and consumer-oriented online services (Bhattacherjee, 2001a; Nevo & Furneau, 2009). One of the key streams of IS research uses ECT to explain the adoption and continued use of IS and relies on the premise that these behaviors are result from users‘ satisfaction. For example, Wixom and Todd (2005) applied ECT to the study

24

of usage intentions in the context of data warehousing while Bhattacherjee (2001a) used it to study continuance intentions among online banking users (Nevo & Furneau, 2009).

In this study, the customer satisfaction construct as well as the usage construct are drawn from the ECT (Oliver, 1980; 2010) and the related ECM by Bhattacherjee (2001b). Satisfaction is characterized by ECT as either an outcome state following the consumption experience or more widely as an evaluative process encompassing the entire consumption experience (Yi, 1990). With regards to continued usage, due to the similarity between re-purchasing products/services in a consumer context and the continued use of technology by consumers, ECT posits an equivalent relationship in the continued online usage context (Bhattacherjee, 2001a; 2001b).

According to Bhattacherjee and Barfar (2011), ECT has been criticized by Ortiz de Guinea and Marcus (2009) for one of its underlying premise that views IT continuance as an intentional, reasoned or purposeful behavior; this ignores the role of emotions and habit. To address this shortcoming, emotional value has been incorporated in this study as an antecedent to usage.

2.2.4

Perceived Risk Theory

The Perceived Risk Theory was first introduced by Bauer (1960) to explain consumer behavior. According to this theory, consumers perceive risk because they face uncertainty and potentially undesirable consequences as a result of purchase or usage of products/services. This means that the more risk consumers perceive, the less likely they will purchase/use a product or service (Bhatnagar, Misra & Rao, 2000). The perceived risk construct in this study is derived from the perceived risk theory and adapted to the online retailing context.

25

Performance Risk

Financial Risk

Opportunity/Time Risk

Perceived Risk

Safety Risk

Social Risk

Psychological Risk

Figure 2.4: The six dimensions of perceived risk Source: Cunningham (1976).

The core constructs of the theory have been decomposed by researchers into several perceived risk dimensions. For instance, Cunningham (1967) conceptualized six dimensions of perceived risk: performance, financial, opportunity/time, safety, social, and psychological risk, while Jacoby and Kaplan (1976) came-up with six components of consumers‘ perceived purchase risk: Performance Risk, Financial Risk, Physical Risk, Convenience Risk, Social Risk, Psychological Risk. Bhatnagar et al. (2000) have argued that two types of risk exist when buying over the internet; product risk and financial risk. These risks are thought to be present in every choice situation but in varying degrees, depending upon the particular nature of the decision (Taylor, 1974). Moreover, different individuals have different levels of risk tolerance or aversion (Bhatnagar et al., 2000).

26

Perceived risk has been applied in various studies of the consumer technology use context. For instance, an early study of telephone shopping by Cox and Rich (1964) found that consumers perceive higher risks in new innovative channels. In the ecommerce context, perceived risk has been applied in studies such as internet banking adoption (Tan & Teo, 2000), usage of e-commerce services (Liebermann & Stashevsky, 2002) continued usage of internet banking (El-Kasheir, Ashour & Yacout, 2009), online consumers‘ purchasing behavior (Zhang, Tan, Xu & Tan, 2012) amongst others.

2.2.5

Theory of Consumption Value

The Theory of Consumption Values (TCV) is a consumer behavior theory that was developed by Sheth, Newman and Gross (1991a; 1991b). Over the years, TCV has evolved into a popular marketing theory and has been widely applied in various contexts, including IS. The theory focuses on explaining why consumers choose to use or not to use a specific product or service, arguing that consumer decisions are made based on perceived value.

The TCV has five core constructs which are conceptualized as five different types of values (functional value, social value, epistemic value, and emotional value, and conditional value) that underlie consumer choice behavior. A particular choice may be determined by one value or influenced by several values (Sheth et al., 1991a; 1991b). In this study, the perceived value construct is drawn from the TCV by Sheth et al. (1991a; 1991b) and adapted to the online retailing context.

27

Functional Value

Conditional Value

Social Value

Consumer Choice Behavior

Emotional Value

Epistemic Value

Figure 2.5: The five values influencing consumer choice behavior Source: Sheth, Newman, and Gross (1991b).

Kalafatis, Ledden and Mathioudakis (n.d.) re-specified three fundamental propositions that underpin the TCV: (1) consumer choice is a function of multiple consumption values; (2) the values make differential contributions in the choice situation, and (3) the values are independent of each other. Thus, all or any of the consumption values can influence a decision and can contribute additively and incrementally to choice; consumers weight the values differently in specific buying situations, and are usually willing to trade-off one value in order to obtain more of another.

TCV‘s strong point is its analytical strength, which helps practitioners to understand consumer decision making. This enables them to develop practical strategies that address real market conditions (Gimpel, 2011). TCV has been used in several IS studies on technology adoption decisions (Kim, Lee & Kim, 2008).

28

However, the theory‘s main limitation is due to the fact that it applies only in cases of individual, voluntary and rational or systematic decision situations (Sheth et al., 1991a, 1991b); therefore, it cannot be used to predict the behaviour of two or more individuals and is thus restricted to individual end-user/consumer acceptance contexts.

2.3

Empirical Literature Review

The empirical review is made up of related literature regarding the hypothesized relationships between the various study constructs: i) customers‘ perceptions, ii) usage, iii) customer satisfaction and iv) customer demographic characteristics. These relationships are discussed in the following sections.

2.3.1

Usage of Online Retailing Services by Consumers

A review of usage literature shows that it is a complex construct with multiple conceptualizations (Burton-Jones & Straub, 2006). As such, the context within which it is used is important in determining its relevant measures and dimensions. In this research, the usage of online retailing services is studied at the individual level (i.e. the consumer context) as a post-adoption phenomenon that is also referred to as continued use. Therefore, usage serves as the dependent variable in line with Bhattacherjee and Barfar (2011), who argue that the goal of IT continuance research (or IT post-adoption research) is to predict actual behaviors (and not intentions). In their opinion, continuance research should operationalize and measure IT usage behavior rather than end at intention.

According to Turner, Kitchenham, Brereton, Charters and Budgen (2010), the actual usage of information technology can be measured using both objective and subjective forms. Objective measures are usually generated from logs of usage generated by the

29

software itself. In comparison with objective measures of actual usage, subjective measures of usage are based on the individual opinion that is usually established using a questionnaire. Examples of subjective measures of technology use include self-reported usage measures of the frequency or intensity of using the technology in question, in line with Legris, Ingham and Collerette (2003).

Consequently, for this study, usage of online retailing services shall be measured objectively, instead of through perceptual measures such as use intentions and selfreported use. Usage behavior shall be obtained from system log records in line with Venkatesh et al. (2003) who in their longitudinal study captured system data from four organizations over a six-month period and measured actual usage behavior over a six month period as duration of use via system logs.

While there are several factors that affect usage behaviour, this study restricts itself to examining the antecedent role of customers‘ perceptions on usage of online retailing services, in line with extant research that has established the significant relationship between perceptions and system usage. Some of the earliest studies were carried out by Lucas (1975) and Schewe (1976). For Lucas, who carried out an extensive MIS user behavior study, favorable user attitude and perceptions of IS lead to high levels of use of the same. On the contrary, Schewe established that user satisfaction is not associated with IS usage.

Bhattacherjee (2001b) carried out an empirical analysis of the antecedents of continued usage of an e-commerce service. The study developed a model of B2B e-commerce continuance behaviour which was empirically tested in a survey of online brokerage users. The study found that consumers' continuance intention is determined by their

30

satisfaction with initial service use as well as its perceived usefulness. However, it only used the ECT as its basis and thus lacked moderating variables. By employing continuance intention as the DV, the study also failed to assess actual usage.

In another study, DeLone and McLean (2004) adapted their updated model of IS success to an e-commerce context through an extensive review of relevant literature. According to the model, e-commerce usage is directly influenced by satisfaction with the ecommerce system. The major weakness of their study was its conceptual nature since it failed to undertake empirical testing of the proposed relationships between the six dimensions. Instead, two case studies were used to demonstrate how the model can be used to guide the identification and specification of e-commerce success metrics.

Barnett, Kellermanns, Pearson and Pearson (2006-2007) replicated and extended the landmark study of technology acceptance and use conducted by Straub, Limayem and Karahanna-Evaristo (1995). The study empirically examined the TAM and clarified conceptual ambiguities that had hampered system usage research. It found that perceived ease of use was a significant predictor of objective system use, while perceived usefulness was a significant predictor for self-reported usage behavior. The study‘s key weakness was its basis on the TAM, which is more suited for initial usage context.

Bhattacherjee, Perols and Sanford (2008) conceptually extended the IT continuance model by adding continuance behavior as the DV instead of BI. Data for the study was collected via a longitudinal survey of document management system usage among administrators and staff personnel at a governmental agency in Ukraine. However, the study failed to demonstrate the strong intention-behavior association in the IT continuance context, since intention explained only 26% of the variance in continuance

31

behavior. It therefore called for additional constructs that may be able to predict continuance behavior better.

Petter, DeLone and McLean (2008) carried out a meta-analysis where they summarized empirical studies that have investigated the relationship between satisfaction and usage constructs. They classified the level of support for the relationship as strong, moderate, or mixed in order to summarize the empirical results across all studies. According to their analysis, 17 of 21 studies showed a positive, moderate support for the satisfactionuse relationship, 1 showed a positive mixed, while the rest (3) showed no support at all. However, this study failed to address the relationship between customer perceptions and usage.

A recent study by Ramayah and Lee (2012) investigated the role of the users‘ satisfaction in influencing e-learning success by empirically establishing the impact of satisfaction on e-learning usage among students in a public university in Malaysia. The study employed an adaption of the DeLone and McLean‘s (2002, 2003) extended model, and analyzed the response data using the structural equation modelling (SEM) method. The findings of the study supported the hypothesis that user satisfaction is positively related to usage continuance. However, since it was based on the Delone and McLean model (2003), the study fails to address the role of moderating variables in the perception-usage relationship.

2.3.2

Antecedent Role of Customers’ Perceptions on Online Retailing Service Usage

To induce more initial adopters to continue using their services, service providers have to establish what motivates consumers to return repeatedly. One of the factors found in

32

prior consumer technology adoption studies to influence continued use is the customers‘ perceptions (Venkatesh & Davis, 2000; Venkatesh et al., 2003). In general, one of the ways in which perceptions towards e-commerce are determined is through the user‘s past online experiences (Im, Kim & Han, 2008); these perceptions subsequently go on to influence consumers‘ purchase decisions (Osman, Yin-Fah & Hooi-Choo, 2010). In this study, the customer perception construct serves as the independent variable and has theoretical foundations from three perceptual constructs identified in literature as playing an antecedent role via-a-vis usage of online retailing services. These are perceived attributes, perceived risk and perceived value.

2.3.2.1 Relationship between Perceived Attributes and Usage of Online Retailing Services Perceived attributes have been found to influence consumer behavior vis-à-vis technology use. The antecedent role of perceived attributes/characteristics of an innovation on its adoption was first suggested by Rogers (1962) and was later refined by Moore and Benbasat (1991) in the context of individual IS usage. A review of literature reveals several conceptualizations of the perceived attributes construct. In this study, it has three dimensions (perceived usefulness, perceived compatibility and perceived ease of use) drawn from work on the TAM (Davis, 1989), DOI (Rogers, 1995), PCI (Moore & Benbasat, 1991).

Usefulness of an online retailing service enhances consumer shopping activities and consists of aspects such as the relevance of information, the breadth of offerings amongst other things. Compatibility has to do with the fit between the individual current circumstances (e.g. experience, values, needs, and habits) and the features of the online retailing service. A lack of compatibility hampers the adoption of innovations (Rogers,

33

1995). Ease-of-use relates to the simplicity with which consumers can operate an online retailing system to perform shopping activities such as browsing, communicating and carrying out actual transaction such as ordering and payment. It is the polar opposite of complexity.

A review of prior literature reveals different usage outcomes based on the antecedent role of the perceived attributes of an IS. For instance, Parthasarathy and Bhattacherjee (1998) empirically examined usage in the online service context and found that the perceived service attributes such as usefulness and compatibility determine usage behavior.

Moreover, Bhattacherjee‘s (2001b) empirical analysis of the antecedents of e-commerce service continuance demonstrated that perceived usefulness is a key determinant of customer‘s continued usage intention (CUI). This study‘s main weaknesses are its employment of the ECT as its sole basis as well as its lack of moderating variables. It also used continuance intention as the DV, which has been criticized as being weakly correlated with actual use.

A study by Saeed and Abdinnour-Helm (2008) examined the effects of IS characteristics and perceived usefulness on post-adoption usage of information systems. The context was that of a web-based student information system that students use to manage their academic work. Data was collected from 1032 respondents and used to empirically test the model. The results showed that perceived IS usefulness is a good predictor of postadoption usage. However, despite testing for the moderating role of gender and experience, it failed to examine the moderating effect of age, income and education level on model relationships.

34

El- Kasheir et al. (2009) empirically established factors affecting continued usage of internet banking among Egyptian customers. The study, which was based on several intentional models and employed a sample of users of internet banking services, found perceived ease-of-use to be the strongest predictor of intentions to continue usage of internet banking services. This study‘s main weakness was its use of continued intention instead of actual usage as the DV since it has been criticized as being a poor proxy for actual use.

2.3.2.2 Relationship between Perceived Risk and Usage of Online Shopping Services Perceived risk is a subjective consumer behavior concept that relates to the uncertainty and consequences associated with a consumer‘s action. A perception of risk with regards to a particular activity/transaction (e.g. purchasing or using a product or service) dissuades a consumer from taking further action in that regard (Bhatnagar et al., 2000). The notion of perceived risk as a key antecedent to consumer behavior has been established in prior research. For instance, Sharma, Durand and Gur-Arie (1981) showed that the willingness to purchase products is inversely related to the amount of perceived risk associated with a purchase decision.

By and large, perceived risk is conceptualized as a multi-dimensional construct in several studies (Cox & Rich, 1964; Jacoby & Kaplan, 1972; Bettman, 1973; Bhatnagar et al., 2000, Zhang et al., 2012). In this study, the perceived risk construct has three dimensions that have been derived from a review of relevant literature. These are i) financial risk (Jacoby & Kaplan, 1972; Bettman, 1973, Bhatnagar et al., 2000), ii) performance risk (Jacoby & Kaplan, 1972; Bettman, 1973) and iii) personal/privacy risks drawn from work by Jarvenpaa and Todd (1997).

35

Financial risk has to do with the potential financial loss a consumer is likely to incur as a result of overpaying (being overcharged) or due to fraud (Bhatnagar et al., 2000) whereby what is paid for is not received. Consumers assume financial risk when paying for products/services electronically e.g. in online shopping or electronic auctions. Performance risk has to do with concerns that the online service or the desired (or even paid for) product will not function as expected. For consumers who provide personal information during online transactions, the risk of having this information compromised through identity theft and credit card information going to the hands of hackers comprises personal/privacy risk. For many shoppers, online retail raises concerns about privacy and security. These concerns have often been cited as potential barriers to online retail (Forsythe & Shi, 2003; Shim et al., 2001).

Moreover, in the online retailing context, the intangible nature of online transactions poses a risk for consumers, impeding further use of online purchasing services (Bhatnagar et al., 2000; Hansen, 2007). Previous research on its antecedent role also suggests that perceived risk negatively impacts internet shopping. For instance, Liebermann and Stashevsky (2002) investigated the role of perceived risks as barriers to e-commerce usage in Israel amongst both users and non-users. The study, which employed a cross-sectional design and had a sample of 465 employed adults, only considered barriers to e-commerce usage, which were mapped as perceived risk components. The model

was tested empirically against field data and showed that Internet credit card stealing and supplying personal information (i.e. privacy risk) affects both current and future e-commerce usage.

36

However, an empirical study by El-Kasheir et al. (2009) on factors affecting continued usage of internet banking among Egyptian bank customers established that that perceived risk had no relationship with customer continued intention to use the service. Data was collected from 65 respondents using mall interception technique and multiple regression analysis was used to test the research hypothesis. This study‘s main weaknesses was its employment of mall interception to collect data as well as its use of continued intention instead of actual usage, since it has been criticized as being a poor proxy for actual use. In the context of B2C e-commerce, a study by Zhang et al. (2012) on the dimensions of consumers‘ perceived risk and their influences on online consumers‘ purchasing behavior demonstrated five independent dimensions (perceived health risk, perceived quality risk, perceived time risk, perceived delivery risk and perceived after-sale risk) which affect significantly online consumers‘ purchasing behavior. The results also showed that the other three dimensions - perceived privacy risk, perceived social risk and perceived economic risk - are the less relevant factors. 2.3.2.3 Relationship between Perceived Value and the Usage of Online Retailing Services Perceived value is a broad and abstract concept comprised of various components (Bolton & Drew, 1991) that refers to the benefits ascribed to the purchase/use of a product or service. It‘s a widely used business concept that aggregates perceptions about product/service benefits and tradeoffs and is thus considered as the pivot in relationship marketing and customer loyalty (Casalo, Flavian & Guinaliu, 2008). Therefore, understanding consumers‘ value perceptions of their online experience is crucial in enhancing the use of internet as an alternative marketing channel (Andrews et al., 2007). Accordingly, it has attracted a lot of attention from both marketing and IS fields as a significant determinant of consumers‘ decision-making behavior in various contexts, including usage of online retailing services.

37

Prior studies contend that perceived value is a complex construct that is multidimensional in nature (e.g. Sheth et al., 1991b; Sánchez-Fernández & Iniesta-Bonillo, 2007). Accordingly, the perceived values construct in this study has four dimensions drawn from relevant literature, namely i) monetary value, ii) convenience value, iii) social value and iv) emotional value.

Monetary value derives from the customer‘s evaluation of the total financial cost of using the online retailing service (including the price paid) relative to the benefits received. Convenience value pertains to the perceived ease with which users are able to find information and/or products on the online retailing service, pay and, have them delivered as expected by the consumer (i.e. minimization of the overall shopping effort), while social value is derived from the collective/group significance ascribed to the use of the e-commerce by an individual. According to the concept, usage of online retailing services should be congruent with the norms of a consumer‘s friends or associates. On the other hand, emotional value is related to various affective states induced by usage of online retailing services. These can be positive (e.g., confidence or excitement) or negative (e.g., fear or anger).

Similarly, consumers in the online shopping context also have diverse perception of value, as argued by Hu and Chuang (2012) in their study of the relationship between value perception and loyalty (re-purchase) intention toward an e-retailer website in Taiwan. SEM was used to test the data from 243 students and 418 workers. The findings concluded that utilitarian value is more important than hedonic value in terms of influencing loyalty intention for online shopping. The study thus recommended the use

38

of various website elements and attributes as a way of offering either more hedonic or more utilitarian value to online buyers in order to attract and retain more visitors.

Yen (2011) carried out an empirical study on the impact of perceived value on continued usage intention in social networking sites (SNS) amongst savvy Facebook users in Taiwan. Mediated regression analysis was used to analyze data from 205 respondents. The findings revealed that PV, including information value, sociable value, and hedonic value, has a positive impact on CUI. This study‘s limitation is due to the fact that it only considered three ―get‖ values for measuring PV, ignoring ―give‖ components such as monetary value.

It would therefore be expected that if an individual perceived an e-commerce system to have a high value, (s)he would be more willing to try (accept) and use it. Therefore, identifying and delivering value for potential customers is crucial for e-commerce service providers in order to induce continued use of their services by consumers.

2.3.3

Mediating Role of Customer Satisfaction on the Relationship between Customer Perceptions and Usage of Online Retailing

Customer satisfaction is another multi-attribute concept that is originally based on a study by Katz (1960) explaining the role of attitudes in shaping social behaviour. According to the study, the underlying dimensions of attitude include: affect (feelings), behaviour (actions), and cognitions (learning and beliefs). It has evolved into a widely used business concept that aggregates customer evaluations about a product/service. Consequently, understanding of what creates a satisfying customer experience has become crucial, more so due to its role in influencing repeat customers behavior (e.g. repurchase, continued usage).

39

While customer satisfaction has been a popular topic in marketing practice and IS research, it is also one of the most controversial concepts due to its multiple conceptualizations and meanings. Despite many attempts to measure and explain customer satisfaction, there still does not appear to be a consensus regarding its definition. This lack of a clear and broadly accepted conceptual and operation definition, has resulted in the arbitrary development of satisfaction measurement instruments, and conclusions about interactions with other constructs are problematic (Caruana, 2002).

For this study, customer satisfaction is used a post-initial usage evaluation and is treated as a mediating variable influencing the relationship between customer perceptions and continued usage of online retailing services. The customer satisfaction construct in this study is drawn from the ECT (Oliver, 1977; 1980, 2010) and related ECM by Bhattacherjee (2001a) which adapted the ECT to the IS context. IS researchers have made notable use of ECT in an effort to better understand end-user satisfaction with information systems and related services. In this regard, prior studies use ECT to explain the adoption and continued use of IS and rely on the premise that these behaviors are the result of user satisfaction (Bhattacherjee, 2001a; Nevo & Furneaux, 2009; Hossain & Quaddus, 2012).

In the service context, customer satisfaction can be viewed from the i) transactional and/or ii) cumulative/relational orientations/perspective. In earlier studies, satisfaction has been defined from transactional perspective (e.g. Oliver, 1980; 1993), where it is based on a one time, specific post-purchase evaluative judgment of a service encounter. However the conceptualization of satisfaction as a customer‘s overall/cumulative

40

assessment/evaluation construct based on purchase and consumption experiences over a time period (Bitner & Hubbert, 1994) has become more dominant in research.

In terms of the diagnostic and predictive value of satisfaction measurement, cumulative satisfaction is more useful and reliable determinant of long-term use than transactionspecific in that it is based on series of purchase and consumption occasions rather than just one occasion of transaction. This is also the case in the IS context, where satisfaction is often regarded as the basis of system usage continuance, while dissatisfaction may cause users to discontinue the system use (Bhattacherjee, 2001a; Nevo & Furneaux, 2009). Therefore, this study employs the cumulative approach to conceptualize satisfaction which is evaluated from the time of registration with the e-commerce service.

Giese and Cote (2000) reviewed relevant satisfaction literature with the view of identifying its conceptual domain. They defined the customer as the ultimate user of a product/service and came to the conclusion that satisfaction was comprised of three basic components, (i) a response (cognitive or emotional) pertaining to (ii) a particular focus (expectations, product, consumption experience, etc.) determined at (iii) a particular time (after consumption, after choice, based on accumulated experience, etc).

The mediating role of customer satisfaction in the service context was demonstrated by Bolton and Lemon (1999), who developed and empirically tested a dynamic model of customers‘ usage of services, where usage was employed as both an antecedent and consequence of satisfaction. The aim of the study was to identify causal links among customer‘s prior usage, satisfaction evaluations, and subsequent usage. The study was significant in that it provided an in-depth examination of the dynamic relationship

41

between customer satisfaction and customer usage. According to its findings, customers who have high levels of cumulative satisfaction with a continuously provided service in the current time period will have higher usage levels of the service in a subsequent time period.

Bhattacherjee (2001b) carried out an empirical analysis of the antecedents of electronic commerce service continuance. The study examined the key drivers of consumers' intention to continue using B2B electronic commerce services using multiple theoretical, synthetic perspectives to develop a model of continuance behaviour which was empirically tested using a field survey of online brokerage users. The study found that consumers' continuance intention is determined by their satisfaction with initial service use as well as their perceived usefulness of service use.

Devraj, Fan and Kohli (2002) carried out a study on the antecedents of B2C e-commerce satisfaction where they developed and empirically tested a model for consumer satisfaction with the e-commerce channel using constructs from TAM, TCA and SERVQUAL. Subjects purchased similar products through conventional as well as EC channels and reported their experiences in a survey after each transaction. The findings of the study demonstrated the influence of perceived ease of use and perceived usefulness on satisfaction in the B2C e-commerce context.

DeLone and McLean (2003; 2004) conceptually extended the DeLone and McLean IS Success Model (1992) to measure e-commerce success. The mediating role of customer satisfaction was put forward in their study, where they proposed that ease of use influences user satisfaction which subsequently directly influences usage of e-commerce services.

42

A study by Serenko, Turel and Yol (2006) in the mobile phone services established that customer satisfaction was influenced by perceived value. With regards to the C2C context of e-commerce, Jones and Leonard (2007) carried out an empirical study which was aimed at establishing the factors that impact satisfaction in C2C e-commerce. According to the study‘s findings, TAM, TCA, and SERVQUAL all impact satisfaction in C2C e-commerce. On the other hand, Ortiz de Guinea and Marcus (2009) noted that satisfaction may drive IT usage directly.

Chen, Mocker, Preston and Teubner (2010) investigated the confirmation of expectations and satisfaction with the internet shopping context of e-commerce in Taiwan by integrating various theories (ECT, SCT, and TAM) and testing the research hypotheses using empirical data that was collected from a sample of 342 responses. The results suggest that satisfaction is influenced by perceived usefulness, and that both satisfaction and perceived usefulness determined consumer‘s repurchase intention.

The mediation effect of end-user satisfaction on the relationship between PV and CUI was established in an empirical study by Yen (2011) on the impact of perceived value on continued usage intention in social networking sites (SNS) amongst savvy Facebook users in Taiwan. Mediated regression analysis was used to analyze response data from 205 users. The findings revealed that the mediation effect of end-user satisfaction is only significant to social value and hedonic value but not information value. This study however didn‘t consider/ignored ―give‖ components such as monetary value.

As evidenced in the literature, customer satisfaction is critical for retaining current users/customers. As a result, a fundamental understanding of the role of customer

43

satisfaction is of great importance in e-commerce. This study therefore seeks to establish whether customer satisfaction has a mediating effect on the customer perception-ecommerce usage relationship among consumers in Kenya.

2.3.4

Moderating Effect of Customer Demographic Characteristics on the Relationship between Customer Perceptions and Usage of Online Retailing Services

Explaining repurchase intention (in this case continued usage) using satisfaction alone may not suffice (Capraro, Broniarczyk & Srivastava, 2003) since it has been reported that only 15–35% of satisfied customers return (Reichheld, 1996). It is therefore important to examine the role of potential moderators in attaining a better understanding continued system use in the online context. Of particular interest to this study is the moderating role of demographic factors.

The employment of moderators may potentially increase the predictive validity of a model under investigation, and explain the inconsistent findings in various disciplines (Judge & Bono, 2001). For this reason, moderator variables have enjoyed a surge of popularity in the marketing literature in recent years, and scholars have acknowledged their importance for predicting consumer behaviour (Baron & Kenny, 1986; Sharma et al., 1981). The importance of moderators arises from their ability to enhance understanding of the relationship between relevant independent variables and dependent variables, as well as seemingly established relationships (Walsh, Evanschitzky & Wunderlich, 2008).

In the IS context, Sun and Zhang (2006) argue that low explanatory powers and factor inconsistencies of IS models may be due to the exclusion of important moderating variables reflecting individual differences. Several studies have therefore called for the

44

inclusion of individual characteristics into the study of individual technology acceptance and use (e.g. Venkatesh et al., 2003). In view of that, customer demographic factors will serve as the moderating variables of this study.

According to Hoffman, Novak and Schlosser (2000), key elements of a consumers‘ demographic profile that have been found to influence their online behaviour include variables such as income, education, age. Moreover, age, income, education offer significant information regarding the demographic characteristics of the targeted population. For this reason, these three variables have been included in previous studies that examined the usage of various information systems, including online retailing services. This study shall therefore seek to establish the moderating role of 3 demographic measures (i.e. age, income and educational level) on the relationship between customer perceptions and usage of online retailing services.

Previous online purchasing research has examined the three demographic characteristics. For instance, Bhatanager et al. (2000) examined the moderating effect of age amongst other demographic factors (gender, marital status and years on the internet) in a previous study on how risk, convenience and demographics affect internet shopping behavior. They found mixed results on the moderating effect of age on internet shopping behavior. Other online purchasing studies report that e-commerce purchasers are younger, more educated and have higher income than non online purchasers (Ratchford, Talukdar & Lee, 2001).

In the context of mobile phone services, an empirical investigation into the moderating roles of demographic variables, namely age, income, and gender, in forming perceptions and behavioral outcomes with was carried out by Serenko et al. (2006). Structural

45

equation modeling techniques and split-sample approach for moderation analysis were applied to a dataset of 1,253 mobile phone users in the U.S. While age and income were found to have a significant effect on several of the model‘s relationships, the study proposed that gender has a very limited effect on the examined relationships in the mobile services context. However, the study didn‘t go as far as to examine the customer perception – usage relationship.

However, a study by El-Kasheir et al. (2009) that empirically established factors affecting continued usage of internet banking among Egyptian bank customers found that demographic variables such as age, gender, marital status, education level and income level had no relationship with customer continued intention to use the service. Analysis of variance (ANOVA) was used to test the research hypothesis that differences in the demographic characteristics of the bank customers can affect their continued intention to use internet banking. It main weakness was its use of mall interception techniques of data collection, which is prone to selection bias.

More recently, Hernández, Jiménez and José Martín (2011) conducted a study on whether age, gender and income really moderate online shopping behaviour. The study‘s aim was to establish if individuals‘ socioeconomic characteristics – age, gender and income – influence their online shopping behaviour. The individuals sampled were experienced e-shoppers i.e. individuals who often make purchases on the internet. The results showed that socioeconomic variables moderate neither the influence of previous use nor the perceptions of e-commerce; in short, they do not condition the behaviour of the experienced e-shopper.

46

The conflicting outcomes of various studies point to the need for more research in this important area. Therefore, this study seeks to establish the moderating influence of customer characteristics on the perception-usage link in online retailing services. 2.4

Summary of Empirical Literature and Research Gaps

This section contains a summary of literature reviewed regarding usage of online retailing services and research gaps. Appendix 6 presents a summary of the empirical reviews as well as the important research gaps that have been identified in this section of the study.

2.5

Conceptual Framework

The conceptual framework for this study is made up of various consumer behavior and technology adoption constructs, their variables, indicators and is based on the premise that customer perceptions have an effect on usage of online retailing services but this effect is mediated by customer satisfaction and moderated by demographic factors as described in the empirical literature review in the previous section. The framework is a graphical representation of how these constructs of interest are interconnected and is depicted in Figure 2.1.

The presumed interrelationships amongst the study‘s constructs are organized into three sub-models. The first proposed relationship models the main effects of the predictor variable (customer perceptions) on the criterion variable (usage of online retailing services). Since the customer perception variable is made up of three constructs (perceived attributes, perceived risk and perceived value), the study proposes that each of them has a direct effect on the DV (usage of online retailing services).

47

INDEPENDENT VARIABLE

MEDIATING VARIABLE

CUSTOMER SATISFACTION

H4

- Level of Satisfaction

CUSTOMER PERCEPTIONS H5

Perceived Attributes - Usefulness - Compatibility - Ease of Use

H1

Perceived Risk - Financial risk - Performance risk - Personal (Privacy) risk

ONLINE RETAILING SERVICE USAGE H2

Perceived Value - Monetary value - Convenience value - Social value - Emotional value

- Active Use - Inactive Use

H6

DEPENDENT VARIABLE

H3

DEMOGRAPHIC FACTORS - Age - Income - Education level

MODERATING VARIABLE

Figure 2.6: Schematic Diagram Source: Researcher (2013).

48

The second relationship is mediation effect model of customer satisfaction on the relationship between the predictor and criterion variables.

The main concern of

mediators is how cognitive mechanisms operate. In short, mediation ―explain[s] how external physical events take on internal psychological significance‖ (Baron & Kenny, 1986). As argued by Wu and Zumbo (2008), mediators have dual roles — an outcome variable (effects) of the independent variable and an independent variable (causes) that occur before the dependent variable. In this case, it proposed that customer perception has a direct effect on customer satisfaction, and subsequently, customer satisfaction has a significant effect on usage of online retailing services.

The mediation effects model for this study is based on Spencer, Zanna and Fong (2005) concept of ―experimental-causal-chain design‖, whereby the mediator typically functions like a DV for the manipulated IV and in the same way, the mediator is subsequently manipulated to act like an IV for the outcome variable (usage). According to this concept, a researcher conducts two separate manipulated experiments; one aims to establish the causal relationship of X on M, and the other aims to establish the causal relationship of M on Y. They point out that the strength of this design is that by manipulating both the independent variable and the mediator, one can make strong inferences about the causal chain of events in psychological processes.

The third relationship models the interaction effect of the moderating variable (customer demographics) on the relationship between the predictor and criterion variables. According to Wu and Zumbo (2008) a moderator‘s function is to explain the strength and direction of the causal effect of the IV on the DV. It thus serves as a supplementary

49

variable for improving a hypothesized bi-variate causal relationship, and less as a causal variable responsible for the outcome effect. In regard to causal order, a moderator variable is prior to the dependent variable and has no causal relationship with the independent variable (figure 2.1). In short, moderation specifies various conditions under which the direction and/or strength of the relationship varies (Baron & Kenny, 1986). For this study, three demographic factors — age, income and education level — are conceptualized as having a moderating effect on the customer perception – usage relationship.

50

CHAPTER THREE RESEARCH METHODOLOGY

3.1

Introduction

This chapter presents the research methodology. It includes the research design, empirical model of the study, target population, sampling and sample size, data collection instruments. In addition, the techniques that were employed in analyzing data as well as the ethical issues arising from the study are outlined.

3.2

Research Philosophy

The philosophical perspective grounds the methodological logic and criteria for a research study (Crotty, 2003). For this study, the philosophical foundation employed was pragmatism, a practical substitute to positivism and anti-positivism (Goldkuhl, 2004). This is because pragmatism does not restrict one‘s choice to between positivism and interpretivism insofar as methods, logic and epistemology are concerned (Creswell, 2003; Pansiri, 2006).

Essentially, pragmatism presupposes that objectivist and subjectivist perspectives are not mutually exclusive. Hence, a mixture of ontology, epistemology and axiology is acceptable to interrogating social phenomena - of importance is what works best to address the research problem. Pragmatist researchers thus prefer to use both quantitative and qualitative data as this allows for better grasp of social reality (Wahyuni, 2012). Instead of laying emphasis on methods, the research problem is considered as the most important issue; this frees researchers to choose the methods, techniques and procedures of research that best meet their needs and purposes (Creswell, 2003). 51

In practical terms, pragmatism embraces the two extremes:- quantitative methods espoused by positivism/post-positivism and qualitative methods espoused by interpretivist proponents (Pansiri, 2006). For this reason, pragmatism is regarded as the basis of mixed-method research (Tashakkori & Teddlie, 1998; Teddlie & Tashakkori, 2003) popular in the social and behavioural sciences (Maxcy, 2003; Pansiri, 2006).

Consequently, this study employed a similar pragmatic approach. As a result of the pragmatic approach adopted by the study, both method and data triangulation were achieved through the use of qualitative as well as quantitative methods of data collection. In this study, survey served as the main quantitative data collection method, supplemented by key informant interview as the qualitative data collection method, in line with Bryman (2006).

3.3

The Research Design

According to Sekaran and Bougie (2010), the research design addresses important issues relating to a research study such as purpose of the study, location of the study, type of investigation, extent of researcher interference, time horizon and the unit of analysis. In view of this, this study adopted a mixed design made of descriptive – specifically cross-sectional survey design – and explanatory research design (Saunders, Lewis & Thornhill, 2009) to establish the relationship between customers‘ perceptions and the usage of online retailing services in Nairobi, Kenya.

Descriptive design involves assessing the study phenomena without the ability to control or manipulate variables, and thus require the researcher(s) to collect data and determine relationships without inferring causality (Swanson & Holton, 2005). Saunders et al. (2009) argue that descriptive studies should be regarded as a means to

52

an end and not an end in itself. Consequently, this study employed description prior to delving into explanation of phenomena.

Explanatory research design serves to test hypotheses derived from the theory and thus test causality between the independent and dependent variable (Saunders et al., 2009). The explanatory approach of the study which followed tested the causal relationships between the variables to determine their significance.

3.4

The Empirical Model

The model used in this study is based on the premise that customer perceptions have an effect on usage of online retailing services but this effect is mediated by customer satisfaction and moderated by demographic factors. These causal relationships between the study‘s variables were organized into three sub-models that are explained in the following section:

3.4.1

The Direct Effects Model

The first sub-model was the direct-effects model, whose relationship is the direct effect of the predictor variable (customers‘ perceptions) on the criterion variable (usage of online retailing services). Since the customer perception variable was made up of three constructs (perceived attributes, perceived risk and perceived value), the study presumed that each of them has a direct effect on the DV (usage of online retailing services).

The direct effect model was empirically analyzed using the logistic regression (Logit) analysis method, a statistical technique of modeling the non-linear relationship between continuous IVs and dichotomous dependent variables (Liou, 2008). Logistic regression has become an important statistical procedure employed in various social and 53

behavioral studies (Moosbrugger et al., in press) to estimate the coefficients of a probabilistic model involving a set of independent variables that best predict the value of the dependent variable. A positive coefficient increases the probability, while a negative value decreases the predicted probability of the outcome being in either of the two dependent categories (Mazzarol, 1998).

Logit analysis was deemed suitable for this study because of the binary/dichotomous nature of the dependent variable y (usage), which can have either of two values representing different outcomes categories: the value 1 which denotes active user with a probability of P, or the value 0 which denotes inactive user with a probability of 1 - P. The empirical model estimated the conditional probability that the dependent variable is either one or the other. This is illustrated in the Equation 1, where Pi is the conditional probability of observing whatever value of y is observed for a given observation.

……………………………. (1)

The formula for the probability itself is equation 2 shown below

P( y  1) 

e

B0  B1 X1  B2 X 2 .....  Bk X k 

- (B0  B1 X1  B2 X 2 .....  Bk X k  ) 1 e

(2)

Whereby: y = The dichotomous dependent variable y = Estimated regression equation = B0 + B1X1 + B2X2 +…+ BkXk + ε1 P (y = 1) = The conditional probability of an individual being classified as belonging to either of two outcome categories: 1 or 0 54

e = Exponential, the quantity 2.1828+, the base for natural logarithms X1 - Xk B0 = Intercept Term B1- k = Logistic regression coefficients for predictor variables X1 - Xk = Predictor variables ε1= Error Term

Consequently, the logistic regression model that was used to establish the direct effects of the predictor variables on the criterion variable is expressed as:

logit P( y  1)  B0  B1X1  B2 X 2  B3 X 3   ………. (3)

Whereby: y = The dichotomous DV (usage of online retailing services) with 1 (active user) or 0 (inactive user). y = Estimated regression equation = B0 + B1X1 + B2X2 + B3X3 + ε1 P (y = 1) = The conditional probability of an individual being classified as belonging to either of two outcome categories: 1 (active user) or 0 (inactive user).

e = Exponential, the quantity 2.1828+, the base for natural logarithms X1, X2, and X3 B0 = Intercept Term B1- 3 = Logistic regression coefficients for predictor variables X1 = Perceived Attributes X2 = Perceived Risk X3 = Perceived Value ε1= Error Term

55

It is noted that the left hand side of the equation is not the dependent variable, y, itself; but the so-called ‗log odds‘ or ‗logit‘ of y.

3.4.2

The Mediated Effect Model

The second relationship in the conceptual research model is the mediation effect of customer satisfaction on the relationship between the predictor and criterion variables. The mediation effect of customer satisfaction (M) on the relationship between the predictor and criterion variables was obtained from two regression equations (4 & 5) in line with Spencer et al., (2005). The first equation (Equation 4) used simple linear regression, a data analysis technique for identifying underlying correlations among data in research (Nimon, 2010). Equation 4 is illustrated below:

M = B0 + B1P1 + ε1 ………....................................................................………..…..... (4)

Where: M = Customer Satisfaction (Dependent variable) B0 = Constant B1 = Linear regression coefficients P1 = Customer Perceptions (Composite Value) ε1= Error Term In equation 5, the mediating variable, customer satisfaction (M) was re-conceptualized as an independent variable affecting usage. The effect of customer satisfaction on usage was established using Equation 5, a binary logistic regression shown below.

Logit [ p ( y = 1) ] = B0 + B1M + ε1 ….............…………...………..………..…..... (5)

56

Where: Y= Online Retailing Service Usage P (y = 1) = Probability of belonging to either 1 or 0 B1 = Logistic regression coefficient M = Customer Satisfaction (Mediating Variable) ε1= Error Term

Equations 4 and 5 represent simple cases of correlational and are thus useful for making statements about the relationship between two variables. Neither is useful for establishing evidence of causality; to do this, one must also attempt to establish proper time order and to control for potentially confounding variables.

3.4.3

The Interaction Effect Model

The third relationship modeled the interaction effect of the moderating variable (demographic factors) on the relationship between the independent variable (customer perceptions) and the dependent variable (usage). For this study, three demographic factors — age, income and education level — were conceptualized as a composite value Z. This was obtained using a logistic regression equation. In order to do this, three data computation procedures were performed. First, a composite value (CPER) of the IV (customer perceptions) was derived. Second, the moderating variable Z (demographic factors) was also computed as a continuous composite value (DEMF) of the three demographic factors (age, income and level of education) employed in the study. Third, the researcher computed a new variable to represent the interaction; this is called CPERDEMF; it is obtained by forming the product CPER × DEMF. Subsequently, a logistic regression was performed using CPER, DEMF, and CPERDEMF as predictors of usage.

57

Accordingly, the interaction effect of the moderating variable Z (demographic factors) on the relationship between the predictor and criterion variables was obtained using the following logistic regression equation:

Logit [ p (y=1) ] = B0 + B1P1 + B2Z + B3(P.Z) + ε1 …….………………………. (6) Where:

Y = Usage of online retailing services p (y = 1) = Probability of belonging to either 1 or 0 B1- 3 = Logistic regression coefficients P1 = Customer Perceptions (Composite Value) Z = Moderating Variable (Customer Demographics) PZ

=

Interaction Effect

ε1= Error Term

In total, three empirical models were estimated. The first model (equations 1, 2 & 3) reported the results for the direct effects only, while the second model (equations 4 & 5) introduced the mediation effects of the mediating variable M, whereas the third model (equation 6) captured the interaction effects of the moderating variable Z on the IV-DV relationship.

3.5

Operationalisation and Measurement of Study Variables

This section defined the criterion, predictor, mediating and moderating variables used to operationalize this study as well as the measures used in their assessment. Table 3.1 presents a summary of the different variables, their indicators, their operational definitions and instruments used to assess each of the variables.

58

Table 3.1 Operationalization and Measurement of Study Variables Predictor Variable Variable

Online Retailing Service Usage

Indicator

Continued Usage

Usefulness

Perceived Attributes

Compatibility

Ease-of-Use

Nature

Operationalisation

Criterion/ Dependent Variable (DV)

The current utilization of one or more features of e-commerce services by registered users of online retailing firms.

Independent Variable (IV)

The degree to which consumers believe that using an e-commerce service will enhance their activities.

4 items; 7 point Likert scale. Davis 1989; Davis et al., 1989)

Independent Variable (IV)

The degree to which using an e-commerce service is perceived as being consistent with the values, needs, and habits of potential user.

4 items; 7 point Likert scale (Moore & Benbasat, 1991)

Independent Variable (IV)

Independent Variable (IV)

The degree of anxiety regarding the perceived financial loss as a result of e-commerce usage.

Independent Variable (IV)

The level of uncertainty regarding the perceived performance of an ecommerce system.

Personal risk

Independent Variable (IV)

The level of anxiety regarding the perceived compromise/insecurity of personal information as a result of ecommerce usage.

Monetary value

Independent Variable (IV)

Convenience Value

Independent Variable (IV)

Financial risk

Perceived Risk Performance risk

Perceived Value

The degree to which an e-commerce system is perceived as relatively easy to understand and use.

The financial utility derived from ecommerce usage The time, place and execution utility derived from usage of

59

Measure 1 item, Binary, 1 = Active Usage, 0 = Inactive use; (Venkatesh et al., 2003)

6 items; 7 point Likert scale (Davis 1989; Davis et al., 1989; Forsythe et al., 2006) 3 items; 7 point Likert scale Zhang et al., (2012); Forsythe et al. (2006) 5 items; 7 point Likert scale. Forsythe et al. (2006)

3 items; 7 point Likert scale Forsythe et al. (2006); Tan (1999)

4 items; 7 point Likert scale; (Sweeney & Soutar, 2001) 5 items; 7 point Likert scale; (Chang et al., 2012; Mosavi & Ghaedi, 2012)

Question No.

Hypothesized direction

N/A

_

B1-B4

Positive

B5-B8

Positive

B9-B14

Positive

B15-B17

Negative

B18-B22

Negative

B23-B25

Negative

B26-B29

Positive

B30-B34

Positive

the e-commerce system

Social value

Customer Satisfaction

Demographic Factors

Independent Variable (IV)

The perceived favorability/ approval derived from one‘s social milieu regarding e-commerce usage

The utility derived from the feelings or affective states that an ecommerce service generates

Emotional value

Independent Variable (IV)

Level of Satisfaction

Mediating Variable (MV)

The level of satisfaction with the e-commerce system

Age

Moderating Variable (MV)

Age group

Income

Moderating Variable (MV)

Level of Income

Education

Moderating Variable (MV)

Level of Education

5 items; 7 point Likert scale (Tan,1999; Sweeney & Soutar, 2001)

7 items; 7 point Likert scale (Thompson et al., 1991; Compeau & Higgins 1995b; Compeau et al. 1999, Sweeney & Soutar, 2001) 4 item, 5 point Categorical scale (Westbrook, 1980; Oliver & Bearden, 1983; Oliver & Westbrook, 1982; Swan et al, 1981) Ordinal 1 item; Ordinal (Venkatesh et al., 2003)

B35-B39

B40-B46

C1- C5

1 item; Ordinal (Hernández,2011)

A2.

Positive

1 item; Ordinal

A3.

Positive

The area of study was Nairobi County in Kenya. Nairobi, one of the 47 counties in the country, is unique in that it serves as the capital city of Kenya (Institute of Economic Affairs (IEA), 2011). It is the largest city in East Africa, with a population of over three million. Nairobi is also an international, regional, national and local hub for commerce, transport, regional cooperation and economic development, connecting eastern, central and southern African countries (United Nations Human Settlements Programme (UN-

60

Positive

Negative

The Study Area

HABITAT), 2006).

Positive

A1.

Source: Researcher (2013)

3.6

Positive

As the capital of the country and the seat of national government, the city generates over 45% of the national gross domestic product (GDP) and is thus a major contributor to the Kenya‘s economy. It provides employment for its residents and commuters from its environs, employing 25% of Kenyans and 43% of the country‘s urban workers (UNHABITAT, 2006). At present, the city is growing faster than ever as it develops into a regional economic center with the potential to become Africa‘s next major business hub (IBM, 2012).

The rationale for choosing Nairobi County as the area of study was three-fold. First, Nairobi is a cosmopolitan and urbanized area in comparison to the rest of the country. Second, there are several local as well as international firms that offer e-commerce services. Third, the primary target market for most of these firms is Nairobi and its environs (i.e. the Nairobi Metropolitan Area), given that the bulk of internet users in Kenya reside in Nairobi.

3.7

Target Population

The target population for the study was made up of online retailing firms in Nairobi, Kenya. To arrive at the number of online retailers in Nairobi, the researcher relied on Kenya ICT Board (2012) records and Kenya Postel Directories (2012), which show that there are 25 registered online retailing firms in Nairobi, Kenya (see Appendix 2). However, the accessible ones were 6, which formed the target population for this study. Accordingly, the respondents for this study were the 18,147 registered users drawn from these six online retailing firms in Nairobi, Kenya.

This included 12 key

informants who are regarded as expert sources of information (Marshall, 1996). They included owners/CEOs, senior managers and employees of these online retailers as well

61

as consultants who were identified with the help of the online retailing firms as being particularly knowledgeable and accessible.

Table 3.2: Distribution of Target Population No 1.

Target Population Firm 1

2.

No. of Respondents 4868

931

Firm 2

3.

Firm 3

4.

Firm 4

5.

Firm 5

6.

Firm 6

Total

6

6470

1909

1447

2522 18,147

Usage Category Active Users

1022

Inactive Users

3846

Active Users

111

Inactive Users

820

Active Users

2076

Inactive Users

4394

Active Users

278

Inactive Users

1631

Active Users

138

Inactive Users

1309

Active Users

363

Inactive Users

2159

Total

18,147

Source: Kenya ICT Board (2012); Kenya Postel Directories (2012)

The distribution of six online retailing firms and their respective numbers of registered users is illustrated in Table 3.2, whereby the users are categorized into usage categories (active and inactive users). 3.8

Sampling Design and Procedure

Sampling design and procedure for this study concerned the sampling techniques used in the study to determine a representative sample from the general population. These are explained in the following sections.

62

3.8.1

Sampling Technique

The study employed both probability and non-probability sampling techniques to draw samples from the target population. To achieve this, the study used the nested sampling approach which involves using sample members selected for one stage of the study as a sampling frame for choosing a subset/sample for an ensuing stage of the study (Collins, Onwuegbuzie & Jiao, 2007). In this study, sampling for the questionnaire phase preceded that of the interview, thus providing a sampling frame for the smaller, more focused qualitative phase. In total, the sampling frame for this study was made up of the 18, 147 registered members (active and in-active users) of six online retailing firms in Nairobi, Kenya. This is illustrated in Table 3.3.

Stratified random sampling, a probability sampling technique, was used to select the sample from the 18, 147 respondents. Stratified random sampling was employed because the sampling frame was not homogeneous since the sample contained subgroups thereby necessitating a fair representation of these sub-groups in the sample size (Ahuja, 2005). This technique ensures that observations from all relevant strata are included in the sample (Lemm, 2010). Stratified sampling also guarantees that every possible sample matches the population distribution on strata-defining characteristics (Mallet, 2006).

To this end, proportional stratification technique was initially employed to select the elements from the respective strata. In proportional stratified sampling, the population is divided into groups or strata. Samples are then selected by strata, in proportion to strata sizes (Mallet, 2006). A sample with proportionate stratification is chosen such that the distribution of observations in each stratum of the sample is the same as the distribution of observations in each stratum within the population (Lemm,

63

2010). Proportionate stratification uses the same fraction (multiplier) for each subgroup (Latham, 2007) to insure representation of all sub-groups. The sampling fraction, which refers to the size of the sample stratum divided by the size of the population stratum (n/N), is equivalent for all strata (Lemm, 2010). This is illustrated in Table 3.3.

Subsequently, a sample was randomly drawn from each strata and categories using random sampling method. In simple random sampling, every possible combination of population elements is equally likely to be selected (Mallet, 2006), thereby eliminating possible bias. For this study, a computerized random number generator was used to select the respondents out of the whole population. This was aimed at eliminating bias in the sample selection.

Non-probability sampling was used to select participants for the key informant interview using the stratified purposeful sampling scheme from among those who participated in the questionnaire-based survey. In this technique, the sampling frame is initially separated into strata to obtain fairly smaller homogeneous groups from which a purposeful sample is selected from each stratum (Collins et al., 2007).

For this study, there were 6 strata which were based on the number of online retailing firms that participated in the study. The sample of key informants was purposively selected from each stratum and consisted of the registered users of these firms who had expressed an interest in participating in a follow-up interview.

Potential key informants were identified with the help of representatives of the online retailing firms and thereafter contacted by the researcher. The basis for selection was (i) their level of activity on the website, (ii) having adequate information about online retailing and (iii) willingness to participating in a follow-up interview. Campbell (1955) 64

recommended that key informants should (1) take up roles that make them conversant with the researched focus and (2) be available to share their insights with the researcher. They are however not considered as representing the sampled units statistically (John & Reve, 2001). For that reason, selection of potential key informants was based on their correspondence with the selection criteria enumerated by Campbell (1955).

3.8.2

Sample Size Determination

The sample size was determined using Yamane‘s (1967) formula for calculating the sample size since it is relevant to studies where probability sampling is used. According to the formula, n is the sample size, N is the population size and e is the margin of error. A 95% confidence level and e = 0.05 were assumed for the equation in this study.

n =

N 1 + N (e)2

For this study, N = 18,147 and ε = 0.05. At 95% confidence level, this translated to a sample size of 391 respondents out of a target population of 18, 147.

n=

18,147

= 391.37 ~ 391

1 + 18,147 (0.05*0.05) Selection into the sample was based on two key parameters of interest. First, to be considered for this study, individuals must have been registered users of the selected online retailing firms. Second, the selected individuals comprised those who are registered on the service for more than three months. Table 3.3 shows the sampling frame as well as the distribution of users sampled from the respective online retailing firms.

65

Table 3.3: Sampling Frame and Sample Distribution

Size

Frequency

Category

Active Users 1. Firm 1

2. Firm 2

3. Firm 3

4. Firm 4

5. Firm 5

6. Firm 6

Total

Proportionate Sample

Population

Strata

4868

931

6470

1909

1447

2522

1022

Inactive Users

3846

Active Users

111

Inactive Users

820

Active Users

2076

Inactive Users

4394

Active Users

278

Inactive Users

1631

Active Users

138

Inactive Users

1309

Active Users

363

Inactive Users

2159

Total

18,147

18147

Multiplier

0.0215 0.0215 0.0215 0.0215 0.0215 0.0215 0.0215

Category

Frequency

Active Users

22

Inactive Users

83

Active Users

3

Inactive Users

18

Active Users

45

Inactive Users

94

Active Users

6

Inactive Users

35

Active Users

3

Inactive Users

28

Active Users

8

Inactive Users

46

Total

391

Source: Researcher (2013)

3.9

Data Collection Instruments

Both primary and secondary data was collected in this study. Primary data was collected using a mixed-mode approach with the help of two instruments: (i) a selfadministered questionnaire and (ii) an interview guide for key informants. The questionnaire was used to collect quantitative data while the interview guide was used for qualitative data collection. Mixed-mode approach has several advantages. According to De Leeuw (2005), it is an affordable option that compensates for the flaws inherent in each individual mode, it can provide more choice and flexibility for respondents, while improving timeliness and minimizing non-response and nonresponse bias. Furthermore, it is also more efficient as compared to other alternatives.

66

3.9.1

Self-Administered Questionnaire

In essence, the questionnaire was composed of three different sections (A, B and C) and it consisted of 54 questions that were close-ended with ordered responses. The measures were adopted from previous studies (Swan et al., 1981; Oliver & Westbrook, 1982; Oliver & Bearden, 1983; Thompson et al., 1991; Compeau & Higgins 1995b; Compeau et al. 1999, Sweeney & Soutar, 2001; Venkatesh et al., 2003); Hernández et al., 2011) and reworded to suit the context of the current study. The measures were organized based on the research questions and specific objectives. Each scale item was modelled as a reflective indicator of its hypothesized latent construct

The survey questionnaire was marked because the respondents were divided into two groups (active users and inactive users) depending on whether or not one was using online retailing service at the time of the survey. Depending on the usage, the respondent was presented with a corresponding survey questionnaire. A sample of the final questionnaire is shown in Appendix 3B of this thesis.

3.9.2

Key Informant Interview Guide

The key informant interview was a follow-up to the questionnaire survey. Therefore, qualitative data was collected using the key informant technique with the help of an interview guide (Appendix 3C).

For this study, the interview guide was composed of 9 questions that were open-ended. Some of the measures were adopted from previous studies as well as the findings of the quantitative study, and some additional open questions to elicit more information if possible so as to suit the research questions and specific objectives. A sample of the final interview guide is available in Appendix 3C of this thesis.

67

The first category of questions (questions 1 – 5) was based on the two variables (satisfaction and usage). The second category (questions 6 & 7) explored 9 constructs namely: usefulness, convenience, ease of use, financial risk, performance risk, privacy risk, monetary value, convenience value and social value. The questions were aimed at ascertaining if the variables do affect the satisfaction and usage of online retailing services. The third category (questions 8 & 9) contained provocative questions that were meant to elicit more information that was required regarding the state of the online retailing sector in Kenya as a whole.

In addition to the primary data, secondary data was collected from a variety of both print and online sources including published reports, book chapters, sectoral directories and trade publications such as industry magazines and newsletters.

3.9.3

Validity of the Data Collection Instruments

Validity of the questionnaire was assessed using both criterion-related and content validity. Criterion-related validity was demonstrated through the questionnaire items which were derived from measures validated in prior research (Nunnally & Bernstein, 2004; Cooper & Schindler, 2008) and standardized and adapted to the context of this study as is indicated in Table 3.1. Content validity was achieved using a panel of five experts in the field who were asked to give their views and suggestions on how to improve the questionnaire (Nunnally & Bernstein, 2004; Cooper & Schindler, 2008). The five experts evaluated the questionnaire and found that the questions were relevant to the study variables.

68

In addition, the questionnaire was pilot-tested on 30 selected respondents who are online retailing users and who were subsequently excluded in the main survey. Pilottesting of the questionnaire was conducted before the actual research so as to get an indication of the expected responses with a view to identifying ambiguous and unclear questions as well as to detect possible weaknesses in the design and instrumentation as suggested by Cooper and Schindler (2008).

Some of the pre-testers voiced their concerns regarding the length of the questionnaire, wondering whether respondents‘ attention could be maintained. Notwithstanding these concerns regarding the number of items, all were found to be relevant and therefore none was deleted. However, some items relating to convenience value, social value and level of satisfaction had to be improved due to their perceived similarity. An example of this questionnaire can be found in Appendix 3B.

For the key informant technique, the main threats to validity or credibility that were identified were addressed through the methods recommended by Creswell and Miller (2000).

These methods include triangulation, peer review/debriefing and member

checking.

The first techniques that was employed in enhancing validity of the qualitative study was the use of multi-method strategies for data collection and analysis (method and data triangulation) as recommended by Easterby-Smith et al. (2002). It was realized through the use of a combination of qualitative as well as quantitative methods of data collection including document analysis, surveys and interviews. Gray (2004) affirms that triangulation is a practical means of strengthening of a study since its aids in overcoming the major flaws of the respective methods employed.

69

The other technique used was peer review or debriefing, which is described by Creswell and Miller (2000) as ―the review of the data and research process by someone who is familiar with the research or the phenomenon being explored.‖ For this study, the two supervisors, two e-commerce managers as well as two IT consultants were engaged as peer reviewers. Before the interviews were carried out, the reviews were asked to evaluate the interview guide so as to ensure that the questions were clear, relevant and comprehensive. Some changes regarding the logical flow and choice of respondents were made to the initial sample based on the feedback received. During the thematic analysis, they were also asked to assess the proposed categories, themes and interpretations so as to improve the grasp of the study‘s findings by ensuring that it provides a rational and practical account of the phenomena.

The final technique through which the validity of findings was established is member checking. Member checking is a process for obtaining feedback from a few key informants in which collected data is ‗played back‘ to the informant to check for perceived accuracy and reactions (Cho & Trent, 2006). It occurs throughout the research process. For this study, the researcher sent research participants a copy of the documented interview to confirm that it reflected their perspective of the subject matter. After the thematic analysis, the researcher prepared a brief summary of the findings and shared it with the available key informants.

3.9.4

Reliability of the Data Collection Instruments

For this study, Cronbach‘s Alpha coefficient (α) statistical procedure was used to assess reliability of the quantitative measures as recommended by Mugenda and Mugenda (2003). As a rule of thumb, reliability of 0.7 and above is recommended for most research purposes to denote the research instrument as reliable (Roberts, Priest & 70

Traynor, 2006). Using this cut-off value, all but one of the measures in the questionnaire exhibited internal consistency by having Cronbach‘s alpha values greater than 0.7. The single variable that did not meet this cut-off is social value, which had a Cronbach‘s alpha value of 0.538. However, the Cronbach‘s Alpha for social value does meet the guidelines suggested by Hair, Anderson, Tatham and Black (2006), who recommended reliability level of 0.5 and above.

Table 3.4 Reliability of Questionnaire Items Factor

Perceived Attributes

Measure

Reliability (Cronbach‘s alpha) 0.954

Usefulness

4 items (B1 – B4)

Compatibility

4 items (B5 – B8)

0.954

Ease-of-Use

4 items (B9 – B14)

0.950

Financial Risk

3 items (B15 – B17)

0.801

Performance Risk

5 items (B18 – B22)

0.702

Personal Risk

3 items (B23 – B25)

0.885

Monetary Value

4 items (B26 – B29)

0.839

Convenience Value

5 items (B30 – B34)

0.954

Social Value

5 items (B35 – B39)

0.538

Emotional Value

7 items ( B 40 – B46)

0.975

Level of Satisfaction

5 items (C1 – C5)

0.941

Perceived Risk

Perceived Value

Customer Satisfaction

Number of Items

Source of data: Survey (2013)

As a result, this variable was not excluded in the data analysis in spite of its weak reliability, meaning that all the item-to-total correlation values exceeded the 0.50 cutoff value suggested by Hair et al. (2006). The remaining ten variables had an alpha value of 0.7 and above indicating that they were reliable as indicated by Roberts et al. (2006).

71

For the key informant interviews, the main threats to reliability/dependability were the lack of standardization of the interview processs, the choice of informants/ informant bias as well as interviewee/participant bias (Saunders et al., 2009).

One of the techniques used by the researcher to build reliability with regards to interview standardization was usage of an interview schedule to guide the interview process. The researcher also outlined the procedures employed in the study, especially the data gathering procedures and processes. This will enable future researchers to replicate the work, and perhaps arrive at similar outcomes (Shenton, 2004).

Informant bias resulting from the sort of entities that were approached and agreed to participate in the interview was overcome through the sampling approach that was used (Saunders et al., 2009) as well as the criteria used to select interview participants which stipulated the nature of persons identified as informants (LeCompte & Goetz, 1982).

Interviewee or response bias may result from the interview participants saying what they thought their superiors required them to say out of fear of losing their job. One of the ways through which it was addressed was by ensuring that the interviews were anonymous. Further to this, the responses were presented in a way that they could not be linked to a specific organization or individual so as to jeopardize participants‘ anonymity (Elfving & Sundqvis, 2011).

3.10

Data Collection Procedures

To begin with, the researcher received a letter of introduction from Kenyatta University Graduate School to facilitate the study. Afterwards, a research permit approving the researcher to conduct the research was sought from the National Commission for

72

Science, Technology and Innovation (NACOSTI). In addition, the authorization and consent to collect information from study respondents was sought and granted from the management of the sampled online retailing firms before administering the questionnaire to the respondents.

Consequently, primary data was collected using two methods: a cross-sectional, mixedmode survey of users of online retailing services in Nairobi, Kenya for quantitative data that was supplemented with in-depth interviews of key informants for qualitative data; secondary data was collected via a review of existing literature and data from relevant print and online sources.

3.10.1 Quantitative Data Collection Procedures Quantitative data collection began by contacting respondents via e-mail and telephone to inform them of the nature and objective of the study as well as to solicit their participation in the survey. To ensure maximum response, the questionnaire was administered using a mixed-mode method, as recommended by De Leeuw (2005). Despite the fact that there have been reservations regarding mixed-mode methods of questionnaire administration, it seems not to have a statistically significant effect on findings (De Leeuw & Hox 2011).

To begin with, the questionnaire was sent to respondents via e-mail as an attachment to the selected respondents. This method was preferred as it was cheaper and allowed for faster data collection (Cooper & Schindler, 2008). Emailing ensures significant cost savings in terms of postage and paper materials can be made (Phellas, Bloch & Seale, 2012). It has also been shown to be useful in surveying individuals who may be unwilling to participate in personal or phone interviews, but who might respond to an email survey when it is convenient to them (Simsek et al., 2005). 73

Each questionnaire was sent with a covering letter (Annexure 3A) from the researcher explaining the purpose of the study and soliciting their participation. The questionnaire was marked so as to differentiate between active and inactive users. It included a brief introduction that explained the purpose of the study as well as solicited the recipients‘ participation. After the questionnaire had been distributed, email reminders were sent at a weekly interval to those who did not respond so as to try and ensure that higher response rates were achieved. Rogelberg & Stanton (2007) found that higher response rates result in findings which have greater credibility among key stakeholders.

In order to increase the survey response rate (RR), the e-mail questionnaires were followed-up with a self-administered questionnaire that was delivered to nonrespondents using the drop-and-pick method as recommended in previous studies (Rojas-Méndez & Davies, n.d., Ibeh et al., 2004), whereby respondents are contacted in person and asked to fill in a questionnaire at their most convenient time. Picking up the completed questionnaires was scheduled at a specified time as recommended by Paxson (1992). This allowed for personal contact with the respondent and provided the opportunity to explain the purpose of the survey and thereby increase the motivation to respond. The respondent may also seek clarifications during such instances, thus improving the quality of the data that is collected (Phellas et al., 2012).

The drop-and-pick method was preferred because it reduces non-response bias through reduction of non-coverage, non-contact or refusal to participate (Paxson, 1992). On average, two follow-ups were made before the questionnaires were ready to be picked for analysis.

74

3.10.2 Qualitative Data Collection Procedures On the other hand, qualitative data collection was conducted using key informant interviews with the help of an interview guide that was developed by the researcher based upon the findings from the quantitative data analysis and the literature review that had been carried out previously. It is a useful data collection technique (Homburg, Klarmann, Reimann & Schilke, 2012), that can either be used on its own or in combination with other methods (Marshall, 1996). One of the advantages of using the key informant technique is in connection with the quality of data that can be collected in considerably short period of time (Lincoln & Guba, 1985).

The researcher conducted interviews where a set of questions based on themes arising from the literature review as well as the quantitative data analysis were asked; it included some previously formulated open-ended questions aimed at eliciting pertinent information as recommended by Creswell (2003). Following Marshall‘s (1996) suggestion, subjects who were identified with the help of the participating firms were invited to participate through the official cover letter (Appendix 3A) that was prepared and delivered at their offices. Those who responded positively by expressing were later contacted via telephone to arrange a convenient date and place for the interview.

The face-to-face interviews took the researcher about 40 minutes to conduct and were held at the respondents‘ office for the sake of their convenience. The interview schedule was delivered prior to the actual interview to give them ample time to prepare. The interview tried to capture respondents‘ explicit and implicit knowledge of the subject matter and participants were free to deviate; the interviewer intervened only to clarify issues or move on to a new theme. The sample interview schedule is in Appendix 3C.

75

The interview guide was composed of 9 open-ended questions that allowed for interviewees‘ comments. The questions – which were adopted from previous studies explored the same themes as in the questionnaire and were reworded to suit the context of the current study. They were organized based on the research objective, specific objectives as well as research hypotheses (H1 - H6).

The method used for recording interviews for documentation and later analysis was note taking whereby the researcher took down detailed notes so as to ensure that insightful comments and details are not forgotten or lost. At the end of the interview, the interviewer reviewed the notes for legibility and coherence. Where this was not the case, the interviewer asked the respondent to clarify his/her remarks. In took close to 10 weeks to collect the required data for the study.

3.11

Data Analysis and Presentation

Due to the mixed-method that was used to collect data, the study employed a sequential technique to analyse and present the data. This involves sequential analysis of quantitative and qualitative data, whereby the initial results of the analysis of the first data set are used to inform the analysis of the second set of data (United States Agency for International Development (USAID), 2013). The section below expounds further.

3.11.1 Quantitative Data Analysis Quantitative data was analysed using both descriptive and inferential statistics. Descriptive statistics that were used include frequency of the distribution, mean and standard deviation. Descriptive data analysis prepares the data for further inferential analysis. Inferential statistics in this study involved conducting both logistic and linear regression analysis of the response data to test the causality of the IV and DV.

76

Prior to the analysis, the data collected during the research was coded and entered into SPSS to create a dataset for analysis. Afterwards, variable re-specification was undertaken by collapsing the response data into fewer variables in line with the specific research objectives/questions. All the research variables were then defined and label in a codebook, as recommended by O‘Neill (2009). Appendix 5A illustrates the codebook listing all the variables included for the statistical analysis, as well as their labels and the codes ascribed to each answer category given.

Before regression analysis, diagnostic tests comprising multi-collinearity and normality test were carried out so as to establish if the independent variables (IVs) in the study model are inter-related. Multi-collinearity arises when two or more independent variables are highly correlated with each other (De Fusco, 2007). For this study, the collinearity test was conducted using correlation analysis, tolerance and variance inflation factor analysis. Normality was also tested for the linear regression equation using the Kolmogorov-Smirnov (KS) test. According to Ul-Islam (2011), a key goal in linear regression in checking for normality is to ensure that the t-statistic is giving us the correct message that whether the independent variable is a significant explanatory variable or not. All the above analysis were done using SPSS version 19.

The variables, tested hypotheses, regression models as well as the expected outcomes are summarized in Table 3.5. In additionally, diagnostics tests were carried out on both the logit and linear regression models to evaluate their (i) overall significance/fit and (ii) significance of predictor variables. These tests included a multi-collinearity test, goodness-of-fit test and a normality test. SPSS software was used to perform both diagnostic tests and the regressions. Tables, figures and narratives were used to present the data.

77

Table 3.5: Inferential Data Analysis Techniques Hypothesis

Hypothesis Test

Statistical Model

Expected Outcome

logit P( y  1)  B0  B1X1  B2 X 2  B3 X 3  

Where: Hypothesis 1: H0:Perceived attributes does not have a significant positive influence on usage of online retailing services

X 0: β = 0 X 0: β ≠ 0 Reject H0 if p < 0.05, otherwise fail to reject the H 0

Y = Usage of online retailing services P (y = 1)= Probability of belonging to either 1 or 0

β1- 3 = Logistic regression coefficients X1 = Perceived Attributes X2 = Perceived Risk X3 = Perceived Value ε1= Error Term

If the regressed B1 coefficient for the B1X1 product term is statistically significant (i.e. p < 0.05), this is interpreted as evidence of X1 having a significant effect on Y.

logit P( y  1)  B0  B1X1  B2 X 2  B3 X 3  

Where:

Hypothesis 2: H0:Perceived risk does not have a significant negative influence on usage of online retailing services

X 0: β = 0 X 0: β ≠ 0 Reject H0 if p < 0.05, otherwise fail to reject the H 0

78

Y = Usage of online retailing services P (y = 1)= Probability of belonging to either 1 or 0

β1- 3 = Logistic regression coefficients X1 = Perceived Attributes X2 = Perceived Risk X3 = Perceived Value ε1= Error Term

If the regressed B2 coefficient for the B2X2 product term is statistically significant (i.e. p < 0.05), this is interpreted as evidence of X2 having a significant effect on Y.

logit P( y  1)  B0  B1X1  B2 X 2  B3 X 3  

Where: Hypothesis 3: H0:Perceived value does not have a significant positive effect on usage of online retailing services

X 0: β = 0 X 0: β ≠ 0 Reject H0 if p < 0.05, otherwise fail to reject the H 0

Y = Usage of online retailing services P (y = 1)= Probability of belonging to either 1 or 0

β1- 3 = Logistic regression coefficients X1 = Perceived Attributes X2 = Perceived Risk X3 = Perceived Value ε1= Error Term

If the regressed B3 coefficient for the B3X3 product term is statistically significant (i.e. p < 0.05), this is interpreted as evidence of X3 having a significant effect on Y.

M = β0 + β1P1 + ε1 Hypothesis 4: H0:Customer perception does not have a significant influence on customer satisfaction with online retailing services

Where: X 0: β = 0 X 0: β ≠ 0 Reject H0 if p < 0.05, otherwise fail to reject the H 0

79

Y= M = Customer Satisfaction

β0 = Constant β1 = Linear regression coefficient P1 = Customer Perceptions (Composite Value) ε1= Error Term

If the regressed B1 coefficient for the B1P1 product term is statistically significant (i.e. p < 0.05), this is interpreted as evidence of P1 having a significant effect on Y (M).

Logit [ P (y = 1) ] = β0 + β1M + ε1 Hypothesis 5: H0:Customer satisfaction does not have a significant influence on usage of online retailing services

X 0: β = 0 X 0: β ≠ 0 Reject H0 if p < 0.05, otherwise fail to reject the H 0

Where: Y= Usage of online retailing services P (y = 1)= Probability of belonging to either 1 or 0

If the B1 coefficient for the B1M product term in the regression is statistically significant (i.e. p < 0.05), this is interpreted as evidence of M having a significant effect on Y.

β1 = Logistic regression coefficient M = Mediating Variable (Customer Satisfaction) ε1= Error Term

logit P( y  1)  B0  B1P  B2 Z  B3PZ  

Hypothesis 6: H0:Demographic factors do not have a significant influence on the relationship between customer perceptions and usage of online retailing services

Where: X 0: β = 0 X 0: β ≠ 0 Reject H0 if p < 0.05, otherwise fail to reject the H 0

Y = Usage of online retailing services P (y = 1)= Probability of belonging to either 1 or 0

β1-3 = Logistic regression coefficients P1 = Customer Perceptions (Composite Value) Z = Moderating Variable (Customer Demographics) PZ = Interaction Effect

ε1= Error Term

80

If the regressed B3 coefficient for the PZ product term is statistically significant (i.e. p < 0.05), this is interpreted as a significant interaction between P 1 & Z as predictors of Y. .

3.11.2 Qualitative Data Analysis To correspond to the sequential approach and to complement the quantitative findings, directed content analysis technique was used to manually analyze qualitative data from twelve key informant interview transcripts. Directed content analysis allows for the use of key concepts/variables identified from prior quantitative research as initial coding categories in a qualitative study (Potter & Levine-Donnerstein, 1999). These coding categories are then operationalized using definitions that are based on the theoretical review (Hsieh & Shannon, 2005).

Accordingly, pre-determined codes were adopted from seven previous ones used for quantitative data analysis (usage, perceived attributes, perceived risk, perceived value, customer perceptions, customer satisfaction and demographic factors). In addition, another three new codes (industry prospects, challenges/problems and policy recommendations) that were developed by the researcher based on the literature review formed part of total code categories. Through the inductive coding method, these initial/preliminary codes were used to group/cluster raw data from the transcripts for further analysis so as to make meaning and draw insights for further analysis (Ary, Jacobs & Sorensen, 2010).

As put forward by Hsieh and Shannon (2005), data that did not fall into these predetermined codes was isolated and later analyzed to ascertain if they represent a new category or a subcategory of an existing code. Appendix 5B illustrates the codebook for the

qualitative

data

analysis,

description/operationalization.

81

including

the

codes

and

their

respective

Once coding of the transcripts was completed and all items with a particular code placed together, the sets of items were reviewed to ensure that they belonged together since some may fall in more than one category. At this stage, the researcher also started considering how the codes could be merged into broader categories (Ary et al., 2010). Thematic coding methods were considered as they are often used for reducing text data into convenient summary categories or themes used to draw conclusions about a sample (Krippendorff, 1980; Weber, 1990; Jackson & Trochim, 2002). These methods are mostly used with denser types of text, such as in-depth interview transcripts where richer context can result in the identification of repeated themes (Jackson & Trochim, 2002).

In line with the constant comparative method (Ary et al., 2010), the clustered data underwent thematic analysis whereby similar data was further labeled and analyzed so as to create meaning and thereafter merged into broader categories or emerging themes. Here, the researcher relied on a combination of qualitative analysis methods to determine and classify the themes as suggested by Ryan & Bernard (2003). They entail the use of observational techniques such as identifying and sorting of words and key phrases, and checking for repetition, similarity and differences in expressions or statements

After sorting the data into broad categories, the researcher carried out further analysis of the various categories to determine whether some could be merged into themes (Ary et al., 2010). Consequently, additional thematic analysis of the transcripts was carried out manually in order to establish the key themes. In this study, the interview information was sorted into five broad themes: (i) usage diversity, (ii) prevailing attitudes, (iii) usage drivers, (iv) market development and (v) market prospects.

82

Appendix 5C shows the codebook for the thematic analysis, as well as the definitions given to each concept and significant statements attributed to respondents regarding each thematic area.

3.12

Ethical Issues

The main ethical issues in this study revolved around confidentiality, honesty among respondents/participants and data collection in general. As far as possible, the researcher addressed these key ethical concerns.

To begin with, confidentiality mainly concerned the identity of online retailing services/companies (as well as that of their customers who served as respondents). All six online retailing firms granted the researcher access to their customers, but preferred not to be identified by name given the highly competitive and developing nature of the online retailing industry, compelling the researcher to describe the firms in a general manner. Moreover, since the respondents did not want their identities disclosed, care was taken to guarantee anonymity of the research participants.

The second confidentiality issue is ensuring the anonymity of the interviewee in relation to the disclosed information: Some of the information shared by the interviewee during the interview could jeopardize his or her position. There‘s therefore need to protect respondents by non-disclosure of their identity and from those whose interests conflict with those of the interviewee (DiCicco-Bloom & Crabtree, 2006).

Also, honesty among the research participants was necessary for the success of this study. For this reason, the researcher insisted on honesty on the part of all of the respondents. This was not to be taken for granted since not all participants may have

83

been aware of what constitutes ethical behavior. According to Zickmund & Babin, (2010), honest cooperation is the main obligation of the research participant.

In terms of data collection, the researcher sought permission from the purposively selected online retailing companies as well as notified all potential respondents beforehand regarding the nature and objective of the study. This was also aimed at encouraging participation.

84

CHAPTER FOUR RESEARCH FINDINGS AND DISCUSSION

4.1 Introduction This chapter comprises of data collection details as captured using the research questionnaire, key informant interview guide and documentary sources of secondary data as well as the analysis of those findings. Questionnaire feedback was analyzed using both descriptive and inferential statistics and has been summarized and presented in the form of tables, charts and narratives. On the other hand, interview feedback and documentary evidence were analyzed using content analysis and presented in narrative form.

4.2

Descriptive Data Analysis

Descriptive statistics provide a summary of the characteristics of response data (Wilson, 2006). The descriptive statistics that were used for analysis of the key variables in this study include frequency of the distribution (in terms of counts and percentages) as well as measures of central tendencies (mean) and dispersion (standard deviation).

4.2.1

Response rates

From three hundred and ninety one (391) respondents who are registered as users of 6 online retailing services in Nairobi County, Kenya, two hundred and eighty-seven (287) were able to participate in the study by completing and returning the questionnaire. However, a number of these questionnaires (34) were poorly/improperly filled, while another 13 arrived too late, necessitating their exclusion from the study. Ultimately, the final respondents amounted to 240, equivalent to a 61.38% response rate. This is depicted in Table 4.1. 85

Table 4.1: Distribution of responses

Response

Stratified Sample Strata 1. Firm 1

Proportionate Sample

105

2. Firm 2 21 3. Firm 3

139

4. Firm 4

41

5. Firm 5

31

6. Firm 6

54

Total

391

Category

Frequency

Frequency

Percentage

Active Users Inactive Users Active Users

22

7.69

83

23 39

18.97

3

6

1.02

Inactive Users Active Users Inactive Users Active Users Inactive Users Active Users Inactive Users Active Users Inactive Users Total

18 45 94 6 35 3 28 8 46 391

8 40 46 8 15 1 15 12 27 240

2.56 16.42 21.02 1.02 7.18 0.52 7.69 2.05 13.85 100

Source: Survey Data (2014)

While there is still no consensus on what percentage of response rate should be acceptable for reporting and analysis in research, the rule of thumb is the higher the better (Rojas-Méndez & Davies, n.d.). Accordingly, Babbie (1990) argued that ―a response rate of at least 50 percent is generally considered adequate for analysis and reporting, a response rate of at least 60 percent is considered good, and a response rate of 70 percent or more is very good‖. More recently, Rubin and Babbie (2011) suggested that a 50 percent response rate is considered adequate for reporting and analysis. This means that the response data was more than adequate for analysis and reporting.

4.2.2

Demographic Characteristics of Online Retail Users

Table 4.2 shows a summary of the demographic characteristics of the respondents based on Section A of the questionnaire.

86

Table 4.2: Demographic characteristics of the sample (n = 240) Variable

Category

Frequency

Percentage

Age

18-23 Years 24-29 Years 30-35 Years 36-41 Years 42-47 Years 48 years and above Total High School Cert. Diploma Bachelor‘s Degree Masters‘ Degree Doctorate Professional Other Total Less than KSh 24,999 KSh25,000 – 49,999 KSh50,000 – 74,999 KSh75,000 – 99,999 KSh100,000 – 124,999 KSh125,000 & above Total

30 105 77 24 4 0 240 1 37 139 51 8 3 1 240 31 45 54 43 32 35 240

12.5 43.8 32.1 10.0 1.7 0 100.0 0.4 15.4 57.9 21.3 3.3 1.3 0.4 100 12.9 18.8 22.5 17.9 13.3 14.6 100.0

Level of Education

Monthly Income

Source: Survey data (2014) In terms of the respondents‘ (n=240) ages, the majority (43.8 %) were between 24 – 29 years while the minority (1.7 %) were between 42 – 47 years of age. None were older than 48 years of age. When it comes to education, a majority of the respondents (57.9 %) have a Bachelor‘s degree, followed by 51 (21.3 %) who have a Master‘s degree and 37 (15.4 %) who have a diploma. Only 1 (0.4 %) had a high school certificate, while 3 (1.3 %) had a professional qualification. With regards to the monthly income of the respondents, the majority (22.5%) earned between KSh 50,000 – 74,999, whereas the minority (12.9 %) had a monthly income of less than KSh 24,999 per month.

87

Taken as a whole, the demographic information showed that the respondents are predominantly young, relatively well educated and with relatively high levels of income. These findings concur with past studies regarding online population (Bellman et al., 1999) and e-shoppers in particular, which established that the online shoppers are generally younger, with high level of income and more educated (Li et al., 1999; Vrechopoulos et al., 2001; Dholakia & Uusitalo, 2002).

4.2.3

Customer Perceptions of Online Retail Users

The key customer perception variables of interest to this study were perceived attributes (PATT), perceived risk (PRSK) and perceived value (PVAL) as well as the composite variable customer perceptions (CPER). Table 4.3 provides a summary of the descriptive statistics for the three predictor variables as well as that of the composite variable.

Table 4.3: Descriptive Statistics Results for Customers’ Perceptions Variable PATT PRSK PVAL CPER

Measure Mean Std. Deviation Mean Std. Deviation Mean Std. Deviation Mean Std. Deviation

Statistic 4.3899 1.54140 4.1256 1.27855 4.2589 1.16118 4.2585 .62906

Source: Survey data (2014) The results indicate that perceived attribute had the highest mean score (4.389), closely followed by perceived value (4.258) and perceived risk (4.125). This high value concurs with several studies have empirically established that the perceived attributes is a key element in usage behavior of online services (Parthasarathy & Bhattacherjee, 1998; Bhattacherjee, 2001b). The mean score of the customer perception variables was also reasonably high (4.26). 88

On the other hand, the standard deviation for perceived attributes (1.54) is the largest, followed by perceived risk (1.28) and perceived value (1.16). According to Rumsey (2011) a large standard deviation isn‘t necessarily a bad thing; it just reflects a large amount of variation in the data set that is being studied. This indicates that the responses for perceived attributes are further away from the mean than those of perceived value.

Descriptive statistics (mean and sum) of to the 10 individual perceptual indicators we also carried out in order to establish the overall emphasis placed by respondents on each. This is presented in Table 4.4. Table 4.4: Descriptive Statistics Results for Individual Perceptual Indicators No. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Variable Usefulness (USFL) Compatibility (CMPT) Ease-of-Use (EOUS) Financial Risk (FINR) Performance Risk (PRFR) Personal Risk (PRSR) Monetary Value (MOVL) Convenience Value (COVL) Social Value (SOVL) Emotional Value (EMVL)

Mean 4.25 4.19 4.17 3.87 4.09 4.86 4.29 5.11 2.89 4.02

Sum 825.25 813.00 809.17 754.00 793.60 943.67 832.25 992.00 560.60 780.29

Rank 4 5 6 9 7 2 3 1 10 8

Source: Survey data (2014) The results reveal that convenience value (COVL) had the highest scores (sum = 992.00; mean = 5.11) for an individual indicator, implying that consumers placed the utmost importance on the convenience of using online retailing services. This is corroborated in a study by Robinson, Riley, Rettie and Rolls-Wilson (2007) which established that the key motivation for online shopping is the convenience of round-theclock shopping and having the items delivered at one‘s door step. On the opposite end, social value had the lowest scores (sum = 560.60; mean = 2.89).

89

4.2.4

Customer Satisfaction of Online Retail Users

In addition, the study assessed the customer satisfaction of users of online retailing services. Table 4.5 provides a summary of the results of the descriptive statistics for the customer satisfaction variable.

Table 4.5: Descriptive Statistics Results for Customer Satisfaction Variable

Measure

Statistic

N

240

Mean

3.1650

Std. Deviation

1.0866

CSAT

Source: Survey data (2014)

The results in Table 4.5 indicate that customer satisfaction had a mean score of 3.165. This moderate value could imply that users are not very enthusiastic about the online retailing services. This is a cause for concern since customers‘ overall satisfaction is an indication of how well customers like their experience with using the website, and it is probably the best indication of their willingness to return to the site again if they are to make another purchase in the category (Jiang & Rosenbloom, 2005). With regards to the standard deviation, the value was 1.087.

4.2.5

Usage of Online Retailing Services

The study also assessed the usage of online retailing services. Table 4.6 provides a summary of the results of the descriptive statistics for the usage variable.

Table 4.6: Descriptive Statistics Results for Usage of Online Retailing Services Variable USAGE

Category

Frequency

Percentage

Active User

129

53.8

Inactive User

111

46.3

Total

240

100.00

Source: Survey data (2014)

90

The study results in Table 4.6 reveal that the number of active users of online retailing services who responded was 129 (53.8 %), which was moderately higher than the number of respondents making up inactive users, who amounted to 111 respondents (46.3 %). This points to a fairly representative number/level of respondents who participated in the study, making it generalizable as argued in the problem statement.

4.3

Regression Analysis and Test of Hypotheses

The previous section dealt with descriptive statistics regarding the customer perceptions affecting usage of online retailing services in Nairobi County, Kenya. However, in order to empirically test the study‘s premise - that there is a relationship between customers‘ perceptions and the usage of online retailing services but this relationship is mediated by customer satisfaction and moderated by demographic factors - inferential statistics using logistic and linear regression methods was conducted as appropriate at 95 percent confidence level ( = 0.05) on the response data.

Both linear and logistic regressions analyze the relationship between independent variable(s) and an independent variable. On its part, linear regression analyzes linear relationships, which require a continuous numerical dependent variable (such as customer satisfaction in this study) that follows a normal distribution. In contrast, logistic regressions require binary dependent variable (i.e. usage in this study), which are coded 0/1 and indicate if a condition is or is not present, or if an event did or did not occur (Sweet & Grace-Martin, 2010).

Therefore, the first technique - logistic regression analysis – was used in the first model (equations 1, 2 and 3 in Chapter 3) to empirically analyze the response data for purposes of establishing the main effects of the predictor variables on the criterion variable as a way of estimating the conditional probability of being either an active or

91

inactive user. In addition, it was used in the third model (equation 6) to establish the moderating effect of variable Z on the IV-DV relationship. The second regression method - binary linear regression - was carried out on the second model (equations 4) so as to establish the effect of customer perceptions on customer satisfaction (M).

4.3.1

Diagnostic Tests

However, before subjecting the data to regression analysis, diagnostic tests were first carried out on the collected data to establish if it conformed to the requisite assumptions. The first diagnostic test was the multicollinearity test, which was done so as to establish if the three IV‘s (perceived attributes, perceived risk and perceived value) are inter-related or not. For this study, the collinearity tests were conducted using correlation analysis, tolerance and variance inflation factor (VIF) analysis. Table A.12 shows the correlation matrix for the three predictor variables. According to DeFusco (2007), multi-collinearity arises when two or more independent variables are highly correlated with each other. It is important to note that there is no consensus in extant literature on the acceptable correlation value/level between two variables, but Cooper and Schindler (2008) recommend a correlation value of 0.8 or greater to denote multicollinearity between two IVs (The IV - constant value is ignored, since collinearity among the predictors is what is under investigation). As is evident in the correlation matrix (Table A.12), there is no correlation value of IVs that is greater than 0.8; the highest correlation value of IVs is 0.519. We can therefore conclude that the correlation between the predictor variables in the model was not significant to warrant dropping any of them.

92

Besides bivariate correlation, the study also employed tolerance and variance inflation factor (VIF) analysis to determine the multicollinearity. This is illustrated in Table 4.7 and Table A.13 in the appendix. Table 4.7: Results of Collinearity Statistics Variable

Tolerance

VIF

Perceived attributes (PATT)

0.261

3.826

Perceived risk (PRSK)

0.468

2.135

Perceived value (PVAL)

0.348

2.877

Source: Survey data (2014) Results of the study reveal no multicollinearity problem for the three IVs: perceived attributes, perceived risk, and perceived value. This is due to the fact that the tolerance values for the three variables are greater than 0.1, while the VIF values are all lesser than 10, which show that there is no collinearity amongst the three predictors (Field, 2005). Consequently, all three variables were retained in current research model and used in the regression analysis.

Another diagnostic test that was carried out is the goodness-of-fit test. In logistic regression, goodness-of-fit tests for proposed models are commonly used to describe how well a proposed model fits a set of observations (Wu, 2010). There are different goodness-of-fit tests all of which have pros and cons. In this study, the HosmerLemeshow (H&L) test was used. The H&L test for goodness-of-fit is a statistical measure that shows how good a model fits the data. According to the test, if the significance measure of the model is less than 0.05, the model doesn‘t fit the data very well. It is significant if it is larger than 0.05. Table 4.8 presents the significance value of the H&L Test (1.0) for the logistic regression of the main effects model, which is greater than the required 0.05, meaning that the model does fit the data very well.

93

For the linear regression (equation 4), the overall fit of the model was assessed using the F-test, in line with Doane and Seward (2009). Since the p-value ˂  = 0.05, the null hypothesis that the proposed linear regression model doesn‘t fit the data very well was rejected, implying that the model is statistically significant and therefore useful in predicting customer satisfaction.

Also, a diagnostic test to ascertain the normality of the customer perceptions (CPER) distribution was undertaken. Normality is an important assumption of many statistical procedures such as t-tests, linear regression analysis, discriminant analysis and ANOVA (Razali & Wa, 2011). This study used the formal normality test, specifically the 1-sample Kolmogorov-Smirnov (KS), to test for evidence of the normality of the customer perceptions distribution, in line with Razali and Wa (2011). The outcome of the test as is shown in Table 4.8 as well as Table A.15 in the appendix. According to the KS test, if the significant value is less than 0.05, there is a significant difference between the population and sample, implying that the data is not normally distributed. Table 4.8: Results of Kolmogorov-Smirnov Normality Test Predictor Variable

Significant Value

Perceived Attributes (PATT)

.000

Perceived Risk (PRSK)

.000

Perceived Value (PVAL)

.000

Customer Perceptions (CPER)

.200

Source: Survey data (2014)

The study results in Table 4.8 show that while the KS test significant-values for the three predictors was 0.000, the overall KS test p-value (.200) for the composite customer perceptions distribution was greater than the significant level (p = 0.05), thus implying that the customer perception variable data is normally distributed.

94

This

means that the customer perception distribution satisfies the assumptions of equal variance and normality.

Since the assumptions of the two regression models are reasonably satisfied, the researcher proceeded to perform inference for the regression models.

4.3.2

Test of Hypotheses

After the successful conducting the preliminary diagnostic tests and confirming that the data complied with the requisite assumptions, regression analysis was performed on the data to test the hypotheses. The relevant hypotheses tests are presented in the sections 4.2.3.1. – 4.2.3.6 and the results are summarized in Tables 4.9 and 4.10..

Table 4.9: Results of Logit Regression Analysis β

t = β/S.E

Wald

P-Value

Perceived Attributes (PATT)

6.006

2.903

8.429

0.004

Perceived Risk (PRSK)

-1.834

-2.148

4.618

0.032

Perceived Value (PVAL)

2.329

2.149

4.621

0.032

Customer Satisfaction (CSAT) Interaction Term (CPERDEMF)

5.171 -0.196

6.488 -0.374

42.056 0.141

0.000 0.708

Variable

240

Observations (n) Nagelkerke R Squared (Main Effects) Classification Rate(Main Effects) Hosmer and Lemeshow (Main Effects) Nagelkerke R Squared (Equation 5) Classification Rate(Equation 5) Nagelkerke R Squared (Equation 6) Classification Rate(Equation 6) Dependent variable is Usage (USAGE)

Note * p≤ 0.05 Source: Survey data (2014)

95

(8 df)

0.974 98.8% 0.097 0.874 95 % 0.616 81.3%

1.0

To begin with, logistic regression analysis was used in the first model (see equations 1, 2 and 3 in chapter 3) for purposes of establishing the direct effects of the predictor variables on the criterion variable as a way of estimating the conditional probability of someone being either an active or inactive user. In addition, it was used in the third model (equation 6) to establish the moderating effect of variable Z on the IV-DV relationship. The results are summarized in Table 4.9

For the logistic regression model summary, the coefficient of determination (R2) was estimated using the Nagelkerke‘s R2, a supplementary goodness-of-fit measure recommended by Pallant (2007). Table 4.8 shows that it was 0.974 for the main effects and 0.874 for equation 5, indicating a very strong relationship between the IVs and the DV. This means that about 97.4% and 87.4% of the variation in the outcome variable is explained by the independent variable.

Additionally, the Wald statistic, was used to determine the ―significance‖ of the contribution of each variable in the model, in line with Chan (2004), whereby, the higher the value, the more ―important‖ it is. The relevant hypotheses tests that were conducted to assess the significance of the Wald statistic tested the null hypothesis at 95% confidence level wherein the acceptability level of the hypothesis test was  = 0.05, as recommended by Burns and Burns (2009). The relevant hypotheses tests are presented in the sections.

The second regression method - simple linear regression - was used out on the second model (equation 4) so as to establish the effect of customer perceptions on customer satisfaction (M). The relevant results are summarized in Table 4.10.

96

Table 4.10: Results of Simple Linear Regression Analysis Variable Customer Perceptions (CPER) Observations (n)

β

S.E.

t = β/S.E

P-Value

1.258

0.077

16.337

0.000

0.531 0.529 269.024

0.000

240

R-Squared (Equation 4) Adjusted R-Squared (Equation 4) F-Stat (Equation 4)

(1 df)

Dependent variable is Customer Satisfaction (CSAT)

Note * p≤ 0.05 Source: Survey data (2014) For the simple linear regression model summary (equation 4), the adjusted R-Squared (R2) value was 0.531, meaning that about 53% of the variation in the DV can be explained by the model. The adjusted R2 is important as it helps to discourage overfitting of the model (Doane & Seward, 2011).

4.3.2.1 Hypothesis 1: Relationship between Perceived Attributes and Usage of Online Retailing Services For the main effects model, the first hypothesis to be tested regarded the relationship between perceived attributes and usage of online retailing services.

H01: There‘s no relationship between perceived attributes and usage of online retailing services in Nairobi County, Kenya. As revealed in Table 4.9, the null hypothesis which proposes that there‘s no relationship between perceived attributes and usage of online retailing services was rejected since β ≠ 0 and p-value = 0.004. This is consistent with past research by Adams, Nelson and Todd (1992), which empirically established perceived attributes such as usefulness and ease-of-use are important determinants of system use and Parthasarathy and Bhattacherjee (1998) which established that the perceived attributes of an online service such as usefulness and compatibility determine usage behavior. 97

Similarly, Bhattacherjee‘s (2001b) empirical study of the antecedents of e-commerce service continued usage demonstrated that perceived usefulness is a key determinant of customer‘s continued usage intention (CUI). This can be interpreted that usage depends on cognitive beliefs (i.e. perceptions) about attributes of online retailing services.

4.3.2.2 Hypothesis 2: Relationship between Perceived Risk and Usage of Online Retailing Services For the second predictor variable of the main effects model, perceived risk, the hypotheses that was tested at 95% confidence level acceptability level  = 0.05 is as follows:

H02: There‘s no relationship between perceived risk and usage of online retailing services in Nairobi County, Kenya.

The research findings depicted in Table 4.9 show that for perceived risk (PRSK), β = 1.834 and p-value = 0.032. Hence, the null hypothesis for H2 is rejected since β ≠ 0 and p-value <  = 0.05. However, the study‘s findings show that perceived risk has a significant negative effect on usage. The result concurs with the findings of previous studies (Jarvenpaa & Tractinsky, 1999; Bhatnagar et al., 2000; Lee et al., 2000; Forsythe, Chuanlan, Shannon & Gardner, 2006; Barnes., Bauer, Neumann & Huber, 2007) that perceived risk is negatively associated with online shopping. It also parallels a more recent study by Liu and Forsythe (2010) who argued that risk is often a barrier to online transactions. This simply means that the greater the perceived risk, the less likely consumer are to use online retailing services in the future.

98

4.3.2.3 Hypothesis 3: Relationship between Perceived Value and Usage of Online Retailing Services The relationship between perceived value and usage was tested using the following hypothesis:

H03: There‘s no relationship between perceived value and usage of online retailing services in Nairobi County, Kenya.

The study results reveal in Table 4.9 that for perceived value (PVAL), β = 2.329 and pvalue = 0.032. Therefore, the null hypothesis was rejected since β ≠ 0 and p-value < . This means that perceived value has a statistically significant effect on the usage of online retailing services. The findings of the study are consistent with the previous research which established that perceived customer value is a significant determinant of online transaction behavior (Chew, Shingi & Ahmad, 2006). As pointed out by Abadi, Hafshejani and Zadeh (2011), users will perceive online shopping to be valuable when they see colleagues, friends and family members use it and get a recommendation of using it from them. Accordingly, the researcher did not drop any of the three IVs from the model since their effect was significant based on the Wald statistic.

4.3.2.4 Hypothesis 4: Relationship between Customers’ Perceptions and Customer Satisfaction with Online Retailing Services The fourth hypothesis to be tested regarded the relationship between customers‘ perceptions and customer satisfaction with online retailing services.

H04: There‘s no relationship between customers‘ perceptions and customer satisfaction with online retailing services in Nairobi County, Kenya.

The study results in Table 4.10 reveal that for the cumulative customer perceptions (CPER), β = 1.258 and p-value = 0.000. Consequently, the null hypothesis was rejected 99

since β ≠ 0 and p-value < , meaning that customer perceptions have a statistically significant effect on customer satisfaction with online retailing services. This outcome lends support to the findings of Bolton and Drew (1994) that empirically established that customer perceptions have a significant positive relationship with customer satisfaction in the service context. More importantly however, it corroborates studies (Westbrook and Oliver, 1981) that advanced the notion that satisfaction may comprise of both emotional (i.e. perceived value) and cognitive (i.e. perceived attributes, perceived risk) determinants. This is in line with Sing (1991) who argued that customer satisfaction can be understood as a collection of multiple satisfactions with various objects that constitute the service system.

4.3.2.5 Hypothesis 5: Relationship between Customer Satisfaction and Usage of Online Retailing Services The fifth hypothesis sought to assess the relationship between customer satisfaction and usage. H05: There‘s no relationship between customer satisfaction and usage of online retailing services in Nairobi County, Kenya.

According to Table 4.9, for customer satisfaction (CSAT), β = 5.171 and p = 0.000. Since p = 0.000 < α = 0.05, the null hypothesis was rejected, meaning that customer satisfaction has a statistically significant effect on usage. This outcome is in line with an earlier studies by Baroudi, Olson and Ives (1986), who argued that user satisfaction could lead to system usage, and Bhattacherjee (2001a), according to whom, satisfied users tend to continue using an IS, while dissatisfied users tend to discontinue IS usage and/or switch to an alternative. It also corroborates a study by Ortiz de Guinea and Marcus (2009) which established that satisfaction drives IT usage directly and relationship marketing literature which relates satisfied consumers with repeated

100

purchases/usage and loyalty (Martin & Camarero, 2009). Table A.28 presents the output for the logistic regression for customer satisfaction and usage.

4.3.2.6 Hypothesis 6: Moderating Effect of Demographic Factors on the Relationship between Customer Perceptions and Usage of Online Retailing Services The sixth and final hypothesis sought to test the moderating effects of demographic factors and is as follows:

H06: Demographic factors do not have a moderating effect on the relationship between customer perceptions and usage of online retailing services in Nairobi County, Kenya.

As revealed in Table 4.9, for the interaction variable (CPERDEMF): β = - 0.196 and p = 0.708, implying that the interaction term is not statistically significant. We thus fail to reject the null hypothesis which proposes that demographic factors have no significant moderating effect on the usage of online retailing services in Nairobi, Kenya. The insignificant moderating effect of demographic factors on perceptions is not surprising, since it is in line with several past technology adoption and usage studies (Szajna, 1996; Gefen & Straub, 1997; Gefen & Keil, 1998; Hernandez et al., 2011). The results also confirm the findings by Bellman et al. (1999), who argue that while demographics appear to influence initial use of the Internet, its effect seems to disappear once people are online on a regular basis. In other words, the behaviour of experienced users is not identical to that of an individual during their initial employment of the IT/IS in question (Gefen, Karahanna & Straub, 2003; Yu, Ha, Choi & Rho, 2005), since the experience acquired modifies the effect of the variables considered. This implies that that once an individual becomes familiar with the IT/IS

101

(in our case online shopping), the experience acquired may nullify the importance of their socio-economic characteristics (Sun & Zhang, 2006). In conclusion, this can be interpreted that demographic factors do not influence the usage behaviour of the experienced online shopper.

4.3.3

Summary of the Test of Hypotheses

The results of the hypothesized relationships are summarized in Table 4.11.

Table 4.11: Summary of Hypotheses Tests Causal Relationship

Hypothesis

Outcome of Hypothesis Test

H01: There‘s no relationship between perceived attributes and usage of Perceived attributes  Usage online retailing services

Rejected Ho

H02: There‘s no relationship between perceived risk and usage of online Perceived risk  Usage retailing services

Rejected Ho

H03: There‘s no relationship between Perceived value  Usage perceived value and usage of online retailing services

Rejected Ho

H04: There‘s no relationship between customers‘ perceptions and Customer perceptions  customer satisfaction with online Customer Satisfaction retailing services

Rejected Ho

H5: There‘s no relationship between customer satisfaction and usage of Customer satisfaction  Usage online retailing services

Rejected Ho

H6: Demographic factors do not have a moderating effect on the Demographic factors  relationship between customer (Customer perceptions  perceptions and usage of online Usage) retailing services

Source: Survey data (2014)

102

Failed to reject Ho

As presented in the table, all but one of causal relationships between the constructs postulated by our model is well supported. The study found support for H1-H5. However, the results failed to support H6, that demographic factors have a moderating effect on the relationship between customer perceptions and usage of online shopping services.

4.4 Content Analysis This section presents findings for the qualitative analysis of data collected during the key informants interviews aimed at providing an in-depth perspective of the effects of customer perceptions on online retailing usage in Kenya. The findings are summarized into five major themes and are presented in the form of narratives.

4.4.1

Interview Participants

From each of the six online retailing firms that agreed to participate in the study, the researcher identified two individuals who were deemed eligible to serve as key informants and who agreed to participate in the semi-structured interviews. In total, the researcher conducted twelve (12) in-depth semi-structured interviews with key decision-makers in each firm. These individuals were selected due to their influential roles/occupations within the firms. The distribution of interviewees is depicted in Table 4.12.

Table 4.12: Distribution of interview participants Role/Occupation

Number of Interviewees

1.

Online retailing entrepreneurs/owners

2

2.

Consultants working for the online retailing firms

2

3.

Managers of the online retailing firms

7

4.

Employees of the online retailing firms

1

Total

12

Source: Survey data (2014) 103

4.4.2

Key Themes

The key informant interview findings gave an in-depth perspective regarding the effect of customer perceptions on online retailing usage in Kenya. The findings largely complemented those found in the quantitative analysis, with no major contradictions. They are summarized in the following section based on the five major themes that were developed via directed content analysis as outlined in Chapter 3.

4.4.2.1 Theme No 1: Usage Diversity The findings revealed that usage of online retailing services is quite prevalent in Kenya, as all the respondents admitted that they have used local online retailing services before to buy and sell goods and services. It was interesting to note the multi-faceted/diverse usage behaviour of respondents: while some used the websites primarily for price comparison (researching), others went further and used them for purchasing items such as books, electronics (television sets and mobile phones) as well as tickets for events.

As one participant stated, ―I used Jumia to look at the price range of a laptop that would fit my budget as well as look at the specifications of that laptop‖. Another interviewee reported that ―… I was looking for products I wanted to purchase [but] I was also selling some goods on OLX‖. Another respondent reported used Mzoori.com, Rupu.com, Jumia and OLX ―to look for products for purchase and was also selling some goods on OLX‖. Interestingly, a majority disclosed that they have used online retailing services within the last three months.

However, one response that stood out was from someone who had not shopped online for more than three months.

As that participant explained, ―…I only use online

retailing stores when I am seeking to buy high cost items because I get an opportunity

104

to compare prices across different e-commerce stores. I do not do this frequently that‘s why I cannot do this (use the websites) every now and then…‖. When asked whether there are any goods they wouldn‘t purchase online, one respondent cited electronics, clothes and furniture (especially sofas). According to the respondent, ―I want to see and feel and test (them) before purchasing‖. Other mentioned food as another item that they wouldn‘t purchase online ―due to its perishability‖.

4.4.2.2 Theme No 2: Prevailing Attitudes There were a number of interesting responses which highlighted the mixed opinions, thoughts and feelings regarding online retailing in Kenya. For instance, when asked ―how satisfied (or dissatisfied) are you with local online retailing services in Kenya?‖, the respondents were roughly split in the middle as some replied in the affirmative while others were negative.

Those respondents who did not have a favorable attitude attributed their negative attitudes to a variety of factors including product availability, delays in delivery, issue with payment modes, mistakes during delivery and poor customer service. According to one owner-manager, ―the inconsistency in product availability coupled with the lack of accurate information on some websites has made online users form a negative impression towards online retailing services‖. This point was emphasized by another dissatisfied respondent who stated that ―quite a number of online retailers cannot guarantee proper inventory management – you order an item online but you cannot have it (delivered) because it is out of stock‖.

Moreover, one employee remarked that ―I‘m very dissatisfied. I have not been able to purchase anything because what‘s available is not within my budget range‖. Yet

105

another respondent who is an owner added that ―I‘m somewhat satisfied…however, it can be quite costly to roll out‖.

On the other hand, one of the owners who responded in the affirmative asserted that ―I am very satisfied…we have come a long way in terms of e-commerce in Kenya‖. Another respondent added by saying ―I‘m very satisfied with online retailing in Kenya because of how fast it is growing and the convenience it brings‖. Another like-minded respondent remarked that ―It‘s starting to be more and more appreciated as people become busy with fast/complicated lifestyles‖.

4.4.2.3 Theme No 3: Usage Drivers The study also identified several determinants of online retailing usage based on the interview responses. Overall, the interview findings indicated that convenience, product "assortment/variety, price (i.e. deals & discounts) and credibility are the most important drivers of online retailing usage. All respondents agreed that when these aspects are missing, it becomes difficult to enhance or even sustain usage of a service.

Further, the participants agreed that their convenience value is very high. As one stated, ―online retailing provides convenience as buyers get whatever they purchase without moving…‖ while another remarked that ―online shopping experience takes away the headache associated with traditional in-store shopping. They [online retailing websites] are very convenient for price comparison purposes‖. As one participant stated, ―I loved the convenience of getting the television set delivered to my doorstep without leaving the comfort of my house!‖.

106

Another important driver was usefulness. One interviewee agreed that online retailing services ―are very useful for price comparison‖ while another remarked that ―online retailing is quite useful as it helps one purchase products after viewing a variety without having to waste time walking around‖. Ease of use was another driver. ―The websites and mobile apps are generally very easy to use‖. According to the respondent, one reason for this is ―modern payment methods such as M-pesa which make it easy to transact though it comes with some level of risk which most Kenyans are aware of and have found ways to overcome‖.

Indeed, the growth in electronic payment services such as credit cards and mobile money offered reliable and flexible payment options which have also contributed the growth of adoption and usage of online retailing. Majority of respondents confirmed that M-pesa is the main mode used for payment when shopping online because it reduces financial risk which has made many shy off from shopping online. One of the consultants interviewed confirmed this by stating that ―Mpesa is easy to use unlike credit cards which can be abused after the purchase‖. As one senior employee in one of the online retailing firms put it: ―There is a chance of personal details getting in the wrong hands, especially for those using visa cards to make payments‖. The respondent also indicated the performance risk involved by saying that ―there are chances that delivery does not occur as promised e.g. delivery within 24 hours turns to delivery after 72 hours‖.

With regards to social value, one owner remarked that ―Online retailing provides a platform where people can be able to comment or review their experiences or products/services. That way shopping decisions are made based on that [social value]…‖.

107

4.4.2.4 Theme No 4: Market Development While the respondents attested to the accelerated pace of change in how consumers are shopping online, most acknowledged that the sector was growing at a slower pace than expected and that more needs to be done to develop the sector in Kenya. In this regard, they suggested several possible strategies that could be employed in order to increase uptake and sustain the usage of online retailing services in Kenya.

One respondent who works as an internal consultant for one of the online retailing firms had three key recommendations: ―Increase the product range and ensure that products are always in stock; delivery to homes would increase the convenience of this value chain; ensure that the regulatory frameworks for protecting customers privacy are enhanced and there is judicial recourse if these are compromised‖. One owner recommended that online retailers should ―build trust; avail a variety of products; offer good prices for products; avail all information on one page; offer return policies; offer cash on delivery and offer warranties‖. Another added that ―there‘s need for more regulation from government to reduce the risks associated with online shopping‖. Moreover, one consultant cited the need for ―data privacy laws to protect consumers‘ sensitive data‖.

This example also shows that companies cannot ignore the risks that users face when using their services as ignoring them carries significant reputational risk through wordof-mouth as a result of negative customer experience. The consensus was that stringent regulation of e-commerce was a necessary component for the sector‘s growth.

108

4.4.2.5 Theme No 5: Market Prospects Last but not least, the respondents gave an analysis of the economic/business potential of the online retailing industry in Kenya. Generally, Kenya is seen as a regional leader in

e-commerce

and

online

retailing in

particular,

the

current

challenges

notwithstanding. As it turns out, most respondents affirmed that the current online retailing activity in Kenya is a pretty strong indicator of its potential to drive online sales in the near future.

For the most part, the consensus was that the prospects for online retailing in Kenya are bright and going by recent developments and trends, it may soon become a viable sales channel. As one respondent remarked, ―it is growing at a very fast pace compared to the rest of the region‖. This was supported by another respondent who added that ―online shopping as a trend is picking up and has a bright future. This is because more internet users are turning to online shopping especially due to its convenience‖.

Moreover, another respondent noted that ―online retailing is gaining momentum; while we have already achieved good progress, a lot more needs to be done to ensure security while transacting, product variety and value added services‖. Last but not least was one consultant who added that ―online shopping is an industry that is growing in Kenya and people are beginning to embrace it. A few things need to be worked on and it will turn into a lucrative industry‖.

109

CHAPTER FIVE SUMMARY, CONCLUSION AND RECOMMENDATIONS

5.1 Introduction This chapter presents a summary and discussion of the main findings of the study with respect to each study objective. Also, conclusions based on the findings are made, followed by recommendations of the study as well as suggestions for further study which are proposed at the end of this chapter. Every attempt was made to represent the facts with completeness and clarity.

5.2

Summary

In the last decade, Kenya has undergone a transformation in its ICT sector, which outperformed every other sector during this time. This remarkable growth has been characterized by introduction of various e-commerce services such as online retailing into the market, which target Kenya‘s rapidly growing internet population. However, this huge increase in internet usage in Kenya during this period has not been matched by a corresponding usage of online retailing services. Reports indicate that the usage of online retailing services in Kenya is still very low, thereby posing an existential threat to the service providers due to the financial sustainability problem of maintaining the loss-making online services. Consequently, continued loss-making may eventually lead to closure of the online service, resulting in waste of effort to develop the service. Stakeholders are therefore keen to establish the reasons as to why customers use online retailing services, especially the continued usage or post-adoption use.

110

Accordingly, research in IT continuance has examined different factors and/or processes that motivate continued usage or discontinuance of IT products or services, following their initial acceptance. Studies show that continued usage of IS can be influenced by individual/psychological, system/technical and organizational factors. However, this study restricted itself to examining individual psychological factors, specifically the antecedent role of customer perceptions on usage of online retailing services, coupled with the mediating role of satisfaction on the perception-usage relationship.

Accordingly, the general objective of this online consumer behaviour study was to empirically determine the relationship between customer perceptions and the usage of online retailing services in Nairobi County, Kenya. It proposed a research framework which blended online consumer behavior and technology adoption constructs, their

variables, indicators and the presumed relationships among them as described in the empirical literature.

This mixed method study made use of a descriptive, specifically cross-sectional, survey design and explanatory, correlational design. Quantitative data was collected by use of a self-reporting questionnaire that was administered to selected respondents, who are individuals registered as users of 6 online retailing services in Nairobi County, Kenya. On the other hand, qualitative data was collected using a semi-structured interview guide and analysis of relevant records and documents.

Questionnaire responses were analyzed using descriptive and inferential statistics; descriptive statistics that were used to summarize the data include frequencies, means and standard deviations. Inferential statistics, which involved both linear and logistic

111

regression analysis, was employed to infer the relationship between the study variables. The model specification was tested using multi-collinearity test, goodness-of-fit test and normality test. Tables, charts and narratives were used to present the data.

Results of the quantitative analysis that was presented in the previous chapter reveal that five out of the six causal relationships between the constructs postulated by the research model are well supported by the study‘s findings. Those variables that had a

positive association with usage are perceived attributes, perceived value and customer satisfaction, while customer perception had a positive effect on customer satisfaction. On the other hand, perceived risk had a significant negative effect on usage. However, the study established that demographic factors do not have a significant moderating effect on the relationship between customer perceptions and usage of online retailing services.

Qualitative data from twelve semi-structured key informant interview transcripts was manually analyzed using the directed content analysis technique into initial coding categories. Subsequently, the researcher carried out further thematic analysis of the various categories which were then merged into themes. Its findings complemented those found in the quantitative analysis, with no major contradictions. Based on the content analysis, 5 key themes were identified as providing meaningful insight into the usage of online retailing services context in the Kenyan context. These are usage diversity, prevailing attitudes, usage drivers, market development and market prospects.

112

5.3

Conclusions

Several important conclusions can be drawn from the findings of this study. These are categorized into six main areas in line with the objectives of the study. These conclusions are detailed in the following section.

For one, the current study has shown that perceived attributes is a key factor motivating the usage of online retailing services in Kenya. This finding underscores the significance of perceived website functionality vis-a-vis online customer decisionmaking behavior and is consistent with earlier studies which empirically examined post-adoption usage in the online service context.

The results have also drawn attention to the role of perceived risk as a stumbling block to online retailing usage in Kenya. It is evident from the study that perceived risk plays a key role in determining continued usage of online retailing services, albeit a negative one. This is consistent with previous research which shows that customer risk perceptions are negatively associated with usage of online retailing stores as consumers are only willing to purchase product/service from an online vendor that is perceived as low risk. In short, the greater the perceived risk, the less likely consumer are to use online retailing services in the future. Therefore, reducing such risk is crucial to online vendors‘ success.

In addition, the study conclusively established that perceived value is positively associated with usage. This concurs with previous studies which established that perceived customer value is an important determinant of online transactions behavior. It is also evident that customer perceived value in online retailing is analogous to offline context. This means that online shoppers opt for and repatronize online retailers who offer superior customer value in the same way as in traditional stores.

113

Furthermore, the findings conclusively reveal that taken as a whole, customer perceptions also have a significant effect on customer satisfaction with online retailing services. This is consistent with past studies according to which people form attitudes (e.g. satisfaction) towards products/services based on four underlying reasons: utilitarian function (perceived attributes), value-expressive function (perceived value), ego-defensive function (risk perceptions), and knowledge function (awareness of/experience with the product/service).

Similarly, the findings reveal that there is a relationship between customer satisfaction and usage of online retailing services, supporting earlier studies which argued that user satisfaction has an influence on system usage. This implies that when a system‘s use fulfills user need, the resultant user satisfaction with the system is expected to result in its greater use. On the other hand, if system usage does not meet user expectations, satisfaction will decrease thus limiting further use. Such dissatisfied users may discontinue system usage altogether and seek other alternatives

On the contrary, demographic factors have no relationship between customer perceptions and usage of online retailing services. This suggests that the explanatory power of the study model is not in any way enhanced by taking into account the moderating effect of the user demographic factors, as has been established in previous post-adoption studies.

5.4

Policy Implications

The empirical findings of this study have implications for practitioners as well as policy makers who want to enhance the likelihood of success of new online retailing services. These recommendations are outlined in the following section.

114

5.4.1

Implications for Practitioners

In view of the significance of perceived attributes, online retailers should enhance their service features/attributes as a way of ensuring success of their services. Therefore, the design of the online retailing services should take into consideration customer-specific needs by personalizing the website to make it more useful, compatible with customer requirements and easy to use for users with various levels of computer skills. Further, providers should differentiate themselves by offering attractive features that are highly regarded by target customers in addition to developing a user-friendly website that provides needed menu options and functionalities. Moreover, e-retailers should come up with a strategy for raising awareness of the relative advantages of online retailing services vis-à-vis other alternatives (e.g. offline shopping) as people are more likely to use such services when they perceive that advantages outweigh disadvantages.

There‘s need for online retailing service providers to reduce the risk perception amongst their users regarding online purchasing. Ignoring it could lead to significant reputational damage as a result of negative customer experience. One effective strategy for mitigating risk perception is reassuring users of safety of their transactions using website elements that assure users of the security and privacy of their information such as mandatory verification of users coupled with offering various secure methods of payment. Another strategy is instituting strict data protection policies that guarantee the privacy of personal information. Also, having security seal icons, offering extensive user information and getting the endorsement of credible third parties will help in building trust in the services.

Without a doubt, online retailers should design and deliver a unique value proposition that has both functional as well as hedonistic appeals. Functional value can be in terms

115

of convenience which is typified by reduced response time, minimized customer effort, and quick completion of transactions, while monetary/price value can be achieved via price comparison and/or offering lower prices than offline channels. Hedonistic appeal can be achieved through appealing to the user‘s emotional value chiefly through the website design.

Further to this, online retailers should have an effective customer satisfaction strategy for purposes of customer retention since satisfied customers will continue using the system in the long run, while dissatisfied users may discontinue system usage altogether and seek other alternatives. Some of the ways that retailers can deal with this issue effectively include the establishment of customer relationship management sections to handle complaints and service recovery, having service agents online with whom they can chat real time when they have a question or require assistance, as well as providing comprehensive and easily accessible self-service information that is easyto-read and understand.

Indeed, it is imperative for online retailing firms to have a good understanding of their target customers‘ needs, wants and demands, since this will not only help in determining the appropriate customer acquisition strategies but also how to enhance the long-term usage of their services. One approach that online retailers can employ is by establishing and growing strategic partnerships with a range of brands that are highly sought after by their users. This will address the challenge of product variety and availability.

5.4.2

Implications for Government

Online retailing offers a novel way to connect Kenya to global markets while creating the much needed jobs for the economy. Owing to this, it behooves the Government to 116

embrace, nurture and facilitate e-commerce usage in Kenya in order to spur further growth in this area.

However, for online retailing to meet the high expectations of users, investors and other stakeholders, the government should urgently address the current bottlenecks hampering online shopping usage in Kenya. For instance, there‘s need for legislation that will address e-commerce users‘ security and privacy concerns with a view to ensuring that organizations respect and protect consumers‘ privacy rights while at the same time offering legal recourse for the victims of fraud and other crimes arising from online activities.

Also, as part of its online consumer protection framework, the government - in consultation with industry stakeholders - should expedite the development of online consumer protection guidelines for e-commerce users. If need be, a quasi-independent multi-sectorial entity could be tasked with overseeing such a program as is the case in other countries.

5.5 Contribution of the Study to Knowledge The results of this study contribute to new knowledge in several ways. Most importantly, this study makes up for the dearth of empirical research on online shopping behavior of consumers in Kenya, specifically with regards to the relationship between customers‘ perceptions and the usage of online retailing services.

Further, the study provides a suitable consumer decision making model that is useful in predicting why individuals continue or discontinue usage of online retailing services.

117

The results of this study have ascertained the usefulness of the non-linear logistic regression model in explaining usage as a dichotomous dependent variable.

Moreover, this study fills the apparent methodology gap in current online consumer behavior literature by employing a rigorous and methodologically sound mixed-method research design in which both qualitative and quantitative approaches are integrated to contribute to a rich and comprehensive study. It therefore provides future scholars with a useful framework of how to incorporate both quantitative and qualitative methods in their online consumer research projects.

Undoubtedly, the study also theoretically contributes to new knowledge with its conceptualization of how customer satisfaction mediates the relationship between customers‘ perceptions and usage and by subsequently empirically validating its mediating role on the relationship between customer perception and usage in online retailing services.

In practical terms, the study findings regarding to the dual role played by customer satisfaction could of help to practitioners in the online retailing field when designing their usage as well as customer retention strategies since online shoppers who are satisfied with online retailers are likely to recommend the online retailer to someone and consider the retailer to be their first choice for future transactions. In contrast, online shoppers who are dissatisfied are likely to switch to competitors in case they experienced a problem with an online retailer, or to complain to other customers.

118

5.6

Suggestions for Further Study

This study primarily sought to establish the relationship between individual perceptual factors and usage of online retailing services, ignoring other determinants such as organizational and environmental factors. Owing to this omission, future studies should incorporate these missing factors as a way of enhancing the validity of the current model. Another area for future study regards the design of the research model; the insignificance of the moderating variable in this study underscores the need to undertake further research to explore the efficacy of other variables in the research model as a way of improving its statistical significance. Perhaps future studies could incorporate additional moderating variables such as psychographic factors in analyzing the usage of online retailing services.

119

References Abadi, H.R.D., Hafshejani, S.N.A., & Zadeh, F.K. (2011). Considering factors that affect users‘ online purchase intentions with using structural equation modeling. Interdisciplinary Journal of Contemporary Research in Business, 3(8), 463– 471. Adams, D.A., Nelson, R.R., & Todd, P.A. (1992). Perceived usefulness, ease-of-use and usage of information technology: A replication. MIS Quarterly, 16(2), 227– 247. Ahuja, R. (2005). Research methods. New Delhi: Rawat Publications. Ahn, J., Lee, J.P., & Park, J. (2001). Risk-focused e-commerce model – A crosscountry study. Working Paper. Retrieved from http://www.misrc.umn.edu/workingpapers/fullPapers/2001/0130_060101.pdf Alba, J., Lynch, J., Weitz, B., & Janisqewski, C. (1997). Interactive home shopping: Consumer, retailer, and manufacturer incentives to participate in electronic marketplaces. Journal of Marketing, 61(3), 38-53. Al-Gahtani, S. (2001). The applicability of TAM outside North America: An empirical test in the United Kingdom. Information Resources Management Journal, 14(3), 37–46. Al-Ghaith, W., Sanzogni, L., & Sandhu, K. (2010). Factors influencing the adoption and usage of online services in Saudi Arabia. EJISDC, 40(1), 1-32. Anderson, L. (2000, March). Retail and wholesale industry. Hoover’s Online. Retrieved from www.hooversonline.com. Anderson, E.W., Fomell, C. & Mazvancheryl, S. (2004). Customer satisfaction and shareholder value. Journal of Marketing, 68, 172-185. Andrews, L., Kiel, G., Drennan, J., Boyle, M.V. & Weerawarden, J. (2007). Gendered perceptions of experiential value in using web-based retail channels. European Journal of Marketing, 41, 5-6, 640-58. Ary, D., Jacobs, L.C. & Sorensen, C.K. (2010). Introduction to research in education (8th Ed.). Belmont, CA: Wadsworth, Cengage Learning. Australian Government Productivity Commission (2011). Economic structure and performance of the Australian retail industry. Productivity Commission Inquiry Report. (No. 56). Retrieved from: http://www.pc.gov.au. Babbie, E. (1990). Survey research methods. Belmont, Ca: Wadsworth. Bajaj, A. & Nidumolu, S.R. (1998). A feedback model to understand information system usage. Information & Management, 33, 213-224.

120

Bansal, H., McDougall, G., Dikolli, S., & Sedatole, K. (2004). Relating e-satisfaction to behavioral outcomes. Journal of Services Marketing, 18 (4), 290-302.

Barnes, S.J., Bauer, H.H., Neumann, M.M., & Huber, F. (2007). Segmenting cyberspace: A customer typology for the internet. European Journal of Marketing, 41 (1/2), 71-93. Barnett, T., Kellermanns, F.W., Pearson, A.W., & Pearson, R.A. (2006-2007). Measuring information system usage: Replication and extensions. Journal of Computer Information Systems, 47(2), 76-85. Baron, R.M., & Kenny, D.A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical consideration. Journal of Personality and Social Psychology, 51, 1173–1182. Baroudi, J.J., Olson, M.H., & Ives, B. (1986). An empirical study of the impact of user involvement on system usage and information satisfaction. Communications of the ACM, 29(3), 232-8. Bauer, R.A. (1960). Consumer behavior as risk-taking. In R.S. Hancock (Ed.). Dynamic Marketing for a Changing World (pp. 389-98). Chicago, IL: American Marketing Association, Bellman, S., Lohse, G., & Johnson, E. (1999). Predictors of online buying behavior. Communications of the ACM, 42 (12), 32-38. Berelson, B. & Steiner, G.A. (1964), Human behavior: An inventory of scientific findings. New York: Harcourt Brace Jovanovich. Bettman, J. (1973). Perceived risk and its components - A model and empirical test. Journal of Marketing Research, 10, 184-90. Bettman, J., Luce, M., & Payne, J. (1998). Constructive consumer choice process. Journal of Consumer Research, 25(3), 187-217. Bhatnagar, A., Misra, S., & Rao, H. R. (2000). Online risk, convenience and internet shopping behavior. Communications of the ACM, 42(11), 98-105. Bhattacherjee, A. (2001a). Understanding information systems continuance: An expectation-confirmation model. MIS Quarterly, 25(3), 351-70. Bhattacherjee, A. (2001b). An empirical analysis of the antecedents of electronic commerce service continuance. Decision Support Systems, 32(2), 201–214. Bhattacherjee, A., & Barfar, A. (2011), Information technology continuance research: Current state and future directions. Asia Pacific Journal of Information Systems, 21 (2), 1–18.

121

Bhattacherjee, A., Perols, J., & Sanford, C. (2008). Information technology continuance: A theoretic extension and empirical test. Journal of Computer Information Systems, 17–26. Bitner, M.J., & Hubbert, A.R. (1994). Encounter satisfaction versus overall satisfaction versus service quality: The consumer‘s voice. In R.T. Rust & R.L. Oliver (Eds.). Service quality: New directions in theory and practice. Thousand Oaks, CA: Sage Publications. Bizcommunity (2011). Kalahari Kenya, Kalahari Nigeria close down. Bizcommunity. Accessed on August 2012, Available at www.bizcommunity.co.ke/Article/111/426/66185.html. Black, W. (1983). Discontinuance and diffusion: Examination of the post adoption decision process. Advances in Consumer Research, 10, 356-361. Bolton, R.N. & Drew, J.H. (1991). A multistage model of customers‘ assessments of service quality and value. Journal of Customer Research, 17(4), 375-84. Bolton, R.N., & Drew, J.H. (1994). Linking customer satisfaction to service operations and outcomes. In R.T. Rust & R.L. Oliver (Eds.). Service quality: New directions in theory and practice (pp. 173 – 200). Newbury Park, CA: Sage Publications. Bolton, R.N., & Lemon, K.N. (1999). A dynamic model of customers‘ usage of services: Usage as an antecedent of satisfaction. Journal of Marketing Research, 36(2), 171–186. Boscheck, R. (1998). New media economics are transforming consumer relations. Long Range Planning, 31(6), 873-878. Bridges, E., & Florsheim, R. (2008). Hedonic and utilitarian shopping goals: The online experience. Journal of Business Research, 61(4), 309-314. Brink, P.J., & Wood, M.J. (1998). Advanced design in nursing research (2nd Ed.). Thousand Oaks: SAGE Publications. Brown, I., & Jayakody, R. (2009). ―B2C e-commerce success: A test and validation of a revised conceptual model. The Electronic Journal Information Systems Evaluation, 12(2), 129 -148. Bryman, A. (2006). Integrating quantitative and qualitative research: How is it done? Qualitative Research, Vol. 6 (1), 97–113. Burns, R. P., & Burns, R. (2009). Business Research Methods and Statistics Using SPSS, London: Sage Publications. Burton-Jones, A., & Gallivan, M.J. (2007). Toward a deeper understanding of system usage in organizations: A multilevel perspective, MIS Quarterly, 31(4), 657679.

122

Burton-Jones, A., & Straub, D.W. (2006). Reconceptualizing system usage: An approach and empirical test, Information Systems Research, 17(3), 228-246.

Capraro, A., Broniarczyk, J. & Srivastava, R.K. (2003). Factors influencing the likelihood of customer defection: The role of consumer knowledge. Journal of the Academy of Marketing Science, 31(2), 164–175. Cardozo, R.N. (1965). An experimental study of customer effort, expectation and satisfaction. Journal of Marketing Research, 2, 244-249. Caruana, A. (2002). Service loyalty: The effects of service quality and the mediating role of customer satisfaction. European Journal of Marketing, 36(7), 811 – 828. Casalo, L., Flavian, C., & Guinaliu, M. (2007). The impact of participation in virtual brand communities on consumer trust and loyalty. Online Information Review, 31 (6), 775-792. CCK. (2011). CCK Statistics Report on the Communication Sector Q4 2010/2011. Retrieved from www.cck.go.ke CCK. (2012). CCK Statistics Report on the Communication Sector Q4 2011/2012. Retrieved from www.cck.go.ke. Chan, Y. H. (2004). Biostatistics 202: Logistics regression analysis, Singapore Medical Journal, 45 (4), 149-153. Chang, H.H., & Chen, S.W. (2009). Consumer perception of interface quality, security, and loyalty in electronic commerce, Information & Management, 46, 411–417. Chen, L.D., Gillenson, M.L., & Sherrell, D.L. (2002). Enticing online consumers: An extended technology acceptance perspective. Information & Management, 39, 705-719. Chen, Z. & Dubinsky, A.J. (2003). A conceptual model of perceived customer value in e-Commerce: A preliminary investigation. Psychology & Marketing, 20 (4), 323-347. Chen, D., Mocker, M., Preston, D.S., & Teubner, A. (2010). Information systems strategy: Reconceptualization, measurement, and implications. MIS Quarterly, 34 (2), 233-259. Chesney, T. (2008). Measuring the context of information system use. Journal of Information Technology Management, 19(3), 9-20. Chew, K.-W., Shingi, P. M., & Ahmad, M.I. (2006). TAM derived construct of perceived customer value and online purchase behavior: An empirical exploration. In R. Suomi, R. Cabral, J.F. Hampe, A. Heikkilä, J. Järveläinen, E. Koskivaara (Eds.). Project E-Society: Building Bricks – 6th IFIP International Conference on e-Commerce, e-Business, and e-Government (13E 2006), October 11–13, 2006, 226, (pp. 215-227). IFIP Advances in Information and Communication Technology. 123

Chircu, A.M., & Kauffman, R.J. (2000). Limits to value in electronic commerce-related IT investments. Journal of Management Information Systems, 17(2), 59-80. Cho, J. & Trent, A. (2006). Validity in qualitative research revisited. Qualitative Research, 6 (3), 319–340. Churchill, G. A., & Surprenant, C. (1982). An investigation into the determinants of customer satisfaction. Journal of Marketing Research, 19, 491-504. COFEK (2013). Nakumatt CEO: Kenya not ready for online shopping. Article available online at www.cofek.co.ke Collins, K.M.T., Onwuegbuzie, A.J. & Jiao, Q.G. (2007). A mixed methods investigation of mixed methods sampling designs in social and health science research. Journal of Mixed Methods Research, 1(3), 267-294. Cottet, P., Lichtlé, M.C., & Plichon, V. (2006). The role of value in services: A study in a retail environment. Journal of Consumer Marketing, 23(4), 219–227. Cooper, D., & Schindler, P. (2008). Business Research Methods (International Edition), Mc-Graw-Hill Education. Cooper, R. B. & Zmud, R.W. (1990). Information technology implementation research: A technological diffusion approach. Management Science, 36 (2), 123-139. Cox, D.F., & Rich, S.U. (1964). Perceived risk and consumer decision making - the case of telephone shopping‖, Journal of Marketing Research, Vol. 1, pp. 2-39. Crego, E.T. & Schiffrin, P.D. (1995). Customer – centered reengineering: Remapping for total customer value. Burr Ridge, Illinois: Irwin. Creswell, J. W. (2003). Research Design Qualitative, Quantitative, and Mixed Methods Approaches. London: Sage Publications. Creswell, J. W. & Miller, D. L. (2000). Determining validity in qualitative inquiry. Theory in Practice, 39 (3), 124 – 130.

Cunningham, S. M. (1967), The major dimensions of perceived risk. In D. Cox (Ed.), Risk taking and information handling in consumer behavior (pp. 82-109). Harvard: Harvard University Press. Daniel, E. & Klimis, G.M. (1999). The impact of electronic commerce on market structure: An evaluation of the electronic market hypothesis. European Management Journal, 17 (3), 318-325. Daniel, J. (2012). Sampling essentials: Practical guidelines for making sampling choices. Sage Publications: Thousand Oaks, CA.

124

Davis, F.D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13 (3), 319-340. Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1992). Extrinsic and intrinsic motivation to use computers in the workplace. Journal of Applied Social Psychology, 22 (14), 1111-1132. Davis, F. D., Bagozzi, R.P., & Warshaw, P. R. (1989). User acceptance Of computer technology: A comparison of two theoretical models. Management Science, 35 (8), 982-1003. De Leeuw, E.D. (2005). To mix or not to mix data collection modes in surveys. Journal of Official Statistics, 21 (2), 233-255. De Leeuw, E.D. & Hox, J.J. (2011). Internet surveys as part of a mixed mode design. In: M. Das, P. Ester and L. Kaczmirek (Eds). Social and behavioral research and the internet: Advances in applied methods and research strategies, (pp. 45– 76). New York: Taylor & Francis Group. De Lone, W.H., & McLean, E.R. (1992). Information systems success: The quest for the dependent variable. Information Systems Research, 3(1), 60-95. De Lone, W.H., & McLean, E.R. (2002). Information systems success revisited. In R. H. Sprague, Jr. (Ed.), Proceedings of the Thirty Fifth Hawaii International Conference on Systems Science (CD-ROM). Los Alamitos, CA: IEEE, Computer Society Press. De Lone, W., & McLean, E. (2003). The DeLone and McLean model of information systems success – A 10 year update. Journal of Management Information Systems, 19 (4), 9-30. De Lone, W., & McLean, E. (2004). Measuring e-commerce success: Applying the DeLone & McLean information systems success model. International Journal of Electronic Commerce, 9 (1), 31-47. Devaraj, S., Fan, M., & Kohli, R. (2002). Antecedents of B2C channel satisfaction and preference: Validating e-commerce metrics. Information Systems Research, 13 (3), 316-333. DiCicco-Bloom, B. & Crabtree, B.F. (2006). The qualitative research interview: making sense of qualitative research. Medical Education, 40: 314–321. doi:10.1111/j.1365-2929.2006.02418.x Dholakia, R. & Uusitalo, O. (2002). Switching to electronic stores: Consumer characteristics and the perception of shopping benefits. International Journal of Retail and Distribution Management, 27 (4), 154-165. Doane, D. P., & Seward, L. E. (2011). Applied Statistics in Business and Economics (3rd Ed.). New York, NY: McGraw-Hill Irwin.

125

Doherty, N. F., & Ellis-Chadwick, F. (2010). Internet retailing: The past, the present and the future. International Journal of Retail & Distribution Management, 38(11/12), 943-965. Easterby-Smith, M., Thorpe, R., & Lowe, A. (2002). Management Research: an introduction. London: Sage Eid, M. I. (2011). Determinants of e-commerce customer satisfaction, trust, and loyalty in Saudi Arabia, Journal of Electronic Commerce Research, 12 (1), 78 – 93. Elfving, J. & Sundqvis, J. (2011). The supplier selection process in the swedish pulp industry: A contingency perspective. Unpublished Masters Thesis. Luleå University of Technology. Retrieved from https://pure.ltu.se/ws/files/33167404/LTU-EX-2011-33136501.pdf El-Kasheir, D., Ashour, A.S., & Yacout, O.M. (2009). Factors affecting continued usage of internet banking among Egyptian customers. Communications of the IBIMA, 9, 252-263. ISSN: 1943-7765. Ethier, J., Hadaya, P., Talbot, J., & Cadieux, J. (2006). B2C web site quality and emotions during online shopping episodes: an empirical study, Information & Management, 43 (5), 627–639. Fichman, R.G., & Kemerer, C.F. (1999). The illusory diffusion of innovation: An examination of assimilation gaps. Information Systems Research, 10 (3), 255275. Field, A. (2005). Discovering statistics using SPSS (2nd Ed.). London: Sage. Fornell, C, Johnson, M. D., Anderson, E. W., Cha, J., & Bryant, B. E. (1996).The American customer satisfaction index: Nature, purpose and findings. Journal of Marketing. 60, 7-18. Forrester (2011a), US online retail forecast, 2010 To 2015. Forrester Research Inc. Forrester (2011b), European online retail forecast, 2010 to 2015. Forrester Research Inc. Forsythe, S., Chuanlan, N., Shannon, D., & Gardner, L.C. (2006), Development of a scale to measure the perceived benefits and risks of online shopping, Journal of Interactive Marketing, 20 (2), 55 – 75. Forsythe, S.M., & Shi, B. (2003). Consumer patronage and risk perceptions in internet shopping. Journal of Business Research, 56 (11), 867-876. Foxall, G.R., Goldsmith, R.E., & Brown, S. (1998). Consumer psychology for marketing, (2nd Ed.). London: Thomson Learning. Furr, M. (2011). Scale Construction and Psychometrics for Social and Personality Psychology. Thousand Oaks, CA: Sage Publications.

126

Gefen, D., & Straub. D.W. (1997). Gender differences in the perception and use of email: An extension to the technology acceptance model. MIS Quarterly, 21 (4), 389-400. Gefen, D., & Keil, M. (1998). The impact of developer responsiveness on perceptions of usefulness and ease of use: An extension of the technology of the technology acceptance model. Database for Advances in Information Systems, 29 (2), 3549. Gefen, D., Karahanna, E., & Straub, D.W. (2003). Inexperience and experience with online stores: The importance of TAM and trust. IEEE transactions on engineering management, 50 (3), 307-21. Giese, J.L., & Cote, J.A. (2000). Defining customer satisfaction. Academy of Marketing Science Review, (1), 1 – 24. Gimpel, G. (2011). Value-driven adoption and consumption of technology: Understanding Technology Decision Making. Unpublished Phd Thesis. Retrieved from http://openarchive.cbs.dk/bitstream/handle/10398/8326/Gregory%20Gimpel.pdf ?sequence=1 (accessed on August, 2012). Grabner-Krauter, S., & Kaluscha, E. A. (2003). Empirical research on on-line trust: A review and critical assessment. International Journal of Human-Computer Studies, 58 (6), 783-812. Grembowski, D. (2001). The Practice of Health Program Evaluation. Sage: Thousand Oaks, CS. Grgecic, D., & Rosenkranz, C. (2011). Reconceptualizing IT use in the post-adoptive context, European Conference on Information Systems (ECIS) 2011 Proceedings. Paper No. 59. Retrieved from www.lse.ac.uk/asp/aspecis/20110059.pdf. Grigoroudis, E., & Siskos, Y. (2010). Customer satisfaction evaluation: Methods for measuring & evaluation service quality. New York: Springer. Ha, H., & Janda, S. (2008). An empirical test of a proposed customer satisfaction model in e-services. Journal of Services Marketing, 22 (5), 399 – 408. Hagel, J., & Singer, M. (1999). Net worth: Shaping markets when customers make the rules. Boston, MA: Harvard Business School Press. Hair, J.F., Babin, B., Money, A.H., & Samuel, P. (2003). Essentials of business research. New York: John Wiley & Sons. Hair, J., Anderson, R., Tatham, R., & Black, W. (2006), Multivariate data analysis (6th Ed.). Prentice Hall: Upper Saddle River, NJ. Hair Jr., J.F., Black, W.C., Babin, B.J., & Anderson, R. E. (2010). Multivariate data analysis: A global perspective. Upper Saddle River, N.J.: Pearson Education.

127

Halcomb, E. J., & Andrew, S. (2005). Triangulation as a method for contemporary nursing research. Nurse Researcher (in press). Hansen, T. (2005). Consumer adoption of online grocery buying: A discriminant analysis. International Journal of Retail & Distribution Management, 33(2), 101–121. Hansen, T. (2007). Determinants of consumers' repeat online buying of groceries, The International Review of Retail, Distribution and Consumer Research, 16 (1), 93-114. Hauser, J., Tellis, G.J., & Griffin, A. (2006). Research on innovation: A review and agenda for marketing science. Marketing Science, 25(6), 687-717. Haynes, P.J. & Taylor, V.A. (2006). An Examination of Strategic Practices in Online Retailing, Journal of Internet Commerce, 5(3), 1-26. Hernández, B., Jiménez, J., & José Martín, M. (2011). Age, gender and income: Do they really moderate online shopping behaviour? Online Information Review, 35 (1), 13 – 133. Hoffman D.L., Novak, T., & Schlosser A. (2000). The evolution of the digital divide: How gaps in internet access may impact electronic commerce. Journal of Computer-Mediated Communication, 3(5), Retrieved from http://jcmc.indiana.edu/vol5/issue3/hoffman.html. Homburg, C., Klarmann, M., Reimann, M. & Schilke, O. (2012). What drives key informant accuracy? Journal of Marketing Research, XLIX, 594–608. Hong, S., Thong, J.Y.L., & Tam, K.Y. (2006). Understanding continued information technology usage behavior: A comparison of three models in the context of mobile internet, Decision Support Systems, 42, 1819–1834. Hossain, M.A. & Quaddus, M. (2012), Expectation-confirmation theory in information systems research: A review and analysis, In Y.K. Dwivedi et al. (Eds.), Information systems theory: Explaining and predicting our digital society, 1 (28), Integrated Series in Information Systems, 440-470. Hoyer, W.D., & MacInni, D.J. (2008), Consumer Behavior (5th Ed.). Mason, OH: South Western – Cengage Learning. Hsieh, S.-F. & Shannon, S. E. (2005). Three approaches to qualitative content analysis. Qualitative Health Research, 15(9), 1277-1288 Hsu, H. (2006). An empirical study of web site quality, customer value, and customer satisfaction based on e-shop. The Business Review, 5(1), 190-193. Hu, F-L., & Chuang, C.C. (2012). A study of the relationship between the value perception and loyalty intention toward an e-retailer website. Journal of Internet Banking and Commerce, 17(1), 1-18.

128

IBM East Africa (2012). A vision of a smarter city: How Nairobi can lead the way into a prosperous and sustainable future. Retrieved from http://www05.ibm.com/za/office/pdf/IBM_-_A_Vision_of_a_Smarter_City_-_Nairobi.pdf IEA, (2011). Nairobi city scenarios. Nairobi: Institute of Economic Affairs. Retrieved from www.ieakenya.or.ke Im, I., Kim, Y., & Han, H.-Y. (2008). The effects of perceived risk and technology type on users‘ acceptance of technologies. Information & Management, 45(1), 1-9. Ibeh, K.I.N., Brock, J.U. & Zhou, J. (2004). Drop and pick survey among industrial populations: Conceptualisations and empirical evidence. Industrial Marketing Management, 33, 155–65. Jackson, K.M., & Trochim, W.M.K. (2002). Concept mapping as an alternative approach for the analysis of open-ended survey responses. Organizational Research Methods, 5 (4), 307-336. Jacoby, J., & Kaplan, L. (1972). The components of perceived risk. In M. Venkatesan (Ed.), Proceedings, 3rd Annual Conference (pp. 383 – 393). Chicago, IL: Association for Consumer Research. Jarvenpaa, S.L., & Todd, P. A. (1996). Consumer reactions to electronic shopping on the world wide web. International Journal of Electronic Commerce, 1(2), 5988. Jarvenpaa, S.L., & Tractinsky, N. (1999). Consumer trust in an internet store: a crosscultural validation. Journal of Computer-Mediated Communication, 5(2). Jiang, P., & Rosenbloom, B. (2005). Customer intention to return online: Price perception, attribute-level performance, and satisfaction unfolding over time. European Journal of Marketing, 39(1/2), 150-174. John, G. & Reve, T. (1892). The reliability and validity of key informant data from dyadic relationships in marketing channels. Journal of Marketing Research, Vol XIX, 517-524 Johnson, M.S., Sivadas, E., & Garbarino, E. (2008). Customer satisfaction, perceived risk and affective commitment. Journal of Services Marketing, 22(5), 353 – 362. Jones, K., & Leonard, L. N. K. (2007), Consumer-to-consumer electronic commerce: A distinct research stream, Journal of Electronic Commerce in Organizations, 5 (4), 39-54. Judge, T.A., & Bono, J.E. (2001). Relationship of core self-evaluations traits – selfesteem, generalized selfefficacy, locus of control, and emotional stability – with job satisfaction and job performance: a meta-analysis. Journal of Applied Psychology, 86, 80–92.

129

Juma, V. (2010, August 16th). Online shopping keeps consumers out of revenue authority‘s reach‖. Business Daily. Retrieved from http://all Africa.com/stories/201008160015.htm. Kalafatis, S.P., Ledden, L., & Mathioudakis, A. (n.d). Re-specification of the theory of consumption values. Accessed on February 28th 2013. Retrieved from: http://eprints.kingston.ac.uk/18098/1/Kalafatis-S-18098.pdf. Kamarulzaman, Y. (2008). Modelling consumer adoption of internet shopping. Communications of the IBIMA, 5(26), 217-227. Katz, D (1960). Functional theory of attitudes. Social Cognition, 11, 163 – 204. Kenya ICT Board (2012). Tandaa grant applicants. Nairobi: Kenya ICT Board. Kenya Postel Directories (2012). Official Nairobi 2013 Edition. Nairobi: Kenya Postel Directories. Kermeliotis, T. (2011, October 10). Web Savvy Africans Fuel growth in Online Shopping. CNN Marketplace Africa. Retrieved from http://edition.cnn.com/2011/10/10/business/online-shopping-nigeria Kim, H., Lee, I., & Kim, J. (2008). Maintaining continuers vs. converting discontinuers: relative importance of post-adoption factors for mobile data services. International Journal of Mobile Communications, 6(1), 108-132. Klein, L. R. (1962). An introduction to econometrics. Prentice-Hall: New Jersey. Kotler, P. (1986). Marketing Management: Analysis, planning and control. Englewood Cliffs, New Jersey: Prentice-Hall International, Kotler, P., & Armstrong, G. (2000) Marketing (5th ed.). Prentice-Hall: Englewood Cliffs, NJ. Koufaris, M. (2002). Applying the technology acceptance model and flow theory to online consumer behavior, Information Systems Research, 13 (2), 205. Krippendorff, K. (1980). Content analysis: An introduction to its methodology (Vol. 5). Newbury Park, CA: Sage. Kumar, V., Aaker, D.A., & Day, G.S. (2002). Essential of Marketing Research (2nd Ed.). Hoboken NJ: John Wiley & Sons Inc. Latham, B. (2007). Sampling: What is it? Retrieved online from http://webpages.acs.ttu.edu/rlatham/Coursework/5377(Quant))/Sampling_Meth odology_Paper.pdf Lawrence, M. & Low, G. (1993). Exploring individual user satisfaction within user-led development. MIS Quarterly, 6, 195-208. LeCompte, M.D. & Goetz, J.P. (1982). Problems of reliability and validity in ethnographic research. Review of Educational Research, 52(1), 31-60.

130

Lee, D., Park, J., & Ahn, J. (2000). On the explanation of factors affecting e-commerce adoption, Working Paper. Retrieved from http://misrc.umn.edu/workingpapers/fullpapers/2000/0025_120100.pdf Legris, P., Ingham, J., & Collerette, P. (2003). Why do people use information technology? A critical review of the technology acceptance model. Information and Management, 40(3), 191–204. Lemm, K. (2010). Stratified sampling. In N.J. Salkind (Ed.). Encyclopaedia of research design (pp. 1451-1454). Thousand Oaks, CA.: Sage Publications. Li, H., Kuo, C. & Russell, M. G. (1999). The Impact of Perceived Channel Utilities, Shopping Orientations, and Demographics on the Consumer's Online Buying Behavior, Journal of Computer-Mediated Communication, Vol. 5, No. 2, 1999. Liebermann, Y., & Stashevsky, S. (2002). Perceived risks as barriers to internet and ecommerce usage. Qualitative Market Research: An International Journal, 5(4), 291-300. Lim, W.M., & Ting, T.H. (2012). E-shopping: An analysis of the technology acceptance model. Modern Applied Science. 6(4), 49 -62. Limayem, M., Hirt, S.G., & Cheung, C.M.K. (2007). How habit limits the predictive power of intention: The case of information systems continuance. MIS Quarterly, 31(4), 705-37. Lincoln, Y.S., & Guba, E.G. (1985). Naturalistic inquiry. Sage Publications. Lindsay, P., & Norman, D. A. (1977). Human information processing: An introduction to psychology. Harcourt Brace Jovanovich, Inc. Liou, F-M. (2008). Fraudulent financial reporting detection and business failure prediction models: A comparison. Managerial Auditing Journal, 23 (7), 660662. Liu, C. (2007), Modelling consumer adoption of the internet as a shopping medium: An integrated perspective. Cambria Press. Liu, C., & Forsythe, S. (2010). Post-adoption online shopping continuance. International Journal of Retail & Distribution Management, 38(2), 97-114. Lucas, H.C. (1975). Why Information Systems Fail. New York, NY: Columbia University Press. Magutu, P.O., Mwangi, M., Nyaoga, R.B., Ondimu, G.M., Kagu, M., Mutai, K., Kilonzo, H., & Nthenya, P. (2011). E-Commerce products and services in the banking industry: The adoption and usage in commercial banks in Kenya, Journal of Electronic Banking Systems, 2011, 1-19.

131

Mawhinney, C.H. (1990). A study of computer use by knowledge workers: user satisfaction versus system use. Proceedings of the Annual Meeting of Decision Sciences Institute. San Diego, CA, 954-956. Mallet, D. (2006). Sampling and Weighting. In R. Grover & M. Vriens (Eds.), The Handbook of Marketing Research (pp. 159-177). Thousand Oaks, CA.: Sage. Marshall, M.N. (1996). The key informant technique. Family Practice, 13(1), 92-97. Martin, S.S., & Camarero, C. (2009). How perceived risk affects online buying. Online Information Review, 33(4), 629-654. Maxcy, S. J. (2003). Pragmatic threads in mixed methods research for multiple modes: the search for multiple modes of inquiry and the end of the philosophy of formalism, in: A. Tashakkori, and C. Teddlie (Eds) Handbook of Mixed Methods in Social and Behavioral Research, pp. 51–89 (Thousand Oaks, CA: Sage). Mazzarol, T. (1998). Critical success factors for international education marketing, International Journal of Educational Management, 12(4). McDougall, G.H.G., & Levesque, T. (2000). Customer satisfaction with services: Putting perceived value into the equation. Journal of Services Marketing, 14(5), 392-410. McKinney, V., Yoon, K., & Zahedi, F. (2002). The measurement of web-customer satisfaction: An expectation and disconfirmation approach. Information Systems Research, 13(3), 296-315. McMillan, J.H., & Schumacher, S. (2001). Research in education. A conceptual introduction (5th Ed.). New York: Longman. Mohsavi, A., & Ghaedi, M. (2012). An examination of the effects of perceived value and attitude on customers‘ behavioral intentions in shopping. African Journal of Business Management, 6(5), 1950-1959. Molla, A., & Licker, P. (2001). E-commerce systems success: An attempt to extend and re-specify the Delone and Mclean Model of information systems success. Journal of Electronic Commerce Research, 2(4), 1-11. Monsuwé, T.P., Dellaert, B.G.C., & De Ruyter, K. (2004), What drives consumers to shop online: A literature review. International Journal of Service Industry Management, 15 (1), 102-121. Moore, G. C., & Benbasat, I. (1991). Development of an instrument to measure the perceptions of adopting an information technology innovation. Information Systems Res., 2, 192–222. Moosbrugger, H., Schermelleh-Engel, K., Kelava; A. & Klein, A. G. (in press). Testing multiple nonlinear effects in structural equation modeling: A comparison of alternative estimation approaches. Invited Chapter in T. Teo & M.S. Khine

132

(Eds.), Structural Equation Modelling in Educational Research: Concepts and Applications. Rotterdam, NL: Sense Publishers. Mordkoff, T. (2011). The assumption(s) of normality. Retrieved from http://www2.psychology.uiowa.edu/faculty/mordkoff/GradStats/part%20I/I.07 %20normal.pdf Moschis, G.P. (1994). Consumer behaviour in later life: multidisciplinary contributions and implications for research. Journal of the Academy of Marketing Science, 22, 195-204. Mottner, S., Thelen, S., & Karande, K. (2002). A typology of internet retailing. Journal of Marketing Channels, 10 (1), 3-23. Mugenda O. M. &Mugenda, A.G. (2003). Research methods: Quantitative and qualitative approaches. Nairobi: African Centre of Technology Studies. Murray, C. (2009). Diffusion of innovation theory: A bridge for the research-practice gap in counseling‖, Journal of Counseling & Development, 87 (1), 108-16. Nemes, S., Jonasson, J. M., Genell, A., & Steineck, G. (2009). Bias in odds ratios by logistic regression modelling and sample size. BMC Medical Research Methodology, 9(56), 1-5. Nevo, D., & Furneau, B. (2009). The role of expectations in information systems development. In Whitworth & de Moor (Eds.). Handbook of Research on SocioTechnical Design and Social Networking Systems Vol. 1 (pp. 298 – 312). New York: Information Science Reference. Nimon, K. (2010), Regression commonality analysis: Demonstration of an SPSS solution. Multiple linear regression viewpoints, 36(1), 10 – 17. Nulty, D.D. (2008). The adequacy of response rates to online and paper surveys: What can be done? Assessment & Evaluation in Higher Education. 33(3), 301–314. Nyshadham, E.A., & Ugbaja, M. (2006), A study of e-commerce risk perceptions among b2c consumers: A two country study, Proceedings of 19th Bled eConference, Retrieved from https://domino.fov.unimb.si/proceedings.nsf/0/597b2c70a054655 cc12571800032f9b0/$FILE/36_Nyshadham.pdf Nysveen, H., Pedersen, P.E., & Thornbjørnsen, H. (2005). Explaining intention to use mobile chat services: Moderating effects of gender, Journal of Consumer Marketing, 22 (5), 247-256. Okuttah, M. (Wednesday, July 2014). Report says Kenya lags behind peers in ecommerce. Business Daily. Retrieved online from http://www.businessdailyafrica.com/Report-says-Kenya-lags-behind-peers-in-ecommerce/-/1248928/2369806/-/y1m5jk/-/index.html

133

Oliver, R.L. (1977). Effect of expectation and disconfirmation on post-exposure product evaluations: An alternative interpretation, Journal of Applied Psychology, 62, 480-6. Oliver, R.L. (1980). A cognitive model for the antecedents and consequences of satisfaction decisions. Journal of Marketing Research, 17, 460-9. Oliver, R.L. (1993). Cognitive, affective, and attribute bases of the satisfaction response. Journal of Consumer Research, 20, 418-30. Oliver, R.L. (1994). Conceptual issues in the structural analysis of consumption emotion, satisfaction, and quality: evidence in a service setting. Advances in Consumer Research, 21(1), 16-22. Oliver, R.L. (1997). Satisfaction: A Behavioral Perspective on the Consumer. McGraw-Hill, New York, NY. Oliver, R. L. (2006). Customer satisfaction research. In Grover, R. & Vriens, M., (eds.), The handbook of marketing research: Uses, misuses and future advances, Thousand Oaks, California: Sage Publications Inc., pp. 569 – 587. Oliver, R.L. (2010). Satisfaction: A Behavioral Perspective on the Consumer, McGraw-Hill, New York, NY. Oliver, R. L. & Westbrook, R. A. (1982). On the factor structure of satisfaction and related post-purchase measures. In R.L. Day & H.K. Hunt, (eds.), Consumer satisfaction, dissatisfaction, and complaining behavior, Vol. 5, Bloomington: Indiana University School of Business, pp. 11-14. Ortiz de Guinea, A. & Markus, M.L. (2009). Why break the habit of a lifetime? Rethinking the roles of intention, habit, and emotion in continuing information technology use. MIS Quarterly, 33(3), 433-444. Osman, S., Yin-Fah, B.C., & Hooi-Choo, B. (2010). Undergraduates and online purchasing behavior. Asian Social Science, 6(10), 133-146. Overby, J.W., & Lee, E.J. (2006). The effects of utilitarian and hedonic online shopping value on consumer preference and intentions. Journal of Business Research, 59, (10-11), 1160-1166. Ozen, H., & Kodaz, N. (2012). Utilitarian or hedonic? A Cross cultural study in online shopping, Organizations and Markets in Emerging Economies, 3 (2), 80 – 90. Pallant, J. (2007). SPSS Survival Manual (3rd Edition). Crows West: New South Wales. Pansiri, J. (2005). Pragmatism: A methodological approach to researching strategic alliances in tourism. Tourism and Hospitality Planning & Development, 2(3), 191-206. Park, J., Lee, D., & Ahn, J. (2004). Risk-focused e-commerce adoption model: A crosscountry study. Journal of Global Information Technology Management, 7(2), 6– 30. 134

Parthasarathy, M., & Bhattacherjee, A. (1998). Understanding post-adoption behavior in the context of online services. Information Systems Research, 9, 362. Paxson, M. C. (1992). Follow-up mail surveys. Industrial Marketing Management, 21, 195-201. Petter, S., DeLone, W., & McLean, E (2008). Measuring information systems success: Models, dimensions, measures, and interrelationships. European Journal of Information Systems, 17, 236–263. Phellas, C.N., Bloch, A., & Seale, C. (2012). Structured methods: Interviews, questionnaires and observation. In Seale, C. (Eds.), Researching society and culture (3rd Edition) (pp. 182-205). UK: Sage Publications. Pitt, L., Berthon, P., & Berthon, J.P. (1999), Changing channels: The impact of the Internet on distribution strategy. Business Horizons, 42(2), 19-28. Pookulangara, S., & Natesan, P. (2010). Examining consumers‘ channel-migration intention utilizing channel of planned behavior: A multi-group analysis. International Journal of Electronic Commerce Studies, 1(2), 97 – 116. Rai, A., Lang, S. & Welker, R. (2002). Assessing the validity of IS success models: An empirical test and theoretical analysis. Information Systems Research, 13(1), 5069. Ramayah, T. & Lee, J.W.C. (2012), System characteristics, satisfaction and e-learning usage: A structural equation model, The Turkish Online Journal of Educational Technology, 11(2), 196-206. Ratchford, B.T., Talukdar, D., & Lee, M.S. (2001). A model of consumer choice of the internet as an information source. International Journal of Electronic Commercial, 5(3), 7-22. Ravald, A., & Groonroos, C. (1996). The value concept and relationship marketing. European Journal of Marketing, 30 (2), 19-30. Razali, N.M., & Wah, Y.B. (2011). Power comparisons of Shapiro-Wilk, KolmogorovSmirnov, Lilliefors and Anderson-Darling Tests. Journal of Statistical Modelling and Analytics. 2(1), 21-33. Reichheld, F.F. (1996). The loyalty effect: The hidden force behind growth, profits, and lasting value. Harvard Business School Press, Boston. Roberts, P., Priest, H., & Traynor, M. (2006). Reliability and validity in research. Nursing Standard. 20 (44), 41-45. Robey, D. (1979). User attitudes and management information system use. Academy of Management Journal, 22 (3), 527 – 538.

135

Robinson, H., Riley, F. D., Rettie, R., & Rolls-Wilson, G. (2007). The role of situational variables in online grocery shopping in the UK. The Marketing Review, 7(1), 89-106. Rogelberg, S. & Stanton, J. (2007). Understanding and dealing with organizational survey non-response. Organizational Research Methods, 10, 195–209. Rogers, E.M. (1962). Diffusion of Innovations. The Free Press. th

Rogers, E.M. (1995). Diffusion of Innovations, (4 Ed.). New York: The Free Press. th

Rogers, E.M. (2003). Diffusion of Innovations (5 Ed.). New York: The Free Press. Rogers, S., & Harris, M. (2003). Gender and e-commerce: An exploratory study. Journal of Advertising Research, 43(3), 1-8. Rogers, W. A., van Ittersum, K., Capar, M., Caine, K.E., O‘Brien, M. A., Parsons, L.J., & Fisk, A.D. (2006), Understanding technology acceptance: Phase 1 – literature review and qualitative model development, Technical Report HFA-TR-0602, Atlanta, GA: Georgia Institute of Technology, School of Psychology – Human Factors and Aging Laboratory. Rohm, A. J. & Swaminathan, V. (2004). A typology of online shoppers based on shopping motivations. Journal of Business Research, 57 (7), 748-757. Rojas-Méndez, J.I. & Davies, G. (n.d.). Drop-off/Pick-up as a method of maximizing response rates in self-administered surveys. Working Paper No. 434. Manchester Business School. ISSN 0954-7401. Retrieved from http://www.mbsportal.bl.uk/secure/subjareas/resmethods/mubs/wp/138940WP4 34_01.pdf Roslow, S., Li, T., & Nicholls, J.A.F. (2000). Impact of situational variables and demographic attributes in two seasons on purchase behavior. European Journal of Marketing, 34(9/10), 1167-1180. Rubin, A. & Babbie, E. R. (2011), Research Methods for Social Work (7th Ed). Belmonst, CA: Cengage Learning. Rumsey, D.J. (2011). Statistics for Dummies (2nd Ed.). ISBN: 978-0-470-91108-2 Ryan, G. W. & Bernard, H. R. (2003). Techniques to Identify Themes. Field Methods, 15(1), 85–109. Saeed, K., & Abdinnour-Helm, S. (2008). Examining the effects of information system characteristics and perceived usefulness on post adoption usage of information systems. Information and Management, 45(6), 376–386. Sahaf, M. A. (2008), Strategic marketing: Making decisions for strategic advantage, New Delhi: Prentice Hall of India.

136

Sánchez-Fernández, R., & Iniesta-Bonillo, M. A. (2007). The concept of perceived value: A systematic review of research, Marketing Theory, 7(4), 427–451. Saunders, M.N.K., Lewis, P., & Thornhill, A. (2009). Research methods for business Students (5th Ed.). Harlow, United Kingdom: FT Prentice Hall. Schaffer, E. (2000). A better way for web design, InformationWeek, 784(1), 194-194. Schatz, E. (2009). Nesting semi-structured interviews in surveys or censuses: More than the sum of the parts. Working Paper, Institute of Behavioural Science, University of Colorado at Boulder. Schaupp, L.C., & Bélanger, F. (2005), A conjoint analysis of online consumer satisfaction. Journal of Electronic Commerce Research, 6(2), 95 – 111.

Schewe, C.D. (1976). The management information system user: An exploratory behavioral analysis. The Academy of Management Journal, 19(4), 577-590. Schiffman, L.G., & Kanuk, L.L. (2010). Consumer behavior: Global edition, London: Pearson Higher Education. Sharma, S., Durand, R.M., & Gur-Arie, O. (1981). Identification and analysis of moderator variables. Journal of Marketing Research, 18(3), 291-300. Sharma, P., Chen, I.S.N., & Luk, S.T.K. (2012). Gender and age as moderators in the service evaluation process. Journal of Services Marketing. 26 (2), 102–114. Shenton, A. K. (2004). Strategies for ensuring trustworthiness in qualitative research projects. Education for Information, 22, 63–75. Seddon, P. (1997). A re-specification and extension of the DeLone and Mclean Model of IS success. Information Systems Research, 8(3), 240-253. Serenko, A., Turel, O., & Yol, S. (2006). Moderating roles of user demographics in the American customer satisfaction model within the context of mobile services. Journal of Information Technology Management, 17(4), 173-185. Sheth, J.N., Newman, B.I., & Gross, B.L. (1991a). Consumption values and market choices: Theory and applications. Cincinnati: South-Western Publishing Co. Sheth, J.N., Newman, B.I., & Gross, B.L. (1991b). Why we buy what we buy: A theory of consumption values. Journal of Business Research, 22(2), 159-170. Shih, C.-F., & Venkatesh, A. (2004). Beyond adoption: Development and application of a use-diffusion model. Journal of Marketing, 68(1), 59-72. Shim, S., Eastlick, M.A., Lotz, S.L. & Warrington, P. (2001). An online pre-purchase intention model: The role of intention in search. Journal of Retailing, 77 (3), 397-416.

137

Simon, H.A. (1955). A behavioral model of rational choice. Quarterly Journal of Economics, 69(1), 99-118. Simsek, Z., Veiga, J.F. & Lubatkin, M.H. (2005). Challenges and guidelines for conducting internet-based surveys in strategic management research. In D.J. Ketchen & D.D. Bergh (Eds.). Research methodology in strategy and management, Vol. 2 (pp.179-196). Amsterdam: Elsevier Publishing Limited. Sing, J. (1991). Understanding the structure of consumers‘ satisfaction evaluations of service delivery. Journal of the Academy of Marketing Science, 19(3), 223–244. Smith, T. J. (2008). Senior citizens and e-commerce websites: The role of perceived usefulness, perceived ease of use, and website usability. Informing Science: the International Journal of an Emerging Transdiscipline, 11, 59-83. Specht, N., Fichtel, S., & Meyer, A. (2007). Perception and attribution of employees' effort and abilities: The impact on customer encounter satisfaction. International Journal of Service Industry Management, 18(5), 534 – 554. Spencer, S.J., Zanna, M.P., & Fong, G.T. (2005). Establishing a causal chain: Why experiments are often more effective than mediational analysis in examining psychological processes. Journal of Personality and Social Psychology, 89, 845–851. Stewart, J. (1994). The psychology of decision making. In D. Jennings & S. Wattam, (Eds.). Decision Making: an Integrated Approach. London: Pitman, Straub, D., Limayem, M., & Kurahana-Evaristo, E. (1995). Measuring system usage: Implications for IS system testing, Management Science, 41(8), 1328-1342. Suhr, D.D. (2006). Exploratory or confirmatory factor analysis? Statistics and Data Analysis, 31. Retrieved online from from http://www2.sas.com/proceedings/sugi31/200-31.pdf Sundravej, T. (2006). Factors influencing b2c e-commerce adoption in organizations. IS Research Seminar, University of Missouri at Saint Louis, College of Business Administration. Retrieved from www.umsl.edu/~sundaravejf/ Sun, H., & Zhang, P. (2006). The role of moderating factors in user technology acceptance. International Journal of Human Computer Studies, 64(2), 53-78. Swan, J.E., Trawick, F., & Carroll, M.G. (1981). Effect of participation in marketing research on consumer attitudes towards research and satisfaction with a service, Journal of Marketing Research, 19, 356 – 363. Swanson, R.A., & Holton, E.F.III. (Eds.) (2005). Research in organizations: Foundations and methods of inquiry. San Francisco: Berrett-Koehler. Sweeney, J.C., & Soutar, G.N. (2001). Consumer perceived value: The development of a multiple item scale. Journal of Retailing, 77 (2), 203-220.

138

Sweet, S.A., & Grace-Martin, K. (2010). Data analysis with SPSS: A first course in applied statistics (4th Ed). Pearson. Szajna, B. (1996). Empirical evaluation of the revised technology acceptance model. Management Science, 42(1), 85-92. Swanson, E.B. (1988). Information system implementation: Bridging the gap between design and utilization. Irwin, Homewood, IL. Tan, S.J. (1999). Strategies for reducing consumers‘ risk aversion in internet shopping. Journal of Consumer Marketing, 16(2). Tan, M. & Teo, T.S.H. (2000), Factors influencing the adoption of internet banking. Journal of the Association for Information Systems, 1(5), 1- 44. Tashakkori, A. & Teddlie, C. (1998) Mixed methodology: Combining qualitative and quantitative approaches. Thousand Oaks, CA: Sage. Taylor, J.W. (1974). The role of risk in consumer behavior. Journal of Marketing, 38(2), 54–60 Taylor, S., & Todd, P. (1995a). Assessing IT usage: the role of prior experience, MIS Quarterly, 12, 561–570. Taylor, S., & Todd, P. (1995b). Understanding information technology usage: A test of competing models. Information Systems Research, 6(2), 144–176. Teddlie, C. & Tashakkori, A. (2003) Major issues and controversies in the use of mixed methods in the social and behavioral sciences, in: A. Tashakkori and C. Teddlie (Eds) Handbook of Mixed Methods in Social and Behavioral Research, pp. 3– 49 (Thousand Oaks, CA: Sage). Tongco, M.D.C. (2007). Purposive sampling as a tool for informant selection. Ethnobotany Research & Applications, 5, 147-158. Torkzadeh, G., & Dwyer, D. (1994). A path analytic study of determinants of information systems usage. OMEGA, 22(4), 339-48. Tornatzky, L.G., Eveland, J.D., Boylan, M.G., Hetzner, W.A., Jonson, E.C., Roitman, D., & Schneider, J. (1983). Innovation processes and their management: A conceptual, empirical and policy review of innovation process research. Washington, D.C.: National Science Foundation. Tornatzky, L.G., & Klein (1982). Innovation characteristics and innovation adoption implementation: A meta-analysis of findings. IEEE Transaction on Engineering Management, 29(1), 28–43. Tsiotsou, R. H. & Wirtz, J. (2012). Consumer behavior in a service context. In V. Wells & G. Foxall (Eds.), Handbook of Developments in Consumer Behavior (pp. 147201). Cheltenham, UK: Edward Elgar. Turban, E., & King, D. (2003). Introduction to e-commerce. New Jersey: Prentice Hall. 139

Turner, M., Kitchenham, B., Brereton, P., Charters, S., & Budgen, D. (2010). Does the technology acceptance model predict actual use? A systematic literature review. Information and Software Technology, 52, 463–479. UN-HABITAT (2006). Nairobi urban sector profile. Nairobi, Kenya: United Nations Human Settlements Programme. United States Agency for International Development (USAID), (2013). Technical note on conducting mixed-method evaluations. Monitoring and Evaluation Series. Version 1 (June 2013). U.S. Census Bureau (2004). Monthly retail surveys. (February 23). Retrieved from www.census.gov Van Ittersum, K., & Feinberg, F.M. (2010). Cumulative timed intent: A new predictive tool for technology adoption. Journal of Marketing Research, 47, 808-822. Van Slyke, C., Comunale, C., & Belanger, F (2002). Gender differences in perceptions of web-based shopping. Communications of the ACM, 45(7), 82-86. Van Slyke, C., Lou, H., & Day, J. (2002b). The impact of perceived innovation characteristics on intention to use groupware. Information Resource Management Journal, 15(1), 5-12. Van Slyke, C., Belanger, F., & Comunale, C. (2004). Factors influencing the adoption of web-based shopping: The impact of trust. DATA BASE for Advances in Information Systems, 35(2), 32-49. Venkatesh, V. (2000). Determinants of perceived ease of use: Integrating control, intrinsic motivation, and emotion into the technology acceptance model. Information Systems Research, 11, 342-365. Venkatesh, V., & Brown, S.A. (2001). A longitudinal investigation of personal computers in homes: Adoption determinants and emerging challenges‖, MIS Quarterly, 25(1), 71-102. Venkatesh, V., & Davis, F.D. (2000). A theoretical extension of the Technology Acceptance Model: Four longitudinal field studies‖, Management Science, 46, 186–204. Venkatesh, V., Morris, M.G., Davis, G.B., & Davis, F.D. (2003). User acceptance of information technology: Toward a unified view. MIS Quarterly, 27(3), 425-478. Venkatesh, V., Thong, J.Y.L., & Xin, H. (2012). Consumer acceptance and use of information technology: Extending the unified theory of acceptance and use of technology, MIS Quarterly, 36(1), 157-178. Verhoef, P.C., & Langerak, F. (2001). Possible determinants of consumers‘ adoption of electronic grocery shopping in the Netherlands. Journal of Retailing and Consumer Services, 8, 275-285.

140

Vrechopoulos, A., Siomkos, G., & Doukidis, G. (2001). Internet shopping adoption by Greek Ccnsumers. European Journal of Innovation Management, 4 (3), 142152. Walsh, G., Evanschitzky, H., & Wunderlich, M. (2008). Identification and analysis of moderator variables: Investigating the customer satisfaction-loyalty link. European Journal of Marketing, 42 (9), 977 - 1004 Wareham, J., Zheng, J.G., & Straub, D. (2005). Critical themes in e-commerce research: A meta analysis, Journal of Information Technology, 20, 1-19. Weber, R. (1990). Basic content analysis (2nd ed.). Newbury Park, CA: Sage. Weltevrenden, J. W. J. & Boschma, R.A. (2008). Internet strategies and the performance of Dutch retailers. Journal of Retailing & Consumer Services, 15, 163-178. Wen, C., Prybutok, V.R., & Chenyan, X. (2011). An integrated model for online customer re-purchase intention, Journal of Computer Information Systems, 1423. Westbrook, R., & Oliver, R. (1981). Developing better measures of customer satisfaction: some preliminary measures. In K. Monroe (Ed.). Advances in Consumer Research, Vol. 8. (pp. 94-99). Whyte, G., Bytheway, A., & Edwards, C. (1997). Understanding user perceptions of information system success. Journal of Strategic Information Systems, 6 (1), 35– 68. Wigand, R.T. (1997). Electronic commerce: Definition, theory, and context. The Information Society, 13, 1-16. Wilkie, W.L. (1994). Consumer behavior. New York: Von Hoffman Press. Willemsen, D. (2010). The measurement of customer satisfaction: Existing research, comparison of different methods, and critical appraisals. Scholarly Research Paper. Norderstedt: GRIN Verlag. Wilson, A. (2006). Marketing research: An integrated approach. (2nd Ed.). Gosport: Prentice Hall. Woodruff, R.B. (1997). Customer value: The next source for competitive advantage. Journal of the Academy of Marketing Science, 25 (2), 139-153. World Bank (December 2010), At the tipping point? The implications of Kenya‘s ICT revolution. Kenya Economic Update, (Ed. 3). Nairobi: World Bank Wu, A.D., & Zumbo, B.D. (2008). Understanding and using mediators and moderators, Soc Indic Res, 87, 367–392. Wu, S. (2010), Goodness-of-tests for logistics regression, Unpublished PhD Thesis, Florida State University. Retrieved from 141

http://etd.lib.fsu.edu/theses/available/etd-10272010194400/unrestricted/Wu_S_Dissertation_2010.pdf Xia, W., & Lee, A. (2000). The influence of persuasion, training and experience on user perceptions and acceptance of IT innovation. Proceeding of the 21st international conference on Information Systems, Brisbane, Queensland, Australia. Retrieved April 18, 2006, from The ACM Portal Database. Yamane, T. (1967). Statistics: An Introductory Analysis. (2nd ed.), New York: Harper and Row. Yen, Y.S. (2011), The impact of perceived value on continued usage intention in social networking sites. Proceedings of the 2nd International Conference on Networking and Information Technology, 17, 217-223. Yen, C.H., & Lu, H.-P. (2008). Effects of e-service quality on loyalty intention: An empirical study in online auction. Managing Service Quality, 18(2), 127-46. Yi, Y. (1990). A critical review of consumer satisfaction. In V.A. Zeithaml (Ed.) Review of marketing (pp. 68 – 123). Chicago: American Marketing Association. Yildirim, F., & Cengel, Ö. (2012). The perceived risk and value based model of online retailing. Online Academic Journal of Information Technology, 3(9), 7 -21. yStats.com (2012), Africa internet & b2c e-commerce report 2012, Hamburg: yStats.com Gmbh. & Co. KG. Yu, J., Ha, I., Choi, M., & Rho, J. (2005). Extending the TAM for e-commerce. Information & Management, 42 (77), 965-76. Zeithaml, V.A. (1988). Consumer perceptions of price, quality and value: A means-end model and synthesis of evidence. Journal of Marketing, 52, 2–22. Zeithaml, V.A., Bitner, M.R., & Gremler, D.D. (2006). Services marketing: Integrating customer focus across the firm. New York: McGraw-Hill Irwin. Zhang, L., Tan, W., Xu, Y., & Tan, G. (2012), Dimensions of consumers‘ perceived risk and their influences on online consumers‘ purchasing behavior, Communications in Information Science and Management Engineering, 2 (7), 8-14. Zickmund, W.G., & Babin, B.J. (2010). Exploring marketing research (9th Ed). Mason, OH.: Thomson Higher Education:

142

APPENDICES APPENDIX 1: SUPPLEMENTARY STATISTICAL ANALYSES (SPSS OUTPUT RESULTS) Table A.1: Reliability Output for Usefulness (4 items: B1 – B4) Cronbach's

Cronbach's Alpha Based on

Alpha

Standardized Items

N of Items

.954

.958

4

Source: Survey data (2013)

Table A.2: Reliability Output for Compatibility (4 items: B5 – B8) Cronbach's

Cronbach's Alpha Based on

Alpha

Standardized Items

N of Items

.954

.955

4

Source: Survey data (2013)

Table A.3: Reliability Output Ease-of-Use (6 items: B9 – B14) Cronbach's

Cronbach's Alpha Based on

Alpha

Standardized Items

N of Items

.950

.951

6

Source: Survey data (2013) Table A.4: Reliability Output Financial Risk (3 items: B15 – B17) Cronbach's

Cronbach's Alpha Based on

Alpha

Standardized Items

N of Items

.801

.802

3

Source: Survey data (2013) Table A.5: Reliability Output Performance Risk (5 items: B18 – B22) Reliability Statistics Cronbach's

Cronbach's Alpha Based on

Alpha

Standardized Items .702

N of Items .702

Source: Survey data (2013)

143

5

Table A.6: Reliability Output Personal Risk (3 items: B23 – B25) Cronbach's

Cronbach's Alpha Based on

Alpha

Standardized Items

N of Items

.885

.893

3

Source: Survey data (2013) Table A.7: Reliability Output Monetary Value (4 items: B26 – B29) Cronbach's

Cronbach's Alpha Based on

Alpha

Standardized Items

N of Items

.839

.839

4

Source: Survey data (2013) Table A.8: Reliability Output Convenience Value (5 items: B30 – B34) Cronbach's

Cronbach's Alpha Based on

Alpha

Standardized Items

N of Items

.954

.955

5

Source: Survey data (2013)

Table A.9: Reliability Output Social Value (5 items: B35 – B39) Cronbach's

Cronbach's Alpha Based on

Alpha

Standardized Items

N of Items

.538

.829

5

Source: Survey data (2013)

Table A.10: Reliability Output Emotional Value (7 items: B40 – B46) Cronbach's

Cronbach's Alpha Based on

Alpha

Standardized Items

N of Items

.975

.975

7

Source: Survey data (2013)

144

Table A.11: Reliability Output Level of Satisfaction (5 items: C1 – C5) Cronbach's

Cronbach's Alpha Based on

Alpha

Standardized Items

N of Items

.941

.943

5

Source: Survey data (2013)

Table A.12: Correlation Matrix of the three Predictor Variables

Step 1

Constant

PATT

PRSK

PVAL

Constant

1.000

-.906

-.001

-.742

PATT

-.873

1.000

-.259

.494

PRSK

.178

-.519

1.000

-.134

PVAL

-.767

.491

-.189

1.000

Source: Survey data (2014) Table A.13: Collinearity Output for Tolerance and VIF Coefficients Model

a

Collinearity Statistics

1

Tolerance

VIF

PATT

.261

3.826

PRSK

.468

2.135

PVAL

.348

2.877

a. Dependent Variable: CSAT

Source: Survey data (2014)

Table A.14: Hosmer and Lemeshow Test for the Direct Effects Model Hosmer and Lemeshow Test Step

Chi-square

df

Sig.

1

.097

8

1.000

Source: Survey data (2014)

145

Table A.15: Test of Normality for Predictor Variables

Tests of Normality a

Kolmogorov-Smirnov Statistic

Shapiro-Wilk

df

Sig.

Statistic

df

Sig.

PATT

.110

240

.000

.927

240

.000

PRSK

.100

240

.000

.964

240

.000

PVAL

.091

240

.000

.957

240

.000

240

*

.986

240

.017

CPER

.043

.200

a. Lilliefors Significance Correction *. This is a lower bound of the true significance.

Source: Survey data (2014)

Table A.16: Dependent variable Encoding for Direct Effects Model Dependent Variable Encoding Original Value

Internal Value

0

0

1

1

Source: Survey data (2014)

Table A.17: Model Summary for Logistic Regression for Direct Effects Model Model Summary Step 1

-2 Log likelihood 17.617

a

Cox & Snell R Square

Nagelkerke R Square

.729

.974

a. Estimation terminated at iteration number 11 because parameter estimates changed by less than .001.

Source: Survey data (2014)

146

Table A.18: Classification Table for the Direct Effects Model Classification Table

a

Predicted USAGE Observed Step 1

USAGE

Percentage

0

1

Correct

0

110

1

99.1

1

2

127

98.4

Overall Percentage

98.8

a. The cut value is .500

Source: Survey data (2014)

Table A.19: Case processing summary for Direct Effects Logistic Regression Case Processing Summary Unweighted Cases

a

Included in Analysis Selected Cases

N

Percent

240

98.8

3

1.2

243

100.0

0

.0

243

100.0

Missing Cases Total

Unselected Cases Total a. If weight is in effect, see classification table for the total number of cases.

Source: Survey data (2014)

Table A.20: Logistic Regression Results for the Direct Effects Model Variables in the Equation B

Step 1

a

S.E.

Wald

df

Sig.

Exp(B)

PATT

6.006

2.069

8.429

1

.004

405.759

PRSK

-1.834

.854

4.618

1

.032

.160

PVAL

2.329

1.084

4.621

1

.032

10.269

-27.845

10.471

7.071

1

.008

.000

Constant

a. Variable(s) entered on step 1: PATT, PRSK, PVAL.

Source: Survey data (2014)

147

Table A.21: Model summary for Linear Regression Results for the Relationship between Customers’ Perceptions and Customer Satisfaction

Std. Error of the Model

R 1

.728

R Square

Adjusted R Square

Estimate

.531

.529

.74603

a

a. Predictors: (Constant), CPER Source: Survey data (2014)

Table A.22: ANOVA (F-Test) for Linear Regression Results of the Relationship between Customers’ Perceptions and Customer Satisfaction a

ANOVA Model

1

Sum of Squares

df

Mean Square

Regression

149.726

1

149.726

Residual

132.460

238

.557

Total

282.186

239

F

Sig.

269.024

.000

b

a. Dependent Variable: CSAT b. Predictors: (Constant), CPER

Table A.23: Linear Regression Results of the Relationship between Customers’ Perceptions and Customer Satisfaction Coefficients Model

Unstandardized Coefficients

a

Standardized

t

Sig.

Coefficients B (Constant)

Std. Error -2.193

.330

1.258

.077

Beta -6.642

.000

16.402

.000

1 CPER a. Dependent Variable: CSAT Source: Survey data (2014)

148

.728

Table A.24: Dependent Variable Encoding for Logistic Regression of the Relationship between Customer Satisfaction and Usage

Dependent Variable Encoding Original Value

Internal Value

0

0

1

1

Source: Survey Data (2014)

Table A.25: Classification table: Logistic Regression of the Relationship between Customer Satisfaction and Usage Classification Table

a

Predicted USAGE Observed Step 1

USAGE

Percentage

0

1

Correct

0

106

5

95.5

1

7

122

94.6

Overall Percentage

95.0

a. The cut value is .500 Source: Survey data (2014)

Table A.26: Case Processing Summary for Logistic Regression of the Relationship between Customer Satisfaction and Usage Case Processing Summary Unweighted Cases

a

Included in Analysis Selected Cases

Missing Cases Total

Unselected Cases Total a. If weight is in effect, see classification table for the total number of cases. Source: Survey data (2014)

149

N

Percent

240

98.8

3

1.2

243

100.0

0

.0

243

100.0

Table A.27: Model Summary for Logistic Regression of Relationship between Customer Satisfaction and Usage of Online Retailing Services Model Summary Step

-2 Log likelihood

1

Cox & Snell R Square

76.227

a

Nagelkerke R Square .655

.874

a. Estimation terminated at iteration number 8 because parameter estimates changed by less than .001. Source: Survey data (2014)

Table A.28: Logistic Regression Results for Relationship between Customer Satisfaction and Usage of Online Retailing Services Variables in the Equation B Step 1

a

CSAT Constant

S.E.

Wald

df

Sig.

Exp(B)

5.171

.797

42.056

1

.000

176.063

-16.593

2.626

39.925

1

.000

.000

a. Variable(s) entered on step 1: CSAT. Source: Survey data (2014)

Table A.29: Case Processing Summary for Moderated Effects Model Case Processing Summary Unweighted Cases

a

N Included in Analysis

Selected Cases

Missing Cases Total

Unselected Cases Total

Percent 240

98.8

3

1.2

243

100.0

0

.0

243

100.0

a. If weight is in effect, see classification table for the total number of cases. Source: Survey Data (2014)

Table A.30: Dependent variable encoding for moderation effects model Dependent Variable Encoding Original Value

Internal Value

0

0

1

1

Source: Survey Data (2014)

150

Table A.31: Moderation Effects Model Summary Model Summary Step

-2 Log likelihood

1

Cox & Snell R Square

183.125

a

Nagelkerke R Square .461

.616

a. Estimation terminated at iteration number 6 because parameter estimates changed by less than .001. Source: Survey Data (2014)

Table A.32: Classification table for moderation effects regression Classification Table

a

Predicted USAGE Observed Step 1

USAGE

0

1

Percentage Correct

0

85

26

76.6

1

19

110

85.3

Overall Percentage

81.3

a. The cut value is .500 Source: Survey data (2014)

Table A.33: Logistic Regression Results for Moderation Effects Model Variables in the Equation B

Step 1

a

S.E.

Wald

df

Sig.

Exp(B)

CPER

4.587

1.698

7.302

1

.007

98.235

DEMF

.534

2.187

.060

1

.807

1.707

-.196

.523

.141

1

.708

.822

-18.356

7.070

6.742

1

.009

.000

CPERDEMF Constant

a. Variable(s) entered on step 1: CPER, DEMF, CPERDEMF. Source: Survey data (2014)

151

APPENDIX 2: LIST OF ONLINE RETAILING FIRMS IN NAIROBI, KENYA

NO. NAME 1. Area254.co.ke 2. BuyRentKenya.com 3. Buyandsell.co.ke 4. Cars.co.ke 5. Cheki 6. DailyShark.co.ke 7. Digitalduka 8. DukaWala.com 9. Kenyacarbazaar.com 10. Kilakitu.co.ke 11. Maduqa.com 12. MamaMikes 13. Mimi.co.ke 14. Mottiz.com 15. Mzoori.com 16. N-Soko 17. OLX (formerly Dealfish) 18. PataUza 19. PigiaMe 20. Rupu.com 21. Ravenzo 22. Sokobay.com 23. UzaNunua 24. Jumia 25. Zetu Source: Kenya ICT Board (2012); Postel Directory (2012), others.

152

APPENDIX 3: DATA COLLECTION INSTRUMENTS

APPENDIX 3A: COVER LETTER

Peter Misiani Mwencha P.O. Box 53553-00200 Nairobi, Kenya.

Dear Respondent,

RE : PARTICIPATION IN ACADEMIC SURVEY/INTERVIEW: I am a student pursuing a Doctor of Philosophy degree in Business Administration at Kenyatta University, Nairobi Campus. Part of the requirements of this course is to submit a research thesis on a topic of interest to me. To this end, I am currently conducting a study on the Effect of Customer Perceptions on the Usage of Online Retailing Services in Nairobi County, Kenya. I will therefore be most grateful if you could take time off your busy schedule to respond to the questions in the attached questionnaire or to make yourself available for a brief interview so as to enable me carry out this research. This is only an academic exercise and you are assured of anonymity and confidentiality.

Thank you in advance for your cooperation.

Kind regards,

Peter Misiani Mwencha Admission Number: D86/CTY/21719/2010

153

APPENDIX 3B: QUESTIONNAIRE

KENYATTA UNIVERSITY School of Business

Effects of Customer Perceptions on the Usage of Online Retailing Services in Nairobi County, Kenya

Dear Sir/Madam The purpose of this questionnaire is to collect information on the effects of customer perceptions on the usage of electronic commerce services in online retailing firms in Nairobi County, Kenya as part of a study for the award of PhD at Kenyatta University. I will be most grateful if you could take time off your busy schedule to respond to the questions. This is only an academic exercise and you are assured of anonymity and confidentiality. Thank you.

SECTION A. CUSTOMER DEMOGRAPHIC FACTORS First things first: Tell us a bit about yourself. Please respond to each item by choosing the response that best describes you.

1.

2.

Age:  18- 23

 24 - 29

 30 – 35

 36 - 41

 42 - 47

 48 years & above

Highest Level of Education:  High School Certificate  Masters Degree  Other

3.

 Diploma  Doctorate

Monthly income (gross):  Below KSh 24,999  KSh 50,000-74,999  Ksh 100,000-KSh 124,999

154

 Bachelor‘s Degree  Professional

 KSh 25,000-49,999  KSh 75,000-99,999  Ksh 125,000 & above

SECTION B. CUSTOMER PERCEPTION MEASURES Please indicate the extent to which you disagree or agree with each of the following statements by marking with a cross (X) in the appropriate block provided. Please use the following seven-point rating scale ranging from 1 = “strongly disagree” to 7 = “strongly agree”.

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

CUSTOMER PERCEPTIONS Variable Label Perceived Attributes The system enables me to accomplish what I want more quickly The system makes me more effective The system makes it easier to do what I want I find the system useful The e-commerce service fits my image well. Using the system is compatible with all aspects of my lifestyle. I think that using the system fits well with the way I like to do things. Using the system fits into my lifestyle. I find the system to be clear and understandable. It‘s easy to get the system to do what I want it to do It‘s easy to find what is being sought The system has no hassles Learning to operate the system is easy for me. Overall, I believe that the system is easy to use.

23. 24. 25.

Perceived Risk This service costs more than conventional methods I might be overcharged for using this service I might not receive the product/service that I paid for Inability to touch and feel the item worries me One can't examine the actual product It‘s not easy to get what I want Information takes too long to come up/load The e-commerce service failed to perform to my satisfaction My credit card number may not be secure My personal information may be sold to advertisers My personal information may not be securely kept

26. 27. 28. 29. 30. 31. 32.

Perceived Value This e-commerce service is reasonably priced. This e-commerce service is competitively priced This e-commerce service offers value-for-money Using this e-commerce service is economical I can use this e-commerce service anytime I can use this e-commerce service anyplace This e-commerce service is convenient for me to use

15. 16. 17. 18. 19. 20. 21. 22.

155

1

2

Value Label 3 4 5 6

7

33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46.

I feel that the e-commerce service is convenient for me I value the convenience of using this e-commerce service This service would help me feel acceptable by others This service would improve the way I am perceived Using this service would make a good impression on others My friends and relatives think more highly of me for using this service. This service would give its user social approval I enjoy using the system. Some aspects of the system make me want to use it I feel relaxed about using the system Using the system makes me feel good Using the system gives me pleasure Using the system is fun It‘s exciting to use the e-commerce service

SECTION C. CUSTOMER SATISFACTION MEASURES Please indicate the extent to which you are satisfied with the e-commerce system by marking with a cross (X) on one of the five blocks provided below the position which most closely reflects your satisfaction with the service.

1.

How satisfied were you with the online retailing service initially?  Very Dissatisfied  Somewhat Satisfied

2.

3.

4.

 Slightly Dissatisfied  Very Satisfied

 Neither

To what extent does this online retailer meet your needs?  Extremely well  Pleased  Mixed  Extremely poorly



Satisfied

My experience with this online retailer was very satisfactory  Strong Yes  Yes  No  Strong No



Neutral

Overall, I am _ with the service?  Delighted  Mixed

 

Pleased Mostly Dissatisfied

 Satisfied

5. If I could do it all over again, I would still use this service?  

Strong Yes No

 

Yes Strong No

Thank you very much for your time.

156



Neutral

APPENDIX 3C: KEY INFORMANT INTERVIEW GUIDE

KENYATTA UNIVERSITY School of Business

Effects of Customer Perceptions on the Usage of Online Retailing Services in Nairobi County, Kenya Dear respondent, The purpose of this interview is to collect information from key informants on the effects of customer perceptions on the usage of online retailing services in Nairobi County, Kenya as part of a study for the award of PhD at Kenyatta University. I will be most grateful if you could take time off your busy schedule to make yourself available briefly to answer the questions in this interview guide. This is only an academic exercise and you are assured of anonymity and confidentiality. Thank you. Yours sincerely, Peter M. Mwencha D86/CTY/21719/2010 School of Business Kenyatta University

Working Guide/Manual on Qualitative Information Kindly answer the following questions as truthfully and comprehensively as possible.

1. Have you used any local online retailing website(s) before? [Yes]) [No] a. If yes, give examples and explain what you used the website(s) for? __________________ _________________________________________________________________________ _________________________________________________________________________ _________________________________________________________________________ b. If yes, what factors attracted you to use the online retailing website(s)? _________________________________________________________________________ _________________________________________________________________________ ________________________________________________________________________

157

c. If yes, have you used the website(s) within the last three months? [Yes] [No]

d. If no, explain why not? ______________________________________________________ _________________________________________________________________________ _________________________________________________________________________ _________________________________________________________________________ 2. Are there any products you wouldn‘t buy online? Give examples and please explain why? ________________________________________________________________________ _________________________________________________________________________ _________________________________________________________________________

3. What

issues/concerns

do

you

have

regarding

online

retailing

services?

_________________________________________________________________________ _________________________________________________________________________ _________________________________________________________________________ 4. Which payment method(s) would you prefer (or use) when shopping online? Explain why? _________________________________________________________________________ _________________________________________________________________________ _________________________________________________________________________ _________________________________________________________________________ 5. How satisfied (or dissatisfied) are you with local online retailing services? Give reasons why. _________________________________________________________________________ _________________________________________________________________________ _________________________________________________________________________

6. Please comment on the following aspects regarding online retailing services in Kenya: a. Usefulness:____________________________________________________________ ______________________________________________________________________ b. Convenience:___________________________________________________________ ______________________________________________________________________ c. Ease of Use: _________________________________________________________ ______________________________________________________________________

158

d. Financial Risk: _________________________________________________________ ______________________________________________________________________ e. Performance Risk: ______________________________________________________ ______________________________________________________________________

f.

Privacy Risk: __________________________________________________________ ______________________________________________________________________

g. Monetary Value: _______________________________________________________ ______________________________________________________________________ h. Convenience Value: _____________________________________________________ ______________________________________________________________________ i.

Social Value: __________________________________________________________ ______________________________________________________________________

7. What problems/difficulties/challenges do people encounter when using these websites? _________________________________________________________________________ _________________________________________________________________________ _________________________________________________________________________ _________________________________________________________________________ 8. Give practical suggestions and policy recommendations that would enhance the usage of online retailing services in Kenya? _________________________________________________________________________ _________________________________________________________________________ _________________________________________________________________________ _________________________________________________________________________

9. What is your overall assessment of online shopping in the Kenyan context? _________________________________________________________________________ _________________________________________________________________________ _________________________________________________________________________ _________________________________________________________________________

Thank you very much for your time and attention.

159

APPENDIX 4: RESEARCH AUTHORISATION APPENDIX 4A: CLEARANCE LETTER

160

APPENDIX 4B: RESEARCH PERMIT

161

APPENDIX 5: CODEBOOK APPENDIX 5A: CODE BOOK FOR QUANTITATIVE DATA ANALYSIS Variable

SPSS Variable Name

1.

Usage

USAGE

2.

Perceived Attributes

PATT

3.

Perceived Risk

PRSK

4.

Perceived Value

PVAL

5.

Customer Perceptions

CPER

6.

Customer Satisfaction

CSAT

7.

Demographic Factors

DEMF

Interaction Term between Customer Perception 8.

CPERDEMF and Demographic Factors

Source: Survey data (2013)

162

APPENDIX 5B: CODE BOOK FOR QUALITATIVE DATA ANALYSIS Code/Variable 1. Usage

2.

Perceived Attributes

3.

Perceived Risk

4.

Perceived Value

5. Customer Perceptions

6. Customer Satisfaction

7. Demographic Factors

8. Industry Prospects 9. 10.

Challenges/Problems Policy Recommendations

Operational Definition The utilization of one or more features of an online retailing service by registered users within a certain timeframe. It could be browsing or actual purchase. Users‘ perceptions regarding the online retailing service‘s functional features, properties and qualities. The transaction-related risks that consumers face as a result of using online retailing services. The consumer‘s evaluation of the benefits of online retailing usage. The subjective opinions/beliefs/judgments of an individual regarding an online retailing service based on prior use experience. The customer‘s overall positive evaluation of the online retailing service following initial usage or based on all prior interactions/encounters and experiences with the online retailing service. Are any personal characteristics/attributes of consumers that tend to remain static throughout an individual‘s life time, or evolve slowly over time. This includes age, gender, race, education, income, lifestyle, etc. Forecasts/predictions regarding future developments of the online retailing industry. Could be poor or bright. A hindrance/barrier to online retailing industry growth or development Written policy advice prepared for decision makers regarding a specific issue.

Source: Survey data (2014)

163

APPENDIX 5C: SUMMARY OF MAJOR THEMES Theme

1. Usage Diversity

2. Prevailing Attitudes

Description/Operational Definition

Types/nature and extent of usage/utilization of online retailing services

Opinions, thoughts and feelings regarding online retailing services

Determinants of online retailing usage

3. Usage Drivers

164

Examples of Significant Statements I used Jumia to look at the price range of a laptop that would fit my budget as well as look at the specifications of that laptop. I was looking for products I wanted to purchase [but] I was also selling some goods on OLX. I used Mzoori.com, Rupu.com, Jumia and OLX to look for products for purchase and was also selling some goods on OLX. I only use online retailing stores when I am seeking to buy high cost items because I get an opportunity to compare prices across different e-commerce stores. I do not do this frequently that‘s why I cannot do this (use the websites) every now and then… The inconsistency in product availability coupled with the lack of accurate information on some websites has made online users form a negative impression towards online retailing services. Quite a number of online retailers cannot guarantee proper inventory management – you order an item online but you cannot have it (delivered) because it is out of stock. I‘m very dissatisfied. I have not been able to purchase anything because what‘s available is not within my budget range. I‘m somewhat satisfied…however, it can be quite costly to roll out. I‘m very satisfied with online retailing in Kenya because of how fast it is growing and the convenience it brings. Online retailing is quite useful at it helps one purchase products after viewing a variety without having to waste time walking around. It‘s very convenient as it allows one to purchase by just clicking and having the product delivered at one‘s doorstep. Payment methods such as cash and m-pesa. They are both easy to

pay on delivery as there‘s no risk of losing money and they are fast when transacting. There is a chance of personal details getting in the wrong hands, especially for those using visa cards to make payments. Moreover, there are chances that delivery does not occur as promised e.g. delivery within 24 hours turns to delivery after 72 hours.

4. Market Development

5. Market Prospects

Possible ways of increasing and sustaining the usage of online retailing services

Economic/business potential of the online retailing sub-sector (industry category)

Source: Survey data (2014)

165

Increase the product range and ensure that products are always in stock; delivery to homes would increase the convenience of this value chain; ensure that the regulatory frameworks for protecting customers privacy are enhanced and there is judicial recourse if these are compromised. Build trust; avail a variety of products; offer good prices for products; avail all information on one page; offer return policies; offer cash on delivery and offer warranties. There‘s need for more regulation from government to reduce the risks associated with online shopping Enact data privacy laws to protect consumers‘ sensitive data. Having better user interface, user experience and e-security. Online shopping as a trend is picking up and has a bright future. This is because more internet users are turning to online shopping especially due to its convenience. Online retailing is gaining momentum; while we have already achieved good progress, a lot more needs to be done to ensure security while transacting, product variety and value added services. Online shopping is an industry growing in Kenya and people are beginning to embrace it. A few things need to be worked on and it will turn into a lucrative industry. Its heading in the right direction but we can do better. It‘s still at a pretty early stage. There a few outstanding efforts. However, there are many copy cats.

APPENDIX 6: SUMMARY OF EMPIRICAL REVIEW AND RESEARCH GAPS Variable

Thematic Area

Perceived Attributes

Author

Study

Key Findings

Weaknesses

Knowledge Gaps

Parthasarathy & Bhattacherjee (1998)

Understanding Postadoption Behavior in the Context of Online Services Factors Affecting Continued Usage of Internet Banking Among Egyptian Customers

Perceived attributes such as usefulness and compatibility determine continued usage behavior. Perceived ease-of-use is the strongest predictor of CUI. Perceived risk had no relationship with CUI.

Employed IDT as the sole basis to study continued usage of online services.

There‘s need to include more theories to study consumer usage behavior of e-commerce services. There‘s need to employ actual usage as the DV in future studies instead of continuance intentions.

Perceived risks as barriers to internet and ecommerce usage

Internet credit card stealing and supplying personal information (Privacy risk) affects both current and future e-commerce usage. Perceived privacy risk, perceived social risk and perceived economic risk don‘t influence online shopping behavior Perceived value is an immediate antecedent to customer satisfaction and repurchases intention. Information value, social value, and hedonic value have a positive effect on CUI.

Only privacy risk was established as having an influence on e-commerce usage.

El-Kasheir, Ashour & Yacout (2009)

Liebermann & Stashevsky (2002) Perceived Risk Zhang, Tan, Xu and Tan (2012)

Customer Perceptions

Oh (1999) Perceived Value Yen (2011)

166

Dimensions of Consumers‘ Perceived Risk and Their Influences on Online Consumers‘ Purchasing Behavior Service Quality, Customer Satisfaction and Customer Value: A Holistic Perspective The Impact of Perceived Value on Continued usage Intention in Social Networking Sites

Used mall interception to collect data; Employed CUI as DV instead of actual usage.

The study focused on the online shopping segment of e-commerce in the preadoption context. This study used single-item overall measurement for most variables. Only considered ―get‖ components (social value and hedonic value) ignoring ―give‖ components.

Further studied on the influence of perceived risk dimensions on the usage of e-commerce services are needed. There‘s need to carry out studies focused on other segments of e-commerce in the post-adoption context. Each model construct or variable should be measured with multiple items There‘s need to integrate both give and get components of perceived value.

Bolton and Lemon (1999)

DeLone & McLean (2004)

Customer Satisfaction

Level

of

Satisfaction

Chen, Huang, Hsu, Tseng & Lee (2010)

Yen (2011)

Age

Customer

Income

Factors Education Level

Venkatesh, Morris, Davis & Davis (2003) Hernández, Jiménez & Martin (2011) Riddel & Song (2012)

167

A Dynamic Model Of Customers‘ Usage of Services: Usage as an Antecedent and Consequence of Satisfaction Measuring E-Commerce Success: Applying the DeLone & McLean Information Systems Success Model Confirmation of Expectations and Satisfaction with Internet Shopping: The Role of Internet Self-efficacy The Impact of Perceived Value on Continued usage Intention in Social Networking Sites User Acceptance of Information Technology: Towards a Unified View

Age, gender and income: do they really moderate online shopping behaviour? Role of Education in Technology Use and Adoption: Evidence from Canadian Workplace and Employee Survey

High level of cumulative satisfaction with initial usage will lead to higher usage levels of the service in subsequent periods. Proposed that ease of use influences user satisfaction which subsequently directly influences usage of e-commerce services. Satisfaction is influenced by perceived usefulness and both satisfaction and perceived usefulness determine consumer‘s repurchase intention. End-user satisfaction mediates the relationship between PV and CUI

Age has a moderating influence on the relationship between facilitating conditions and use of IT Age and income have no moderating influence on use of e-commerce Education has an influence on computer usage but not on usage of computercontrolled or computerassisted devices.

No moderating variables used in the study.

There‘s need to establish the moderating effect of customer characteristics on the perception – usage relationship.

Conceptually enhanced the DeLone & McLean IS Success Model (2003) without empirical tests.

There‘s need to empirically validate the relationship proposed in the model.

Ignored the role of perceived value and risk as determinants of ecommerce repurchase intentions.

There‘s need to investigate the role of value and risk perceptions as determinants of usage of e-commerce services in future studies

Only considered ―get‖ components e.g. social value and hedonic value. ignoring ―give‖ components. Was based on an organizational context where usage is not entirely voluntary.

There‘s need to integrate both give and get components of perceived value.

Only establish the moderating effect of age and income but not education Was based on workplace/ organizational context where usage is not entirely voluntary.

There‘s need to establish the moderating influence of education level

There‘s need to establish the moderating influence of age in a voluntary context.

There‘s need to establish the moderating influence of education in a voluntary (i.e. consumer) context.

Continued

Usage

Usage

Bhattacherjee (2001b)

An empirical analysis of the antecedents of electronic commerce service continuance

Perceived usefulness and user satisfaction influence e-commerce continuance intention.

Is based on the ECT. Also lacked moderating variables. Focused on B2B E-Commerce. Used CUI as the DV

There‘s need for an integrated model of ECommerce Usage by Consumers.

Brown & Jayakordi (2008)

B2C e-Commerce Success: a Test and Validation of a Revised Conceptual Model

CUI of an online retail site is directly influenced by perceived usefulness, user satisfaction and system quality. In the post-purchase stage, PU play a more important role than hedonic factors in predicting customer online repurchase intention. Empirically tested and validated that satisfaction is a strong predictor of the continuance intentions of consumers. user satisfaction is positively related to usage continuance

Is limited to the B2C ECommerce context in South Africa. Lacks moderating variables.

Employed the user‘s continuance intention as the DV. In this study, actual usage is employed as the DV. There‘s need for more studies on customer ecommerce usage that has a wider respondent make-up outside of the U.S. This study shall employ actual use as the DV.

Wen, Prybutok and Xu (2011)

Chen & Chou (2012)

Ramayah Lee (2012)

Source: Survey data (2013)

168

&

An Integrated Model for Customer Online Repurchase Intention.

Exploring the continuance intentions of consumers for B2C online shopping: Perspectives of fairness and trust System Characteristics Satisfaction and ELearning Usage: A Structural Equation Model.

Respondents were college students in the U.S., thus limiting generalizability to other populations & contexts. Employed continuance intentions as the DV for ecommerce success in the B2C E-Commerce context. Limited itself by using DeLone & McLean‘s Model (2003) as the sole basis for the study

There‘s need to develop an integrated model for use in explaining IS continuance.