Mount Pleasant, MI 48859. M. RONALD BUCKLEY. University of Oklahoma. GENEVIEVE JOHNSON. American Institutes for Research. JENSEN MECCA.
10.5465/AMBPP.2016.280
EXAMINING MOTIVES IN ORGANIZATIONAL CITIZENSHIP BEHAVIOR: INTRODUCING THE OCB-INTENTIONALITY SCALE ALEXANDRA E. MacDOUGALL College of Business Administration Central Michigan University Mount Pleasant, MI 48859 M. RONALD BUCKLEY University of Oklahoma GENEVIEVE JOHNSON American Institutes for Research JENSEN MECCA SHAKER ABSTRACT This paper details the development of the OCB-Intentionality Scale (OCB-IS). Study 1 reports on scale construction and refinement, scale dimensionality, and individual difference correlates of nine OCB motives. Study 2 examines situational correlates and assesses the predictive utility of the OCB-IS. Results provide preliminary content, construct, and criterionrelated validation evidence. INTRODUCTION Organizations are increasingly searching for employees who go above and beyond their formal, compulsory duties by engaging in discretionary behaviors that benefit others or the organization. This type of behavior, termed organizational citizenship behavior (OCB), has long been touted for contributing to organizational objectives and facilitating organizational functioning (Podsakoff & MacKenzie, 1997). For example, OCB has been shown to reduce labor costs and positively impact service quality, customer service ratings, and profitability (Yen & Niehoff, 2004). Relatedly, citizenship behavior has been linked to increased performance quality, performance quantity, and financial efficiency (Podsakoff, MacKenzie, Paine, & Bachrach, 2000). These examples in mind, it is no wonder that organizational citizens are typically valued as prized assets by organizations. Nevertheless, scholars have begun to question whether our understanding of OCB is onesided. There appear to be a number of boundary conditions on citizenship behavior with meaningful impacts on corresponding outcomes for both the actor as well as the organization. Recent work suggests that employee motives are one such construct that may influence the nature and outcomes of OCB. Whereas it was once believed that citizenship behavior stems from purely altruistic or other-serving motives (Organ, 1988), an emerging school of thought has countered this belief, proffering that citizenship may result from functional, neutral, or even darker motives (e.g., Bolino, 1999; Rioux & Penner, 2001; Bolino, Turnley, & Niehoff, 2004; Bolino, Klotz, Turnley, & Harvey, 2013). Initial strides in understanding these motives has
10.5465/AMBPP.2016.280
proven fruitful, yet there still lacks a comprehensive framework from which to study and understand the motivational origins of OCB. The present effort describes the development and initial validation of the OCBIntentionality Scale (OCB-IS), a measure intended to capture nine discrete motivational precursors to citizenship behavior. Across two studies, we describe the development and continued refinement of the OCB-IS. Further, we establish preliminary validation evidence through the formation of a nomological network surrounding OCB motives and explore the predictive utility of the scale with respect to OCB dimensions and employee-relevant outcomes. INSTRUMENT DEVELOPMENT Development of the OCB-Intentionality Scale followed the construct-based approach (Mumford et al., 1996; 2007). A review of the literature yielded theoretical classification of nine motives, including (1) prosocial values, (2) organizational concerns, (3) image enhancement and maintenance, (4) atonement, (5) obligation, (6) functionality, (7) task avoidance, (8) personal discontent, and (9) social interests. Given this a priori classification, a deductive approach was adopted to contribute to content validation (Hinkin, 1995; 1998). Items from two preexisting scales were used as a base for several of the motive dimensions (Allen & Rush, 1998; Rioux & Penner, 2001). Additional item generation was twofold, including an item-writing team along with an item-review team. Panel members were six senior doctoral students (three per team) chosen for their expertise in item development and review. While panel members were all familiar with OCB generally, they received operational definitions of the nine constructs under consideration for the OCB-IS and discussed each until mutual understanding was reached. Next, the item-writing team proceeded to item generation. Each item writer independently generated items over multiple item writing sessions, yielding a total of 297 items. This preliminary list was reviewed to flag unclear, irrelevant, or otherwise problematic items. This reduced the item pool to 285 items. Next, the item-review team content analyzed each of the 285 OCB-IS items by independently categorizing each item into one of the nine subscales, and rating items for relevance and clarity using a 5point Likert scale. Only items with 100% agreement for categorizations were retained, resulting in 228 items that were retained for use in initial analyses. STUDY 1 Study 1 sought to assess the factor structure and psychometric properties of the OCB-IS, reduce scale length, and provide preliminary construct-related validation evidence. Specifically, individual difference correlates were examined to guide the initial formation of a nomological network surrounding the nine OCB motives. Method Participants comprised 667 employed students drawn from an undergraduate subject pool at a large Midwestern university. Using online survey methodology, participants completed demographic information, the OCB-IS, and a battery of individual difference constructs including personality, emotions, global motives, impression management tactics, and social desirability. Following demographic information, the OCB-IS was always completed first to
10.5465/AMBPP.2016.280
mitigate fatigue effects. The remaining questionnaires were presented to participants in random order. Analyses and Results To examine the factor structure of the OCB-IS, a series of exploratory factor analyses (EFAs) were conducted using maximum likelihood (ML) estimation and a promax (oblique) rotation. Examination of the scree plot along with a priori theoretical expectations led to the decision to retain nine factors. Heeding the recommendation that four to six items are optimal for most constructs (Hinkin, 1998), a number of increasingly stringent criteria were applied to reduce scale length. This process resulted in a nine-factor model with five items per subscale. In the final EFA, the model accounted for 60% of the total variance. The forty-five retained items were found to load above .60 on their respective factors, and cross-loadings did not exceed .15. Further examination of psychometric properties indicated that average inter-item correlations for subscales ranged from .58 to .81 and coefficient alphas ranged from .80 to .93. Finally, interfactor correlations were analyzed to produce evidence of discriminant validity. Correlations among the nine factors ranged from .01 to .55, suggesting that while some motives are related to each other, they are conceptually distinct (e.g., Bedeian, 2014). Next, correlations were analyzed among the nine motives and the aforementioned individual difference traits. These results were largely in line with expectations, providing initial evidence for convergent validity of the OCB-IS. As an example, prosocial value motives exhibited strong negative relationships with Machiavellianism and egoism, but strong positive relationships with agreeableness, conscientiousness, achievement, and affiliation. Conversely, image enhancement motives were found to have the strongest relationships with power, achievement, self-promotion, and ingratiation. Also included in the correlational analyses were gender, age, and social desirability. Results demonstrated that gender, age, and social desirability differentially relate to various dimensions of the OCB-IS, speaking to the importance of controlling for these variables in future efforts. Discussion Study 1 subjected the newly-developed OCB-Intentionality Scale to EFA and scale refinement, thereby reducing the scale from 228 items to 45 items. Results from the EFA provided support for the nine-factor scale structure, indicating that each of the nine theorized motives are distinct. Examination of the correlations among the motives and various individual difference attributes suggest that the motives exhibit unique relationships with personality traits, emotions, global motives, and impression management tactics. STUDY 2 The purpose of Study 2 was threefold. First, to confirm the scale factor structure established in Study 1. Second, to expand the nomological network involving OCB motives through inclusion of situational correlates. Third, to explore the predictive utility by assessing how certain motives differentially predict OCB dimensions as well as citizenship outcomes for the actor.
10.5465/AMBPP.2016.280
Method Participants were 233 employed students recruited in an undergraduate business class. Online survey methodology was adopted to collect participant responses at two points in time. At the first time point, participants provided basic demographic information and then completed the refined 45-item OCB-IS along with measures of social desirability, perceptions of the work environment (i.e., organizational support and organizational politics), and organizational citizenship behavior. During the second time point, roughly one month later, participants were asked to complete questionnaires assessing job stress and work-family conflict. The lapse in time allowed for a more stringent examination of OCB motives on outcomes of interest. Analyses and Results Confirmatory factor analysis. Confirmatory factor analysis (CFA) with maximum likelihood estimation was adopted to examine the extent to which the factor structure of the OCB-IS remained consistent across samples. Three models were tested to compare the fit of our hypothesized nine-factor model with two alternative models. The first model specified nine latent factors, reflecting the nine hypothesized motives, each with five indicators (i.e., scale items). Factor covariances were estimated. The second model specified a reduced three-factor solution derived from examination of inter-factor correlations. Each of the three factors entailed three motives collapsed based on their interrelationships. Again, covariances between factors were estimated. Finally, a unidimensional model was estimated wherein all indicators were captured by one latent variable. Model comparison results revealed that the hypothesized nine-factor solution was the best fitting model. Specifically, fit indices denoted adequate model fit, χ2 (909, N = 233) = 1396.68, p < .001; CFI = .91; TLI = .90; RMSEA = .05 [.043, .053]. Similarly, standardized path loadings all exceeded .50 and were all significant at the p < .0001 level. Taken together, CFA results supported the construct validity of the OCB-IS, given that the factor structure aligned with the constructs it was intended to measure (Floyd & Widaman, 1995). Tests of gender differences. Given that gender arose as a significant correlate of various motives in Study 1, two approaches were adopted to identify gender differences among the nine OCB motives in Study 2. First, a series of independent samples t-tests were conducted, with results revealing differences on four of the nine motives. Females were found to score higher than males on prosocial values, organizational concerns, image enhancement, and social interests. Second, a Multiple Indicator Multiple Cause (MIMIC) approach was applied. MIMIC models provide an avenue to examine the relationships among covariates, latent variables, and indicators simultaneously (Jöreskog & Goldberger, 1975). Gender was added as a covariate by adding paths from gender to each of the nine latent factors (i.e., motives) in the retained CFA model. While the addition of gender to the model did not have a large impact on model fit, χ2 (945) = 1454.95, p < .0001; CFI= .90; TLI=.90; RMSEA=.05 [.043, .053], results largely mirrored findings from the t-tests with the exception of image enhancement motives. Correlational analyses. Interrelationships among the nine OCB motives were explored with respect to perceptions of organizational support and perceptions of politics. Findings largely mirrored our expectations, again proffering evidence of construct validity. Specifically,
10.5465/AMBPP.2016.280
perceptions of politics were found to relate positively to image enhancement, atonement, obligation, functionality, task avoidance, and personal discontent motives, and negatively to prosocial values and organizational concern motives. Conversely, perceptions of organizational support exhibited positive correlations with prosocial values, organizational concern, and social interest motives, and negative correlations with atonement, task avoidance, and personal discontent motives. Multivariate multiple regression. Two multivariate multiple regression analyses were conducted to investigate the impact of OCB motives on outcomes of interest. First, the overall multivariate model exploring the role of motives in predicting various dimensions of OCB (i.e., interpersonal-OCB, organizational-OCB, and individual initiative) was found to be significant, Wilks’ Lambda = .40, F(36, 633) = 6.39, p < .0001. Results indicated that prosocial values and social interests contributed to interpersonally-directed OCB, whereas organizational concern, obligation, and social interest motives contributed to organizationally-directed OCB. Although not expected, atonement was also found to predict OCB-O, albeit negatively. Five of the nine motives (organizational concern, obligation, functionality, personal discontent, and social interests) were found to predict individual initiative. The second overall multivariate model examining the impact of motives on personal outcomes for the OCB actor (i.e., job stress and work-family conflict facets) was also found to be significant, Wilks’ Lambda = .51, F(60, 804) = 2.08, p < .0001. In particular, findings demonstrated that feelings of obligation increased participants’ job stress and work-family conflict, whereas prosocial values and social interest motives reduced scores on dimensions of work-family conflict. Taken together, results from the multivariate multiple regression analyses provide initial criterion-related validation evidence for the OCB-IS with regard to two employeerelevant outcomes. Discussion Study 2 was designed to provide additional validation evidence for the scale developed in Study 1. Using confirmatory techniques, the factor structure established in Study 1 was applied to data collected in an alternate sample, and inspection of model fit indicated that the hypothesized nine-factor solution fit the data well. Further, Study 2 expanded our understanding of the nomological network surrounding OCB motives, particularly with respect to perceptions of the work environment. Lastly, it was established that the motivational precursors to OCB have implications for the type of citizenship behavior an employee may engage in, as well as the personal costs endured as a result. GENERAL DISCUSSION The overarching goal of the present effort was to develop a comprehensive scale to distinguish among none distinct OCB motives. Adopting a construct-based approach, the itemgeneration process was twofold, including an item-writing team and an item-review team. Content analysis by expert judges of the preliminary item pool supported the content adequacy of the scale. Two studies then provided empirical validation evidence for the OCB-Intentionality Scale. Study 1 employed EFA and correlational analyses to explore the factor structure and individual difference correlates of OCB motives, thereby demonstrating the scale’s construct
10.5465/AMBPP.2016.280
validity. Study 2 adopted CFA, correlational analyses, and regression techniques which served to support the construct validity of the scale and to establish preliminary criterion-related validity. Before turning to the broader theoretical and practical advances of this work, several limitations are worth note. First, both Study 1 and 2 comprised undergraduate student samples. While participants were employed, responses may have been biased as compared to that from a full-time, working sample. Second, use of survey methodology, while convenient in nature, risks the potential for common method bias. Third, while the final scale captured the nine hypothesized OCB motives, there may be differing facets within each motive that went unexplored. In this way, multidimensionality of motives may have been lost in this effort. Despite these limitations, there are a number of important implications worth note. Whereas OCB was previously thought to obviate the need for formal mechanisms and rewards (Organ & Konovsky, 1989), the current effort joins other recent research indicating that this may no longer be the case. Rather, some employees strategically engage in OCB to avoid task performance or to gain advantages on the job, whereas others are feeling pressured to do so. This leads to another important implication. To the extent possible, managers should be cognizant of the motivations underlying employees’ OCB. For example, employees constantly engaging in OCB due to perceptions of pressure should be protected, whereas those engaging in OCB for strictly functional or avoidant purposes may benefit from increased oversight. To this end, managers should actively consider their prioritization of task vs. contextual performance in the formation of management policy. CONCLUSION Recent research has called for a more balanced view of organizational citizenship behavior. One avenue to this end involves the elucidation of both bright and dark motivational precursors. The present effort outlines the initial development and validation of the OCBIntentionality Scale, a measurement instrument designed for use in capturing the complex motives that underscore OCB. Results offer preliminary validation evidence for the scale and reveal that motives offer a viable approach to predicting OCB and employee-relevant outcomes. REFERENCES AVAILABLE FROM THE AUTHORS
