A Social Identity Perspective on Political Affiliation

4 downloads 0 Views 256KB Size Report
Jian-hong Zheng. Keywords: Social Identity theory, Political Affiliation, Preference, Identity Salience. Abstract. .... party group at the same strength with party members. Additionally, their .... Michigan, Sheridan Books, 2001. [3] Conover, P. J. The ...
A Social Identity Perspective on Political Affiliation: Implicit and Experimental Evidence Wu-ming He1, Jian-hong Zheng 1,a,* 1

Department of Psychology, Zhanjiang Normal University, Zhanjiang 524048, China a

[email protected]

* Jian-hong Zheng Keywords: Social Identity theory, Political Affiliation, Preference, Identity Salience.

Abstract. Objective: This study investigated the existence of political affiliation as an aspect of social identity and its influence on behavior. Method: Implicit assessment of group preference using IAT and GNAT were conducted and experimental manipulation of identity salience using priming paradigm was introduced. Results: Party members showed in- group preference and out- group derogation. Non-party people’ preference to party group and non-party group did not show difference significantly. Out-group derogation was not found in non-party people. Political affiliation formed a social identity in party members while not in non-party people. Political affiliation as a portion of social identity would affect people ’s specific political evaluation. Conclusions: These findings give more comprehensive insight into our understanding of political affiliation, and provide an example of studying political affiliation from social identity perspective, from both explicit and implicit level, and from an approach of experiment. Introduction Social identity theory. In recent years, social identity has been viewed as the portion of an individual’s self-concept derived from perceived membership in social groups [1, 2],which is multiple and would influence thought and behavior Therefore, social identity theory has been applied to many fields to explain individual’s behaviors in the context of group situation. People’s perception, attitude and behavior are influenced by their social identity as they view themselves different across social groups they belonged to. We here investigate political affiliation as an aspect of social identity, namely political identity, in the context of triggering one’s remember of his/her political membership. It is widely suggested that social identity contains not only objective membership of that group, but also psychological sense to that group, such as psychological attachment or group self- esteem and group commitment [2, 3]. Thus, people with objective party- membership are not necessary to identify their social identity as a group of party members, unless they own a positive attitude to their party group, called in-group preference. We test participants’ political identity under this logic in Study 1 and Study 2, by examining their positivity toward political affiliation. Social identity and priming. Whether a specific identity influences people’s behavior depends on whether it becomes salient in the current situation [4]. This provides the possibility to manipulate social identity by priming. For example, some researchers observed participants’ economic decision making preference across various social identities made salient [5]. Priming as a research method in social psychology is derived from semantic activating system model [6] in cognitive psychology

and was then introduced to implicit social cognition research [7]. There are various methods of priming identity cognition to create a social identity salience, such as imagination [8], or completing background questionnaire [5, 9]. Given that imagination in practice would introduce extraneous variable easily, the Study 3 uses background questionnaire to prime political identity. Social identity and political affiliation. Social identity model has also been applied to politics studies [10]. Correlation between group identification and political perception and evaluation is tested [11]. A recent investigation on college students’ social identity found that student’s political affiliation is more likely to form their social identity [12], which is in line with another study showing that political identity is more accessible [11]. The impact of political affiliation on bank profitability [13] and people’s evaluation to government’s policy has been explored [14]. These studies demonstrate that people with different political affiliation would behave differently. The present study. The previous studies on political affiliation, however, focused only on explicit measurement and were correlational. We fill these gaps by using implicit measurement and investigating causal relationship. Hypothesis 1: People perceive their political affiliation as one of their social identities (Study 1 and 2). Hypothesis 2: Activating political identity would make one’s political evaluation (Study 3). Implicit assessment of political group identity is introduced and causal relationship between political affiliation and political evaluation is investigated. Study 1 Study 1 is to test participants’ attitude toward party group and non-party group. If participants show positivity toward party- membership, they demonstrate a social identity of party affiliation. A novel IAT task and a self-report survey are used as implicit and explicit measure of participants’ attitude, respectively. Participants. Participants were 27 undergraduate students, including 8 party members and 19 non-party people. We would like to note that this seemingly imbalance of number between party and non-party participants is approximate to the real proportion of party members among sampling population. Participants volunteered to participate in the study. Each participant got a small present for reward after the study. Materials. In our IAT task, party group and non-party group were selected as target concepts. Twenty typically used words of positive attribute and twenty widely used words of negative attribute were selected from other researches [15]. College students (N=44) from the population of interest were asked to find out eight positive words and eight negative words that are more typical by rating these words. These eight positive words and eight negative words were used in Study 1 and 2 as attribute concepts. The IAT task is programed by Inquisit software on Windows computers, with stimuli displaying on a 15 inches LCD screen at a refresh rate of 60 Hz. The explicit preference survey contained two seven-point single item scales that were employed to measure participants’ rating on their attitude toward party group and non-party group, respectively. Higher score indicates more positivity toward that group. Procedure. Every participant got an information sheet which informed them that the test is for scientific research and confidential. Then participants individually completed the IAT task, a demographic survey and the explicit preference survey on computer. Result and discussion. The result of implicit and explicit measure of attitude in Study 1 was presented in Table 1. A positive value of D′(D′= incompatible- compatible) indicates a close association of categories in compatible task. A negative value of d′ indicates a close association of categories in incompatible task. The larger is the value of d′, the closer is the association. For our

analyses, the D′of 0.66 indicated greater ease for party members in pairing good with party group and bad with non-party group (compatible task) versus bad with party group and good with party group (incompatible task). As presented in Table 1, party members showed a strong association of party members with positive attribute (D′=0.66, t(7)=4.73, p=0.002). Non-party participant showed a slight association of non-party group with positive attribute (D′=0.35, t(18)=3.53, p=0.002). Both party members and non-party participants demonstrated the in- group preference on implicit level. Additionally, party members showed more preference significantly to party group than non-party participants on explicit level (M1=6.13, M2=4.74; t(25)=2.625, p=0.015). These findings supported the hypothesis 1 that party members identify their party membership as one of their social identities. However, non-party participant’s implicit preference to non-party group was slight and their explicit preference to non-party group was not stronger than party members (M1=6.13, M2=5.79; t(25)=0.994, p=0.330). Non-party people’s social identity of non-party group was vague and weak. Table 1 IAT effect (D′) calculation and explicit measurement for Study 1 Implicit

Explicit Party group

Non-party group

D′

t test

0.66

t=4.73(7);

6.13

(0.393)

p= 0.002

(0.991)

t(25)=2.625,

(0.354)

t(25)=0.994,

Non-party

0.35

t=3.53(18);

4.74

p=0.015

5.79

p=0.330

participant

(0.427)

p=0.002

(1.727)

Party members

M

t test

M

t test

6.13

(0.918)

Study 2 Study 2 was to replicate Study 1 with another implicit measurement method, go/no go association test (GNAT), which allows separable assessment of attitude toward a single object and is a solution to the problem that IAT procedure needs two opposite or complementary objects to run [14]. Participants. Participants were 20 undergraduate students, including 7 party members and 13 non-party people. Each participant got a small present for reward after the study. Materials. We devised a go/no go association test as the implicit measurement of political identity, with the target concepts, attribute words, and computers the same to Study 1, except that we added exemplars of distracter as is a technical requirement of GNAT. Procedure. After informed, participants were first asked to complete the GNAT task and then to complete the demographic survey. Result and discussion. Sensitivity (d′) was calculated by combining data for each pairing, indicating responder’s ability to discriminate targets (signal) from distracters (noise). Greater sensitivity indicates a stronger association between the target category and attribute. One participant’s data was deleted because the values of D′ were equal to zero or less than zero, which means one’s discrimination of target exemplars from distracters is at random level. Sensitivity presented in the Table 2 showed that implicit attitude toward party group between non-party people and party members did not differ significantly. Nor was there significant difference in pair of non-party plus good. However, the pair of non-party plus bad demonstrated a difference, for party members had a greater sensitivity (3.20) than non-party people (2.46) in this dimension (t(17)==2.345; p=0.031). These results were consistent with Study 3 that both non-party people and party members showed preference to party group. The association of non-party group with bad attribute was separated from the incompatible combination of IAT task for party members in Study

1. This further assessment made party members’ attitude clearer that they did not only gain a preference to party group of their own, but also derogation to non-party group. The results of Study 1 and Study 2 confirmed the fact that party members hold an in- group preference and a out- group derogation. This supported the hypothesis that political identity exists in party members. On the contrary, non-party people did not showed derogation to party group. They own a preference to party group at the same strength with party members. Additionally, their positive attitude toward party group (2.88) and non-party group (2.54) was not significantly different (t(11)=-1.513; p=0.158). This finding eliminated the uncertainty of non-party people’s political identity in Study 1. Non-party people did not perceive their political affiliation as one a spect of their social identity, which is reasonable due to its “informal organization” nature. Table 2 Sensitivity (D′) calculations for the four pairs non-party people

party members

Associations

D′(non-party people) vs.

D′

t test

D′

t test

D′(party members)

non-party group+

2.54

t(11)=8.899;

2.47

t(6)=8.00;

t(17)=0.165;

good

(0.989)

p<0.001

(0.817)

p<0.001

2.88

t(11)=14.544;

2.88

t(6)=10.87;

(0.685)

p<0.001

(0.701)

p<0.001

2.46

t(11)=12.070;

3.20

t(6)=14.01;

(0.705)

p<0.001

(0.605)

2.38

t(11)=10.676;

2.69

t(6)=7.48;

t(17)=-.782;

(0.771)

p<0.001

(0.951)

p<0.001

p=0.445

party group+ good non-party group+ bad party group+ bad

p<0.001

p=0.871 t(17)=-0.013; p=0.989 t(17)=-2.345; p=0.031

Study 3 Study 3 conducted an experiment using the priming paradigm to test the hypothesis 2 that people’s political identity would affect their political evaluation. Method. We manipulated the social identity of participants by making their political affiliation salient or not. The dependent variable was participants’ political evaluation on government performance. Participants. Participants were 20 college students with party membership. Each participant was voluntary to participate. All of them were not aware of becoming participants because of their party membership. Materials. The participants’ political evaluation on government performance were measured by political evaluation questionnaire which consists of seven items concerning issues of job services and unemployment, high prices and inflation, individual income tax threshold, ethnic conflic t, nuclear safety, food safety, and environment protection. The priming background questionnaire asked question about confirmation of party membership and time when they took part in the party. The control background questionnaire asked participants’ liking of fruits. Procedure. Participants were randomly assigned to controlled group in which they complete the control background questionnaire, or experimental group in which they finished the priming background questionnaire. Then all participants were asked to give a personal evaluation on the performance of government in seven issues. Result and discussion. The results showed significant difference between control group (5.90) and experiment group (7.60) in evaluating government performance on the issue of ethnicity conflict (t(18)=-2.573; p=0.019). Significant differences were not found in other issues, however.

These finding supported the hypothesis 3. Table 3 T test for the difference between control group and experimental group Issues

M (SD)

t (df=18)

p

5.30 (1.494)

0.000

1.000

5.00 (1.826)

5.20 (1.398)

-0.275

0.786

tax-threshold for individual-income

5.40 (1.713)

6.20 (0.919)

-1.302

0.214

ethnicity-conflict

5.90 (1.853)

7.60 (0.966)

-2.573

0.019

nuclear-safety

6.30 (2.163)

6.90 (1.101)

-0.782

0.444

food-safety

4.20 (1.317)

3.50 (2.224)

0.857

0.403

environment-protection

4.70 (0.949)

3.90 (1.912)

1.185

0.251

control

Experiment

job-unemployment

5.30 (1.494)

price-inflation

General Discussion Study 1 and Study 2 provide implicit and explicit evidence that party members gain not only an objective party membership but also a strong in-group preference to their party group accompanied by out- group derogation to non-party group. These finding support our hypothesis 1. In contrast to previous study [12], we add an implicit level evidences to examine people’s political identity, which provided a more comprehensive perspective. An advantage to assess people’s social identity on implicit level is that we can eliminate the possible situational factors when conducting survey assessment. For people who have party membership, their political affiliation should be taken into account, just as other identities such as their race, gender and nationality, when considering political relationship. This portion of identity would lead to conflict with other parties or non-party people. However, the similar finding is not found in non-party people, which indicate that non-party people do not form a social identity on basis of political affiliation as “non-party members”. This is reasonable, because non-party people are not highly organized. The results of our study were the first time to suggest that non-party people’s political identity may be not existed. Study 3 has demonstrated that political affiliation as an aspect of social identity will influence political evaluation of specific topics. These findings support the hypothesis 2. Party members were more satisfied with government’s dealing with problem of ethnicity conflict when their political identity was made salient. However, when party members’ political identity was not activated, they were more critical in evaluating the ethnicity issue. In a multi-ethnic country, people’s political affiliation should be considered when dealing with ethnicity problems. Different from previous studies [3, 13, 14], Study 3 successfully manipulate participants’ political identity to provide the experimental evidence that the causal relationship between political affiliation and political evaluation exists in specific issue. Social identity theory suggests that people identified with a group would fight for the interest of that group. When asking people to give a political evaluation in realistic situation, the result would be confounded depending on whether there are situational cues that would make people’s political identity salient. In sum, the present study provides both implicit and explicit evidence for the existence of political affiliation as a social identity and tested its causal relationship political evaluation. These findings provide an example of studying political affiliation from social identity perspective, from both explicit and implicit level, and from an approach of experiment.

Acknowledgement This research was financially supported by the Key Funding of Ideological and Political Education for Colleges and Universities in Guangdong Province of China (2009ZZ020). References [1] Tajfel, Henri; Turner, John. "An Integrative Theory of Intergroup Conflict". In Austin, William G.; Worchel, Stephen. The Social Psychology of Intergroup Relations. Monterey, CA: Brooks-Cole. 1979, pp. 94–109. [2] Hogg, Michael A. & Terry, Deborah J. Social Identity Processes in Organizational Contexts. Michigan, Sheridan Books, 2001. [3] Conover, P. J. The influence of group identification on political perception and evaluation. Journal of Politics, 46(1984): 760–785. [4] Deaux, K.. Social Identity. In: J. WoreU (Ed), Encyclopedia of Women and Gender. 1-2. Academic Press, 2001. [5] Daniel J. Benjamin, James J. Choi, and A. Joshua Strickland. (2010). Social Identity and Preferences. Am Econ Rev. 2010 September 1; 100(4): 1913–1928. [6] Collins, A. M., & Loftus, E. F. A spreading activation theory of semantic processing. Psychological Review, 82(1975): 407-428. [7] Fazio RH & Olson MA. Implicit measures in social cognition research: their meaning and use. Annual Review of Psychology, 54(2003): 297-327. [8] Turner, R.N., Forrester, R., Mulhern, B. & Crisp, R.J. Impairment of executive abilities following a social category prime. Current Research in Social Psychology, 11(2005): 29-38. [9] Shih, Margaret, Todd L. Pittinsky, and Nalini Ambady. “Stereotype Susceptibility: Identity Salience and Shifts in Quantitative Performance.” Psychological Science, 10(1999):80-83. [10] Shayo, M. A model of social identity with an application to political economy: Nation, class, and redistribution. American Political Science Review, 103(2009): 147-174. [11] Conover, P. J. The influence of group identification on political perception and evaluation. Journal of Politics, 46(1984): 760–785. [12] Deng Zhiwen & Qing Ding-wen. A Probe into the Social Identity Status of Undergraduates --a case study of an university[J].Journal of Changsha University of Science and Technolo gy(Social Science Edition), 2(2006):120-123 [13] Aburime, T.U. Impact of Political Affiliation on Bank Profitability in Nigeria. Journal of Global Economy, 5(2009): 251-261. [14] Abel, L. & Hornsey, M. Social identity and moral judgement: The impact of political affiliation on the evaluation of government policy. In R. E. Hicks ( Ed.), Personality and individual differences: Current directions (pp. 113-122). Bowen Hills: Australian Academic Press, 2010. [15] Nosek BA, Banaji MR. The go/no-go association task. Social Cognition. 19(2001): 625–66.