Studying Social Identity, Intra Group And Intergroup Perceptions

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

The purpose of this lab was to determine cohesion in and between groups. A sample of respondents filled questionnaires rating their perception of relations both within and between their in-groups and out-groups. The questionnaires rated their perception on a scale of 1-7 representing their level of agreement or disagreement with the matter question. The answers were subjected to statistical analysis using SPSS and results analyzed and interpreted. The results indicated that the most important factors affecting identity ranked in their order of importance are cohesiveness, voice and consensus respectively.

Social identity refers to how human beings make sense of each other. As such, social identity is critical to the construction of culture and by extension society (Turner, 1985). Human beings need to know the affiliations, beliefs, intentions of others to interpret their actions and/or predict their future behavior (Turner et al, 2008). While the qualities aren't observable directly, they can be externally manifested through signals revealing of internal self (Turner et al. 2008). Biologically, such signals include simple mappings that have a direct link between signal and trait; for instance strong body signals strength. The process of social identity is markedly different as individuals hardly build perception of other on a trait by trait basis (Tajfel & Turner, 1979). Rather, they interpret observations and several preexisting prototypes of others to enable them create a richly detailed impression of another. Thus, getting to understand the process by which these prototypes are shaped changed and shared across a group of people that constitute a culture and how individuals apply them in categorizing others is a critical to understanding social identity (Spears, Lea & Lee, 1990).

According to the Social Identity Theory, individuals do not have a "personal self". Rather, they have several "selves" each corresponding to the wide spectrum of group membership. Following from that, an individual will think, act and feel in accordance with his level of self as defined by him/herself, his/her family, group or country (Turner et al, 1987). Typically, an individual has more than one "social identities" apart from the earlier described 'level of self". According to Hogg & Vaughan (2002), social identity is the self concept of an individual resulting from his/her perceived affiliation to social groups. Therefore, social identify is the respective individual's perception of what defines the "us" coupled with his/her internalized group membership as different from the concept of personal identity that describes the knowledge of self as defined by a person's unique attributes. This is where social identity and personal identity diverge (Platow et al, 2006).

The theory of Social Identity further emphasizes that membership of a group results in emergence of an in-group and/or self-categorization and enhancement in a manner that discriminates members of out-groups while and favoring those of the in-group. To illustrate this, Turner and Tajfel (1986) did a study titled Minimal Group Studies. The study indicated that by individuals forming themselves as a group members, it is enough to cause favoritism within the in-group and discrimination of out-groups (Foles, 2006). After formation of a group, persons will seek self esteem by comparing themselves with the out-group as regards one valued dimension thus differentiating the in-group (Krizan & Baron, 2007). This, as Tajfel and Turner (1979) noted, is the reason why an individual's sense of identity is defined not as "I" but as "we". The two researchers identified three variables that contribute to the emergence of in-group favoritism which are;

A) The degree to which the members of an in-group identify with an in-group so as to interpret the membership to the in-group as an aspect of their self-concept (Tajfel & Turner 1979).

B) The degree to which the prevailing situation offers basis for groups comparisons.

C) The comparison group's perceived relevance relative to the status of the in-group: favoritism is displayed when the in-group is vital to their definition of self and when the comparison is particularly meaningful and/or the outcome is contestable (Tajfel & Turner 1979).

The main purpose of this lab was to support the social identity theory. In particular, the lab sought to determine whether it is indeed true that individuals will suppress their identities in favor of the group. The hypotheses of this lab were as follows

First: people in minimal group conditions show in-group identification and group cohesion to the same degree as when people adopt a central social identity, namely their gender identity

Second: people in minimal group conditions do not show in-group identification and group cohesion to the same degree as when people adopt a central social identity, namely their gender identity


Data regarding social identity variables were collected. The sample involved 88 participants. Participants completed a survey to measure their intra-group (cohesiveness, voice, consensus and identification) and intergroup (in-group and out-group trait rating) perceptions. The sample questionnaire attached on the appendix gives the items participants completed for each scale and the SPSS data base shows the scores on each item and the mean score for each scale. Altogether participants completed the following 5 measures for the purposes of the lab report including Group Cohesiveness (COH), Group Voice, Group consensus (Consen) Group identification (Ident) In-group and Out-group Trait ratings. All participants chose to participate, and were given a series of instructions and a task. The task involved a group discussion on a complex discussion problem that would bring out the various aspects to be studied. The approach used 2 Ã- 2 experimental design involving group type (adhoc vs gender) and competition (Competition vs. Non-competition) which served as the key aspects of the study. This was done by forming ad-hoc groups or gender based groups in tutorials and instructing groups to compete to get the best intelligence score in an IQ test.

Each question was assigned a variable code for ease in processing in SPSS. The variables included the various aspects of cohesiveness (including preference for the particular group, sense of belonging, how the members get along, if the members are protective of each other against criticism from outsiders, preference for coalescing after parting, closeness of members) voice

( including expression, input, cooperation, competition, shared objectives, decisions making, teamwork, harmony within groups, consensus) identity including( interest, respect attitudes and values) in-group and out-groups( including borrowing best practices from other groups) intergroup competition and individual traits within the group and outside the group. Respondents were asked to rate their perception of the aforementioned variables in a scale beginning with one for those who strongly agree and ending with 7 or 9 for those who strongly disagree.

To mitigate against errors occasioned by variations in age and gender differences among individuals and groups, data was also collected to reflect the extent to which respondents identify with their respective age and gender categories. This was done in the realization that different individuals will have different level of identification with their genders. The collected data was run in SPSS to reveal underlying trends and statistics.


The results indicated that all the three elements discussed were relatively reliable for making the inferences. The values were cohesiveness (0.82), voice (0.887), consent (0.789), identity (0.852), in-group (0.853) and out-group (0.916). The respective mean and standard deviations for the variables were cohesiveness 5.2971, 0.963, voice 5.955, 0.918, consent 0.5971, 0.8806, identity 5.8086, 0.83386, in-group 7.3766, 0.82015 and out-group 6.6218, 1.11007.

To best reveal the intra-group and intergroup cohesiveness, variables descriptive of each were analyzed differently. As such, the first four variables (cohesiveness, voice, consent, and identification) revealed t-statistic values of 17.187, 27.186, 28.365, and 27.605 respectively at 95% confidence level. On the other hand, the in-group and out-group evaluation revealed a t-statistic values of 36.768 and 18.248 respectively also at 95% level of confidence.

The Pearson correlation coefficients varied across the variable spectrum. Of all the variables used in the analysis, none returned a correlation value less than 0.1. As a rule of thumb, a correlation coefficient of between 0.0-0.30 is considered weak, between 0.30-0.70 considered moderate and 0.70-1.00 strong. Judging by this, identification and cohesiveness, identification and consensus, identification and voice, returned the highest correlation indicating presence of a moderate relationship. The correlation of out-group evaluation and all other variables was the weakest.

To reveal the effect of gender and age considerations on group identity, gender and age variables were put through the Levene's test for equality of variances. Assuming equal variances, gender identity revealed an F value of 1.749 while that of age was 0.847 all at 95% confidence interval. Further, competitiveness check between these two variables also revealed an F value of 5.833. The tests of between subject effects reveal that the differences in variance are statistically significant for the groups.

A 2 (Group Type: Ad-hoc vs. Gender) Ã- 2 (Competition vs. Non-competition) ANOVA was conducted using the Group cohesiveness scale as the dependent variable. The main effect

for cohesiveness was significant, F(1.058, df3) = 0.369, p = 0.44, eta2= 0.20, such that those under the gender condition (M =5.4143 SD =1.0283 ) reported stronger cohesiveness than those under the ad-hoc group condition (M = 5.0723, SD = 1.11306). The main effect of competition condition was significant, F (1.058, df3), p = 0.369, eta2 = 0.20, such that those who were instructed to compete (M = 5.3782, SD =0.83779) reported stronger cohesiveness than those who were not instructed to compete. The interaction was also significant and showed that participants under the ad-hoc non-competitive condition (5.0723, SD = 1.11306) reported significantly less group cohesiveness than participants under the ad-hoc competitive condition (M = 5.3782, SD =0.83779). Accordingly, both hypotheses are supported.


Social identity is a function of several variables. In this lab experiment, five of those variables were analyzed to determine their impact on social identity. All the five retuned a reliability level of more than 0.80 indicating a high internal consistency. All the responses for voice, cohesiveness, consensus and identification had a relative leaning towards strongly agreeing all with a mean more than 5. Equally, their standard deviation was small signifying that there was little spread between the perceptions of individuals. This is a little different when it comes to in-group and out-group where the mean perception lean toward strongly agreeing but perceptional differences between individuals are wider as indicated by the increased standard deviations (Brewer, 1979). This is consistent with the social identity that members of an in-group tend to have the same perceptions and will discriminate against perceptions of members of out-groups as well as try to protect (Brewer & Silver, 1978).

Notably, there are factors that have a greater correlation coefficient than others. For instance cohesiveness, voice, consensus is highly correlated to identification. This coincides with the common perception that individuals are more likely to feel accepted in groups where they naturally fit in (cohesive), are listened to (voice) and reach a consensus on a critical mass of issues (Austin & Worchel, 2005). Again and consistent with the last point, the out-group perception of the individuals within the in-group is immaterial as indicated in the low correlation c-oefficient between in-group and out-group evaluation against all the other four variables. Gender and age have a significant effect on group identity as indicated by the statistical significance factors of 0.188 and 0.359 respectively (Foles, 2006).

The research has shown that when group members face active competition from another group (call it an out-group), they are more likely going to form a united front to counter the common threat. In the process, they identify similarities within themselves as they harness group synergy against the common external threat thereby increasing cohesiveness. This principle has been used in political science where leaders tend to focus their followers on a common external threat-mostly an external enemy- to increase their cohesiveness. This is especially so when internal cohesion was under threat (Foles, 2006).


It is evident from the above that the most important factor affecting group cohesiveness is competition. As noted, when group members face active competition from an out-group, they cluster to form a united front to counter the common threat. Group members then identify similarities within themselves as they harness group synergy against the threat. The students exhibited the same characteristics as has been supported under the social identity theory. Hence, cohesiveness is dependent upon several other factors such as group consensus, voice and acceptance among others.