A Study of The Relation of Gender and Self-esteem in Conformity

Abstract

Conformity studies have been extensive following Asch’s (1956) classic line judgement task. Numerous factors have been purported to affect conformity behaviours. The present study investigates gender differences in conforming behaviours, and how self-esteem is associated with conformity, by replicating Asch’s original experiment. The study further explores if self-esteem is associated with conformity differently between gender groups. Twenty-two psychology undergraduates (eight males and fourteen females) were recruited. An independent t-test and bivariate correlation analyses were used to analyse the data. The results revealed 1) no gender difference in conforming behaviours, 2) a significant negative correlation between self-esteem and conforming behaviours, 3) differential association of self-esteem and conformity between males and females. In all, the results suggest that conformity is influenced by an individual’s sense of self-worth instead of a person’s gender identity; gender may be a moderating variable in self-esteem and conformity research that future studies can explore.

Gender and Self-Esteem Differences in Conformity: Revisiting Asch’s Conformity Test

People are often faced with situations where they are pressurized to conform to certain norms or behaviours. Conformity is the change in behaviours in response to real or imagined pressure from members of the social environment (Gilovich, Keltner, Chen, & Nisbett, 2013). In a classic study by Asch (1956), it was demonstrated that people can give obviously wrong answers even when they knew it was incorrect because of the sense of pressure from others. Specifically, participants in the study were shown three different comparison lines and were asked which line was similar to a target line presented adjacent. The participants were seated among a group of 7 to 9 and had to answer after the others indicated their response. There was only one participant in each group and the rest were confederates who gave unanimous incorrect answers on specified trials, known as critical trials. Approximately 75% of the participants conformed on at least one trial – of these participants, 5% conformed on all critical trials. Qualitative analysis after the experiment revealed that the participants knew the answers were incorrect, but felt the need to conform as they did not want to feel rejected, suggesting that people have a psychological need to fit in with a group.

The yielding to group pressure because of a need to fit in is termed normative social influence, as opposed to informational social influence wherein people are uncertain of a situation and look to others for guidance (Asch, 1956; Gilovich et al, 2013). Clearly, Asch’s study explores the aspect of normative social influence. Numerous studies had since found divergent findings (Bond & Smith, 1996; Mori & Arai, 2010), questioning the external validity of the study. Perhaps the most important limitation to Asch’s study was the use of only male participants, indicating the need for further studies exploring gender differences. Literature on gender differences remains inconsistent – while women was generally found to conform more than men in the past (Bond & Smith, 1996), recent studies suggest no gender differences (Rosander & Eriksson, 2012), or mixed results in conformity between gender groups (Enjanjan, Zeigler-Hill, & Vonk, 2015).

Reviewing Gender Differences

Women were generally found to conform more than men possibly due to gender roles and the conforming towards such roles during the period (1960s – 1990s) when the experiments took place (Rosander & Eriksson, 2012). The social identity of women was arguably more submissive and conforming than present. Good and Sanchez (2010) posited that people conform to gender roles of the society due to intrinsic enjoyment of pulling together an individual’s actual and ideal selves, or due to extrinsic pressure from society. According to the self-discrepancy theory (Higgins, as cited in Gilovich et al., 2013), people are constantly motivated to compare between their actual selves – representing who people truly believe themselves to be – and two other selves, the ideal and ought selves. The former refers to the more positive connotation of people’s ambitions and what others maintain about them; the latter refers to the sense of obligation people feel pressurized to follow. Bond and Smith (1996) meta-analysis seems to substantiate this theory.

In retrospect, the review was also done in the 1990s and conformity behaviours might have changed since then. Agreeably, Rosander and Eriksson (2012) found that women did not conform more than men. The study utilized the Asch’s paradigm, but on the internet where participants were not interacting face-to-face with each other. An additional measure of task difficulty (Easy/Difficult) was introduced. The overall conformity result was like Asch’s original study. Additionally, men were found to conform more than women on difficult task (specifically on difficult and logical questions, for details see Rosander & Eriksson, 2012). This is corroborated by Enjanjan and colleagues (2015) where men with varying levels of self-esteem were found to conform more on difficult trials. Furthermore, women might not have been more conforming but that men tend to report less conformity depending on the context (Rosander & Eriksson, 2012). Recent explanation of why men conform less alludes to the idea that non-conformity portrays uniqueness, thus making an individual stand out and increase prospective opportunities such as leadership roles (Griskevicius, Goldstein, Mortensen, Cialdini, & Kenrick, 2006; Rosander & Eriksson, 2012). Taken together, the findings suggest no concrete direction of gender difference, and hints a relationship between conformity and self-esteem.

Reviewing Self-Esteem in Conformity

Following the self-discrepancy theory, conformity towards ascribed norms may increase self-esteem due to inherent positive feelings. Conversely, conforming due to feelings of pressure may induce lower levels of self-esteem (Good & Sanchez, 2010; Gilovich et al., 2013). Alternatively, this paper aims to identify if self-esteem levels affect people’s conforming behaviours. One of the main reasons people conform is the desire to fit in or to be correct (i.e. normative and informational social influences, Gilovich et al., 2013). Since these desires are associated with self-esteem (Enjanjan et al., 2016), it is intuitive to think that people with low self-esteem conforms more. Truly, studies converge on the notion that individuals with low self-esteem tend to conform more than individuals with high self-esteem (Enjanjan et al., 2016), possibly to protect their weak sense of self and mitigate the damaging impacts of failure (Ardnt, Schimel, Greenberg, & Pyszczynski, 2002). Research on the relationship of gender in self-esteem and conformity remained scarce, with identifiable studies placing self-esteem as the outcome/dependent variable in their study instead of conformity (Good & Sanchez, 2010).

Due to inconsistencies in the literature, this paper expects a non-directional gender difference in rate of conformity. Next, it is expected that self-esteem is negatively correlated with conformity. Finally, this paper further explores if self-esteem levels between gender groups is differently associated with conformity.

Methods

Design

There were two designs of the study – a between-subject quasi-experimental design for test of group difference and a correlational design for test of association. The independent variable for the former test was gender; the dependent variable was social conformity operationalised as the rate of conformity (out of 12 critical trials) of Asch’s (1956) conformity test. Self-esteem, measured using Rosenberg Self-Esteem Scale (RSES; Rosenberg, 1965a), was correlated with the dependent variable. Participants responded to both RSES and the conformity experiment.

Participants

A total of 22 Participants (8 Males, 14 Females), age ranging from 18 – 25 years (M = 20.7, SD = 2.35 years) were recruited from James Cook University as part of their course requirement PY3102. Majority of the participants (N = 19) were First year students, and the remaining (N = 3) were Second year students. All participants were included in the study as all relevant details were filled up in both the demographics form (Gender, Age, and Year of study in college), and all participants completed the experiment.

Materials

Asch’s (1956) line judgement task. Microsoft PowerPoint 2016 was used to create the stimuli and presented through in-class overhead projector in James Cook University Singapore, room C2-06. The stimuli (Appendix A) consist of 18 trials of line judgement test, 12 of which were critical trials whereby the researcher recorded responses of the participants. The confederates were told to give wrong responses unanimously on the critical trials. The target lines were copied exactly (copy-paste function) from the correct comparison line while the remaining lines were created such that the correct answer was always obvious. All lines were between 2 – 10 inches (5.08cm – 25.40cm) following the original study.

Rosenberg self-esteem scale (RSES; 1965a). The RSES is a 10-item scale that measures feelings of self-worth by assessing both positive and negative feelings about the self (Rosenberg, 1965b). The items are scored on a 4-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Items 2, 5, 6, 8, 9 are reverse scored. Item scores are summed and higher scores indicate higher self-esteem (Appendix B). The scale was found to have good internal consistency, a = .91, and demonstrated good internal and external validity (Sinclair et al., 2010).

Procedure

Participants were given an information sheet and were told that their participation is completely voluntary; they could leave at any point in the experiment. Only one participant was present for each experiment. Upon agreeing to informed consent and filling up of the demographics sheet, participants completed the RSES. Participants were told that they were in a group study of visual judgement together with three other confederates, and were assigned to the last seat in the row. The researcher presented the line judgement task on screen (5 metres away) and participants had to answer: “Which comparison line A, B, or C is the same as the target line?” after responses from all confederates. The confederates were instructed to give unanimous correct responses on trials 1, 2, 6, 10, 15, and 16 (randomly generated order, except for 1 and 2, following Asch’s original study), while incorrect responses for all other trials (critical trials). Only responses from the critical trials were recorded. Finally, participants were debriefed and given the true nature of the experiment.

Statistical Analysis

IBM SPSS 22 was used to analyse the data. An independent t test was used to test for gender differences and bivariate correlation analysis was used to test for the relationship between self-esteem and the dependent variable.

Results

The present study explored gender differences in social conformity, and its relationship to self-esteem. Table 1 presents the descriptive statistics of each group. Contrary to the hypothesis, an independent t test revealed no significant differences between male and female, t(20)= -0.067, p = .947, 95% CI [-0.256, 0.240]. A significant moderate negative correlation was found between self-esteem and social conformity, r(20) = -.58, p = .004, indicating that people with higher self-esteem tend to conform less. Finally, this study further explored if self-esteem levels between males and females would be associated differently with rate of conformity. A bivariate correlation analysis between gender groups indicated differential association with conformity – there was a significant strong negative correlation in the female group, r(12) = -.73, p = .003; there was no significant correlation in the male group, r(6) = -.31, p = .45.

Discussion

In all, the results revealed no gender differences in rate of conformity, a significantly moderate negative correlation of self-esteem and conformity, and different associations of self-esteem and conformity in males and females – women with higher self-esteem seems to conform less, while there was no significant correlation in men with varying levels of self-esteem. Contrary to previous findings, it appears that males and females do not differ in their rate of conformity. One possible explanation can be that gender roles in the current globalized society are not as prominent and people are more liberal in their social identity, thus blurring the distinction between gender.

The result was consistent with the hypothesis of negative correlation of self-esteem and conformity, thus supporting previous findings and the notion that conforming behaviours seem to be a form of defensiveness toward threats to the self (i.e. fear of criticism or judgement by others, feeling excluded etc). Additionally, this paper showed that different levels of self-esteem in males and females are associated with conformity, consistent with Enjanjan et al.’s (2016) findings. It is likely that females with low self-esteem value the desire to be liked by others more than men with low self-esteem, while women with high self-esteem is more confident to make autonomous decisions. Conversely, men probably do not consider conforming behaviours as being impactful to their self-esteem as they may feel that uniqueness is a preferable trait rather than being liked by others. However, it is puzzling that no gender difference emerged overall, but differential associations emerged at varying levels of self-esteem in males and females. The findings imply that conformity in people may be related more to an individual’s sense of self-worth rather than to gender, and complex interactions may exist between gender groups and self-esteem levels. The findings may be of interest to people working in groups, especially leaders – to identify the tendency for people to conform and therefore impeding productive generation of ideas.

Limitations and Future Directions

This study is limited due to the small sample size (N = 22), and very low count of male participants (n = 8). Also, all participants were psychology students and they might have guessed the nature of the study. This was controlled by asking for the year of study – with older-year students having the tendency to guess the true nature. Furthermore, due to the quasi-experimental nature, random assignment was not possible thereby confounding the results. Future studies may recruit more participants and include a question at the end of the experiment (“What is the study about?”) to exclude participants who know the true nature of the study. This was not done in the present study due to restrictive sample size. Perhaps with adjustments to the study and with more advanced statistical procedures, complex interactions between gender, self-esteem, and conformity can be discovered.