briefly summarize the bottom article (study) and how it is an example
of a predictable behavior, Why do you think people behave similarly in certain situations?
(answer to the best of your knowledge using below study)
Liuzza, M. T., Macaluso, E., Chiesa, P. A., Lingiardi, V., & Aglioti, S. M. (2019). An fMRI study on the neural correlates of social conformity to a sexual minority.Scientific Reports (Nature Publisher Group), 9(1) doi:http://dx.doi.org.nuls.idm.oclc.org/10.1038/s41598-019-40447-3
An fMRI study on the neural
correlates of social conformity
to a sexual minority
M. t. Liuzza 1,2,3, E. Macaluso2,4, P. A. Chiesa 1,2,5, V. Lingiardi6 & s. M. Aglioti1,2
Social conformity refers to the tendency to align one's own behaviors, beliefs and values to those of others. Little is known about social influence coming from a minority group. To test whether social pressure from sexual minorities triggers avoidance-motivated behaviors, we explored how being influenced by the preferences of gay peers modifies the behavioral and neural reactivity of individuals defined as in- vs. out- groups on the basis of sexual orientation. To this aim, we combined fMRI with
a social conformity paradigm in which heterosexual and gay/bisexual (hereafter non-exclusively heterosexual, NEH) individuals provided with male body attractiveness ratings by a fictitious group
of gay students may or may not alter their previous rating and may or may not conform to the mean. Behaviorally, conformity to the minority preference was found in in-group NEH more than in out-group heterosexuals. Analysis of BOLD signal showed that social pressure brought about increased brain activity in frontal and parietal regions associated with the detection of social conflict. These results show that members of a sexual majority group display a smaller level of conformity when a sexual minority group exerts social influence. However, the neural correlates of this modulation are yet to be clarified.
Social Conformity refers to the act of changing one's own thoughts, feelings, and behavior to match the responses of a given group1, even when such responses appear blatantly wrong2. Generally, social conformity is more pro- nounced for opinions expressed by in-group than by out-group members3. While a number of psychological studies have been conducted on conformity, only recently has this topic attracted the interest of social neuro- science4 and the picture emerging from the few fMRI studies on the neural correlates of social conformity is largely incomplete. In a seminal study, Klucharev Hytönen, Rijpkema, Smidts, & Fernández5 demonstrated that participants exhibit increased neural activity in rostral cingulate cortex and decreased activity in ventral striatum when providing ratings of facial attractiveness that conflict with those of a group. These two brain regions play a crucial role in conflict monitoring and reward processing, respectively. Also, the social conformity-related choice of changing a decision and accepting an unfair offer was associated with increased activity in the medial prefron- tal cortex6. Moreover, it has been demonstrated that activity in the dorsomedial prefrontal cortex (dmPFC) is modulated by changes in preference for T-shirts worn by a liked group of peers instead of a disliked group of sex offenders7. While it is clear that sex offenders represent a social threat, stigmatization may concern groups that do not pose any objective physical harm, such as gay people. Behavioral and neural data concerning reactivity to the social pressure exerted by this type of group is largely lacking.
To address this issue, we combined fMRI with a modified version of an experimental paradigm that has been successfully used to uncover the neural correlates of social conformity5. Our main aim was to investigate whether the natural human tendency to conform to others is contrasted by avoidant behavior towards the stigmatized, minority group. In our study, avoidance was operationally defined not in physical terms, but in terms of willing- ness to align with aesthetic evaluations expressed by a sexual minority group.
Importantly, conformity in our task was linked to a defining characteristic of the minority group, namely attractiveness towards male bodies. We decided to use an attitudinal object related to sexual orientation in order to maximize the effect under investigation. Indeed, we speculated that homophobic motivations should fuel the willingness to not conform to a sexual minority when the attitude towards an object related to sexual orientation could be perceived by a homophobic participant as a sign of sexual deviance.
In order to assess the extent to which the hypothesized non-conformity behavior could be linked to anti-gay prejudice, we collected explicit measures of modern homophobia (i.e., the modern homophobia scale (MHS)8 and right-wing authoritarianism (RWA9) that are consistently associated to sexual prejudice towards lesbians and gay men10,11. Because these measures are also intimately related to social conformity12, we expected them to be negatively related to conformance with gay men ratings of body attractiveness. Moreover, in order to circum- vent the effect that social desirability can exert on self-reported attitudes13-15, we also used implicit measures of anti-gay attitudes16 that assess the strength of the association between positive/negative words and the categories of gay and straight people.
Finally, we expected that some of the brain regions that predict conformist behavior when no in-group vs. out-group coding is at play (e.g., the medial prefrontal cortex)5 also predict non-conformist behavior when a sexual minority exerts the social influence. This result would support the notion that exposure to out-groups17and sexual minorities18 triggers avoidance motivated behaviors.
Materials and Methods
Participants. Following recommendations for the number of observations needed to obtain large effect sizes19, we tested 32 male participants (mean age = 28.81 years, SD = 7.26). Participants rated their sexual orien- tation on a Kinsey Scale20 ranging from 0 ("exclusively heterosexual") to 6 ("exclusively homosexual"): fifteen par- ticipants rated themselves as exclusively heterosexual, one as equally heterosexual and gay, two as predominantly gay (only occasionally heterosexual), and fourteen as exclusively gay. They were divided into two groups based on these responses: 15 exclusively heterosexual (Kinsey = 0; mean age = 25.53 years, SD = 3.94) and 17 non-ex- clusively heterosexual participants (NEH, mean Kinsey = 5.71, SD = 0.77, mean age = 31.88 years, SD = 8.27). Although our original plan was to split the sample evenly, we grouped the bisexual participant with the NEH, as studies have found that bisexuals display the same arousal pattern as gay men21 and because they are perceived (and stigmatized) as a sexual minority22.
All participants had a normal or corrected vision, were free from any contraindication to fMRI and had no history of major psychiatric or neurological problems, as assessed through a questionnaire filled by the partici- pants prior to entering the scanner.
The study was conducted in accordance with the Declaration of Helsinki. The protocol was approved by the independent Ethics Committee of the Santa Lucia Foundation (Scientific Institute for Research Hospitalization and Health Care) on the 17th of May 2013 (Prot. CE/Prog. 403-19). All participants gave written informed consent.
Visual stimuli. 128 virtual male bodies were prepared by using Poser© Pro, a software that creates avatars. We manipulated muscles, thinness, and heaviness of 4 models (Diego, Marcus, Tomo, and Ryan). We then asked an independent group of 16 randomly selected, exclusively heterosexual (Kinsey = 0) female participants (mean age = 27.25, SD = 3.09) to rate the models according to both liking ("how much do you like it?") and beauty ("how beautiful is it?") on a 1 ("not at all") to 7 ("very much") Likert-type response scale. Because the two meas- ures were highly correlated (r = 0.99, P < 0.001), we averaged them in a unique attractiveness value. We then excluded 20 images that had mean ratings approaching the minimum possible rating (i.e., the ones judged as the least attractive). Indeed, images that are universally rated very low would systematically prevent our algorithm to generate lower ratings from the fictitious gay group (see below for more details). The final set of 108 stimuli had an average rating of 3.51 (SD = 1.49, range = 1-7). We chose a sample of women under the assumption that they would be more likely to provide more reliable attractiveness rating for male bodies as compared with men, because their self-reported evaluation should be less prone to self-report bias. In fact, even though ratings of male body attractiveness have been provided by males in previous studies23, we reasoned that some men who hold negative attitudes towards homosexuality might distort their overt evaluations on the beauty and attractiveness of male bodies. To rule out this possibility we obtained a post-hoc validation of the stimuli by asking fifteen heterosexual men (mean age = 35 years, SD = 5.24) to rate for body attractiveness of the original 128 pictures. The average ratings from the male group closely matched the ones provided by the female group (Spearman's rho = 0.90, Pearson's r = 0.90), sharing the 81% of their variance. Also, the two-way intra-class-correlation for the average ratings of the two groups (random raters) was good (ICC2 = 0.75). We also looked at what would have happened had we selected the 20 worst (least attractive) images on the basis of men ratings and found that 15 out of 20 (75%) would have overlapped with the female ones. The Cohens' kappa on the selected images indicated a substantial agreement (Cohen's K = 0.70) between the two groups. Finally, we addressed if the ICC2 within each of the groups for the single ratings was indicative of any possible difference in reliability between the two groups. We found that the ICC2 was fair in the female (ICC2 = 0.4), but poor in the male (ICC2 = 0.14) sample.
Experimental procedure. We employed a modified version of the conformity paradigm used by Klucharev and colleagues5. In their fMRI study, the initial judgments of facial attractiveness provided by females were open to the influence of a group of peers ("average European female participant from Milan and Paris"). Participants provided the first rating and attended to the normative group rating in the scanner. The experiments system- atically manipulated the ratings of the normative group in order to have a similar amount of normative group
pressure towards higher ratings, lower ratings and same ratings. Participants were asked to rate the same female faces again, after the fMRI session to test the effect of the social influence exerted by the normative group.
Here we changed the original procedure in three ways: (1) Participants in Klucharev and colleagues' study were European (Dutch), the average European rating represented a majority ingroup. In contrast, in our study we wanted to assess the influence of a sexual minority group's (gay men) on the conformity behavior of minority in-group vs. majority out-group, defined by sexual orientation (i.e. non-exclusively heterosexual vs. heterosex- ual men); (2) We used male bodies instead of faces with the aim of maximizing the non-conformity effect in homophobic participants. Homophobic individuals may fear assimilation into the prejudiced group when the domain is directly related to sexual orientation24. Participants were then asked to rate the attractiveness of the body-stimuli on a 1-9 Likert-type response scale ("how attractive do you find this body?"); (3) We tested the effect of influence immediately after the exposure to the normative group ratings. We decided to show the bodies immediately after the first ratings because bodies are less salient than faces25 and therefore harder to be remem- bered even at an implicit level. Also, each body presented to the participants belonged to one of four identities, which makes even implicit recognition more unlikely and thus may hinder the effect of social influence on the second rating. In contrast, our procedure should maximize the effect. This is important also considering that our design is based on a relatively small sample size. Moreover, because implicit and explicit bias towards gay men is stronger than bias against lesbians11,16 participants in our study were all male, while in Klucharev et al. they were all female. Our sample might have been thus less susceptible to social conformity than the sample in Klucharev et al., as female participants are more likely to conform after undergoing social influence3 (but see26,27 for a possible alternative explanation of the putative gender differences in social conformity). These considerations motivated us to maximize the conformity effect through a procedure that makes the normative group rating highly salient.
During the fMRI experiment, the participants were asked to rate attractiveness of the male body stimuli and then; (i) in 83% of the trials, were presented with the mean ratings of the same stimuli provided by a putative gay sample (social influence condition); (ii) in 83% of trials were asked to rate the same body stimuli the second time. Crucially, the sample ratings attributed to the gay sample were fictitious and could be higher (Upward social influence), lower (Downward social influence) or the same (Agree) with respect to those provided by the partici- pants. The experimental stimuli had blurred faces in order to encourage participants to focus on flesh properties such as muscularity, a feature that has been shown to be important in determining attractiveness ratings for many gay men28. The ratings were given along a Likert-type response scale ranging from 1 (not at all) to 9 (very much) by moving a dark blue circle frame away from the middle value (i.e., "5") and towards either higher values (right button) or lower values (left button). Participants had three seconds to provide the rating, after which the white circle turned into a bright blue circle frame.
fMRI Results. Effect of social influence. We first investigated the hemodynamic responses related to expo- sure to the Social influence (i.e., Disagree vs. Agree) of the sexual minority group. The results of these analyses showed the activation of a fronto-parietal network, including the right middle frontal gyrus (rMFG), the intra parietal sulcus (IPS) bilaterally, and the left inferior frontal gyrus (lIFG). See Table 1 and Fig. 3.
We did not find any significant cluster for the interaction Group x Social influence effect.Correlations between Disagreement vs. agreement contrasts and behavioral conformity. We found a negative cor- relation between the Disagree vs. Agree contrast estimate (CE) and conformity behavior across all participants in three voxels (p < 0.05, FWE corrected) within a cluster that includes middle superior frontal gyrus/dorsomedial prefrontal cortex (dmPFC).
However, a closer inspection of the scatterplot displaying the relationship between the CE and the Conformity behavior in these voxels revealed the presence of an outlier participant (NEH) whose CE deviated more than three standard deviations from the rest of the sample. When that participant was dropped from the analyses, no significant cluster and/or voxel was found significant when correcting for the FWE.
Social conformity is a robust, evolutionarily rooted phenomenon likely motivated by the need for a given indi- vidual to be approved by other individuals and to avoid conflicts with them. The phenomenon may be par- ticularly patent when those who conform and those who lead others to conform belong to the same social group3. Pioneering studies on the neural underpinnings of social conformity highlight the important role of reward-related (ventral-striatal) and conflict-related (medio-frontal) regions in mediating the adjustment to the opinions of others43,44. Moreover, activity in medio-frontal regions tracks the direction of preference change from one's own opinion towards or away from that of liked vs. disliked groups7. However, to the best of our knowledge, no study has thus far explored the neural systems activated when social conformity is induced by a group defined as in- or out-group on the basis of sexual orientation. To fill this gap, we tested whether a minority group of sexu- ally stigmatized gay men would instigate social conformity to a different extent in a minority group of peers (i.e., exclusively gay or bisexual individuals) as compared to a majority group of exclusively heterosexual individuals. Moreover, we explored how any modulation of social conformity may be reflected in the activity of brain areas involved in social influence. Importantly, the perceptual, cognitive and emotional dimensions tapped by our con- formity task involved a feature conspicuously associated with the group that should lead to conformity, namely the physical attraction towards male bodies that gay men may typically experience.
One important behavioral result of our study is that social conformity, as indexed by the alignment of one's own ratings of male body attractiveness with the ratings given by a sexual minority of gay students, was stronger in gay and bi-sexual as compared to heterosexual people. More specifically, alignment in NEH ratings (as indexed by the delta between the first and second rating in the direction of the social influence) significantly differs from that of the heter- osexual participants. Tellingly, no differential conformity was instigated by a fictitious group whose sexual preference was not specified. Indeed, no differences in the delta between first and second rating in the direction of the social influ- ence was found in the control task where participants had to rate the attractiveness of paintings that did not depict male bodies. This finding rules out any group difference in the general tendency to conform to others and suggests that the observed difference in conformity may be determined by the sexual orientation of the influencing group.
With that being said, we acknowledge that the aesthetic rating of a sexually neutral stimulus may constitute a suboptimal control for our main task. Future studies should focus more systematically on the evaluation of sexually neutral stimuli along with the manipulation of the group that exerts the social influence (majority vs. minority).
It has been argued that results from studies on conformity may be undermined by the phenomenon known as regression to the mean (RTM)34,45, that occurs when an extreme measurement at time 1 is more likely to be followed by a less extreme value at time 2. This makes a natural variation in repeated measures to look like an experimentally-induced change46.
In our study, we controlled for the possible confounding effect of RTM by entering the first rating as a covari- ate of no interest, as recommended by34 (although it has been held that this method may even underestimate the effect of social influence35).
Seminal findings from Moscovici and colleagues have shown that minorities may exert an influence on the majority47 (see48,49 for a recent review). Importantly, however, this form of influence is most effective when infor- mational social influence is at play50. Informational influence occurs when people rely upon others' behaviors and beliefs in order to be accurate1. Crucially, however, our participants performed subjective evaluation of attractive- ness, rather than accurate estimations of objective property of the stimuli (e.g. specific body features). In cases of so-called normative social influence, where opinion tasks are at stake, social identity51 seems to be fundamentally important48. This is consistent with our findings, where the conformity behavior is stronger among participants - NEH - who can identify with the normative group - the fictive gay group.
Unfortunately, the lack of a condition where a sexual majority group exerts the social influence prevented us from testing whether the social influence from out-group causes a decrease in the conformity behavior as com- pared to the ingroup in each group, thus limiting the scope of our findings.
It might be argued that our decision to split the sample into groups (NEH and Het) might be somewhat arbi- trary, since the NEH includes individuals who placed themselves in intermediate positions along the Kinsey scale. Our choice had a two-fold motivation. Firstly, the small number of participants who did not consider themselves as exclusively gay or exclusively heterosexual is too small to allow any reliable estimate (N = 3, one bisexual and two predominantly gay - only occasionally heterosexual). Secondly, from a social identity standpoint, bisexuals are definitively perceived - and likely to perceive themselves - as members of a sexual minority22.
Another potential weakness is that we had selected our best 108 stimuli on an original set of 128 on the basis of the ratings of a sample of heterosexual women but then run the study on a sample of heterosexual and NEH men. Importantly, however, a post-hoc validation of our stimuli on a sample of heterosexual men demonstrated a large overlap in terms of shared variance, agreement on the stimuli selection and in terms of agreement on the ratings. Furthermore, we found that the reliability of the ratings within the female group was fair, whereas the reliability of the ratings within the male group was poor, a result that further justifies our prior selection of the stimuli. It is also important to emphasize that this selection procedure was aimed at alleviating floor effects on the ratings that could have biased the results because of the RTM phenomenon34,35,45,46.
Our behavioral results are in line with the findings summarized in the meta-analysis from Bond and Smith3 who found that when the influencing group is perceived as an out-group, social conformity decreases, although this effect was not statistically significant when controlling for other relevant factors (e.g. culture, gender, date of the study). Moreover, a recent study found that, similarly to what found in our study, conformity was attenuated when influence was exerted by an out-group (e.g., "Trump supporters" when participants were Trump opponents)52. Conformity was even reversed when participants perceived the out-group as more threatening. In our study we did not find evidence for a blatant reversal of conformism, but rather an attenuation of conformism when influence came from an out-group sexual minority. This might be due to the fact that very few of the participants perceived
303,122 students got unstuck by Course
Hero in the last week
Our Expert Tutors provide step by step solutions to help you excel in your courses