A Novel Measure of Social Metaperception for Adolescents with and without High Functioning Autism: Reliability and Validity

Background: Our understanding of social cognitive processes is limited by retrospective, informant-based measurement. However, these dynamic internal processes may be better measured by self-reported ratings of impressions after a social interaction. Theory of mind may be necessary but not sufficient for this ability. A novel self-report measure was designed to capture an individual’s ability to perceive what a social partner thinks of him/her, or social metaperception (Laing et al., 1966). Scores on this measure will be examined in relation to theory of mind, narcissism, and loneliness as characteristics associated with other-perceptions, self-perceptions, and interpersonal relationships, respectively.

Objectives: The goal of this study was to examine reliability and validity of a novel measure of social metaperception in typically developing adolescents and those with high functioning autism (HFA). We hypothesized that metaperception accuracy would be positively associated with theory of mind and negatively associated with narcissism and loneliness.

Methods:  Preliminary analyses were conducted on eleven dyads comprising gender-, age-, verbal IQ-matched adolescents with HFA (9 males, M_age =13.69, SD=1.17) and unfamiliar comparison adolescents (COM; M_age=13.45, SD=1.26). Final sample size is estimated to comprise 20 dyads. Immediately following a five-minute unstructured interaction, each participant completed the Perceptions and Metaperceptions Questionnaire (PAMQ) indexing positive and negative perceptions of the social partner and social metaperceptions, or predictions of peer’s impressions. Participants completed assessments of theory of mind and questionnaires assessing narcissism and loneliness.

Results: The measure demonstrated good reliability, positive perceptions α=.93; negative perceptions α=.75; positive metaperceptions α=.94; negative metaperceptions α=.87.

Metaperception accuracy was calculated by separately standardizing positive and negative metaperception totals, and positive and negative perceptions from peers. Peer standardized perceptions were subtracted from target standardized metaperceptions to compute standardized difference scores (SDSs; De Los Reyes & Kazdin, 2004). SDSs further from zero indicate greater discrepancies (positive=overestimation, negative=underestimation). In addition, when interpretation of SDSs was not possible, absolute values were examined, with larger absolute values indicating greater discrepancies.

Across the full sample, absolute values were significantly different from zero, t(19)=8.47, p<.001, positive impressions and t(18)=6.60, p<.001, negative impressions, indicating discrepancies between metaperceptions and peer’s perceptions for both groups. HFA and COM adolescents did not differ on mean SDSs.

Bivariate correlations were used to examine validity of the PAMQ across the full sample. Surprisingly, SDSs were unrelated to theory of mind. As expected, SDSs related to narcissism, r(16)=.51, p=.046, positive words and r(15)=-.46, p=.08, negative words, reflecting an association between narcissism and both overestimation of peer’s positive impressions and underestimation of peer’s negative impressions. SDSs were related to loneliness, r(17)=-.61, p=.01 for positive words and r(16)=.59, p=.02 for negative words, indicating underestimation of positive impressions and overestimation of negative impressions.

Conclusions: Preliminary analyses demonstrate that the novel PAMQ is a reliable and valid social metaperception index for adolescents with and without HFA. PAMQ scores’ associations with narcissism and loneliness demonstrate that social metaperception relates to within-person attributes and interpersonal relationship characteristics. Final analyses will be conducted separately by group to examine differences. Future studies should include longitudinal analyses to parse direction of effects.