(PS4-40) Examining the Discriminant Validity of Commonly Used Autism and Social Anxiety Screening Measures in Undergraduate Students

An extensive body of research has demonstrated that autistic traits and social anxiety disorder (SAD) symptoms significantly overlap, making it difficult to distinguish between the two conditions and provide an accurate and timely diagnosis (White et al., 2014). For some individuals, an autism diagnosis may not be established in childhood, underscoring the need for adult autism screening measures that can adequately differentiate autism from other conditions with similar features. The Social Responsiveness Scale – Second Edition (SRS-2; Constantino & Gruber, 2012) and Autism Spectrum Quotient (AQ; Baron-Cohen et al., 2001) are frequently used measures of autistic traits in clinical practice. Likewise, the Liebowitz Social Anxiety Scale (LSAS; Liebowitz, 1987) and Fear of Negative Evaluation Scale – Brief Version (BFNE; Leary, 1983) are commonly used SAD screening measures. Given the significant overlap in features of autism and SAD, the discriminant validity of autism and SAD measures should be investigated. The current study examined the degree to which the SRS-2 Adult Self-Report Form and AQ correlate with the LSAS and BFNE in a non-clinical sample of 215 undergraduate students (M_age = 18.93 years, SD = 1.20).

A strong, significant correlation was found between the SRS total raw score (M = 61.78, SD = 26.87) and the LSAS total score (M = 53.21; SD = 23.68; r = 0.58, p < .001). A significant, albeit small, correlation was found between the SRS total raw score and BFNE score (M = 40.35, SD = 13.2; r = 0.27, p < .001). There was a moderate, significant correlation between the AQ total score (M = 20.59; SD = 6.31) and LSAS total score (r = 0.43, p < .001). The AQ total score was not significantly correlated with the BFNE total score (r = .01, p = 0.2).

Previous findings have demonstrated the poor discriminant validity of the SRS and AQ from anxiety measures in clinical samples (Richey et al., 2014; South et al., 2017). The current study extends this finding in a non-clinical sample, further calling into question the discriminant validity of the SRS and AQ as autism screening measures. In the current sample, traits assessed on the SRS overlapped to a greater degree with social anxiety measures than the AQ. In addition, the BFNE was less correlated with the autism measures than the LSAS, indicating that the BFNE measures a more central feature of social anxiety that is less shared with autism. These findings suggest that the use of the Autism Spectrum Quotient and Fear of Negative Evaluation Scale – Brief Version may demonstrate greater utility in discriminating autistic traits from social anxiety symptoms. The findings of the current study highlight the need for the development of autism and social anxiety screening measures that assess core features unique to each condition.