Identifying Analogue Samples of Individuals with Clinically Significant Social Anxiety: Updating and Combining Cutoff Scores on the Social Phobia Inventory and Sheehan Disability Scale

Abstract

The use of analogue samples, as opposed to clinical groups, is common in mental health research, including research on social anxiety disorder (SAD). Recent observational and statistical evidence has raised doubts about the validity of current methods for establishing analogue samples of individuals with clinically significant social anxiety. Here, we used data from large community samples of clinical and non-clinical participants to determine new cutoff scores of 34 on the Social Phobia Inventory (SPIN), a validated self-report measure of social anxiety symptoms, and a new cutoff score of 11 on the Sheehan Disability Scale (SDS), a validated self-report measure of social anxiety symptom-related impairment. We then examined whether using these newly determined cutoff scores alone or in combination improves the identification of individuals who have SAD from those who do not, revealing intriguing trade-offs in sensitivity and specificity and clear recommendations for the use of the new cutoff scores in combination with one another to facilitate future research. Finally, we compared the effects of our new cutoff scores with the original cutoff scores currently used in research on social anxiety by extracting analogue samples of participants with high social anxiety from historical data on seven large groups of undergraduate Psychology research participants from the University of Waterloo spanning the past five years (2018–2023). We observed that the new combined cutoff scores identified markedly fewer students as having high social anxiety, lending credibility to their validity and utility. We also observed a striking increase in levels of social anxiety symptoms in the undergraduate population from before to after the COVID-19 pandemic. Of note, most participants were under 30 and identified as Caucasian or Asian women, indicating that future research is needed to examine whether our findings generalize to diverse populations. Implications and future directions for social anxiety research are discussed.