(PS2-A7) Using Combined Updated Cutoff Scores on the Social Phobia Inventory and Sheehan Disability Scale Enhances the Identification of Analogue Participants with High Social Anxiety

Social anxiety disorder (SAD) is characterized by a yearly prevalence rate ranging between 6.8­­-7.1% (Kessler et al., 2005). In an effort to enhance the accuracy of capturing this prevalence in analogue samples, this study introduces updated cutoff scores for the Social Phobia Inventory (SPIN; Connors et al., 2000) and the Sheehan Disability Scale (SDS; Sheehan, 1983), advocating for their combined utilization. The SPIN assesses trait social anxiety (SA) and is frequently employed to categorize participants as having high SA, with cutoff scores historically spanning 19-30. Analysis of a substantial undergraduate sample (N = 1096) revealed that a cutoff of 30 identified 56.2% of participants as having high SA. The SDS gauges functional impairment arising from a specified mental disorder, and in SAD, means range from 17.1-19.2 (Sheehan and Sheehan, 2008), suggesting that the prior recommended cutoff score of 5 may be inadequate for this population.

In comparisons between community-based participants with an SAD diagnosis (N = 941) and non-clinical and clinical control groups, ROC curve analyses for the SPIN indicated an optimal cutoff score of 28.5 when non-anxious non-clinical controls were the reference group. However, this cutoff score increased to 37.5 when a non-SAD clinical control sample was used. For the SDS, the optimal cutoff was 3.5 with non-anxious non-clinical controls as the comparison group, rising to 13.5 when using the clinical control group. Logistic regression analyses demonstrated that adding the SDS to the model with the SPIN significantly increased the odds of distinguishing individuals with SAD from those in the non-clinical control group.

The study proposes the combined use of the SPIN and SDS for identifying participants with high SA. The newly determined cutoff scores will be applied, both individually and in combination, to select individuals with high SA from five representative undergraduate samples spanning 2018 to 2023. Descriptive information outlining this sampling process will be presented during the poster session. These findings aim to enhance the accuracy of selecting analogue samples of individuals with high SA from community and undergraduate populations.