Patients with social phobia commonly resist face-to-face assessments, and a number of alternative assessment methods based on the internet are being developed. The aim of this study was to identify patients with social phobia on the internet and characterize their condition, using the Social Phobia Inventory (SPIN). In Stage I, this study recruited 1307 participants from the internet, most of whom were well-educated young females, who had remained unmarried and unemployed. The internet-based SPIN demonstrated excellent internal consistency (Cronbach's α = 0.937) and good test-retest reliability (intraclass correlation coefficient = 0.942). In Stage II, we examined the discriminant validity of SPIN via structured telephone interviews. The area under the receiver operating characteristic curve used to discriminate social phobia was 0.871 with an optimal cut-off point of 24 on the total score for SPIN. According to the SPIN scores, 919 of Stage I participants (70.3%) reached the threshold of social phobia, 531 of which (57.8%) had never sought professional help. These results suggest that the internet is a potential avenue through which to find untreated patients with social phobia.
