Scoring the Social Phobia and Anxiety Inventory: Comments on Herbertet al. (1991)

Abstract

The manuscript by Herbert, Bellack, and Hope (1991) entitled "Concurrent validity of the Social Phobia and Anxiety Inventory," presented additional psychometric data on this instrument. Specifically, concurrent validity data for the Social Phobia and Anxiety Inventory (SPAI) were reported. Also, the authors reached a number of conclusions regarding the manner in which the subscales of the SPAI should be used. We do not agree with the authors' conclusions in this regard and wish to point out what we believe are problems with their recommendations. First, Herbert et al. (1991) concluded that the SPAI social phobia subscale score is a slightly better measure of social phobia symptoms than the Difference score. This conclusion primarily was based on Herbert and co-workers' interpretation that the correlations between the SPAI social phobia subscale and the instruments used to establish its concurrent validity are larger than the correlations between the SPAI Difference score and these measures. However, no statistical test was performed on these data. Also, in several cases, the correlation coefficients presented in Table I of the study by Herbert et aL (t991) differ by only one or two points, making it highly unlikely that they are statistically significant. Furthermore, because the SPAI was developed specifically to measure social phobia, and the other instruments used by Herbert et al. (1991) to establish its concurrent validity were not, one needs to use caution in interpreting patterns of correlation.