The Development of a Short Chinese Version of the State-Trait Anxiety Inventory

BackgroundFew studies on anxiety in China have used the full version of the Spielberger State-Trait Anxiety Inventory (STAI) due to its length, despite its numerous advantages. The goal of this study was to develop a short Chinese version of the STAI and examine its reliability and validity among the general Chinese population and psychiatric patients diagnosed with anxiety. MethodA total of 1,142 participants were invited to test the short Chinese version of the STAI; item analysis, validity testing and internal consistency reliability analysis were performed. Subsequently, 40 participants were enrolled to assess retest reliability 1 week later. Finally, 330 participants, including 33 psychiatric patients with anxiety, were used to assess the criterion and empirical validity. The Self-Rating Anxiety Scale (SAS) and Satisfaction with Life Scale (SWLS) were used as criteria, and receiver operating characteristic (ROC) analysis was conducted to evaluate the discrimination of the short version of the STAI between the groups with and without anxiety disorders. ResultThe short Chinese version of the STAI contains six items for each subscale (STAI-S-6 and STAI-T-6). The Pearson correlation coefficients between the two shortened Chinese versions of the STAI and the full-length STAI were 0.970 and 0.962, the Cronbach's alpha coefficients were 0.871 and 0.842, and the retest reliability values were 0.728 and 0.813 (p < 0.001). Confirmatory factor analysis showed that the 2-factor model achieved an adequate model fit: for the STAI-S-6, CFI = 0.986, TLI = 0.974, and RMSEA = 0.075, and for the STAI-T-6, CFI = 0.994, TLI = 0.988, and RMSEA = 0.052. The short Chinese version of the STAI had a significant positive correlation with the SAS score (r = 0.289, 0.313; p < 0.001) and a negative correlation with the SWLS score (r = -0.476, 0.554; p < 0.001). A significant difference in the level of anxiety was found between participants with and without anxiety disorders. The diagnostic accuracy of the STAI-S-6 and STAI-T-6 for major anxiety disorder was acceptable, with areas under the ROC curves of 0.798 and 0.745, respectively. ConclusionThe short Chinese version of the STAI demonstrates sound psychometric properties and is applicable in evaluating the level of anxiety in Chinese populations.