Qian Song,1,2,* Kaili Zheng,1,2,* Zixia Ding,1,2 Zhengmiao Miao,1,2 Zhaoxia Liu,1,2 Ming Cheng,3 Jinyao Yi1,2,4 

1Medical Psychological Center, the Second Xiangya Hospital, Central South University, Changsha, People’s Republic of China; 2Medical Psychological Institute, Central South University, Changsha, People’s Republic of China; 3Department of Clinical Psychology, Hunan Brain Hospital, Changsha, People’s Republic of China; 4National Clinical Research Center for Mental Disorders, Changsha, 410011, People’s Republic of China

*These authors contributed equally to this work

Correspondence: Jinyao Yi, Medical Psychological Center, The Second Xiangya Hospital, Central South University, Changsha, Hunan, 410011, People’s Republic of China, Email jinyaoyi@csu.edu.cn

Background: There exist four short forms of the Social Interaction Anxiety Scale (SIAS): 10-item version (SIAS-10), 6-item version by Peters et al (SIAS-6P), 6-item version by Fergus et al (SIAS-6F), and 5-item version (SIAS-5). This study aims to comprehensively examine the psychometric properties of the SIAS-10, SIAS-6P, SIAS-6F, and SIAS-5 and to determine which one performs relatively better in Chinese population.
Methods: This study enrolled 733 Chinese college students. The unidimensionality of the SIAS-10, SIAS-6P, SIAS-6F, SIAS-5 was examined using Confirmatory Factor Analysis (CFA). Multi-Group CFA was further adopted to assess measurement equivalence across gender. Internal consistency reliability and criteria-related validity were also evaluated. Additionally, the measurement performance of the SIAS-10, SIAS-6P, SIAS-6F, and SIAS-5 was assessed with Item Response Theory (IRT), which estimated the discrimination parameter and the Item Characteristic Curve (ICC) for each item.
Results: Except for the SIAS-5, the SIAS-10, SIAS-6P, and SIAS-6F displayed a good-fit to the one-factor model. Furthermore, the SIAS-10 achieved strict equivalence across gender while other versions did not. The SIAS-10, SIAS-6P, SIAS-6F and SIAS-5 all had acceptable internal consistency and significant correlations with criteria scales. The IRT results showed that the SIAS-10 included more items with higher discrimination and peaked ICCs (indicating more informative), whereas the SIAS-6P included more items with lower discrimination and flat ICCs (indicating less informative).
Conclusion: For assessing social anxiety in Chinese under constrained conditions, the SIAS-10 is recommended to clinicians for it measuring equivalently across gender, reflecting the relevant criteria variables well, and discriminating various levels of social anxiety sensitively.

Keywords: social anxiety, psychometric properties, measurement equivalence, item response theory, Chinese