Abstract

Based on the view that assessment of automatic thoughts is integral to cognitive theory and practice of psychopathology, the present studies were undertaken to develop the Speech Anxiety Automatic Thoughts Questionnaire (SAATQ) and to examine its reliability and validity. In study 1, a total of 83 negative automatic thoughts were drawn from thought listings resported by social phobic clients and pilot subjects while experiencing speech anxiety, and then 32 items (SAATQ-32) were finally selected on the basis of factor analyses in twice on 415 undergraduate students' responses to the 83-item SAATQ. The SAATQ-32 had three factors labeled Guess of Audience Slight and Ridicule, Self-Reproach and Lowered Self-Efficacy, and Perfection Seeking and Question of Audience Responses', and the three-factor model was thought to be the best model in terms of goodness-of-fit and model interpretation. In study 2, all three factors had high values for the coefficient of congruence which assesses the similarity of factor pattern coefficients across two independent samples. Therefore, the three-factor model was cross-validated. On the contrary, the four-factor model wasn't cross-validated since two of four factors hadn't high congruence values. The SAATQ-32 was highly reliable in terms of internal consistency, split-half reliability, and test-retest reliability. And it had good concurrent, construct, and discriminant validities. In conclusion, the SAATQ-32 appears to be a highly reliable, valid measure to assess negative automatic thoughts associated with speech anxiety. Lastly, both the implications of the present studies and the utilization of the SAATQ-32 in the cognitive study and treatment of speech anxiety were discussed.