ABSTRACT
The purpose of the current studies is to identify safety behavior dimensions relevant to test anxiety, to develop a questionnaire to assess those dimensions, and to examine the validity of that questionnaire. Items were generated from interviews with college students ( N = 24). Another sample ( N = 301) completed an initial 33-item measure. Another sample ( N = 151) completed the final 19-item version the Safety Behaviors in Test Anxiety Questionnaire and provided access to their academic records. Interviews and expert evaluations were used to select items for the initial pool. An examination of item distributions and exploratory factor analysis were used to identify dimensions and reduce the item pool. Confirmatory factor analyses were used to validate the factorial structure. Correlational analyses were used to examine criterion validity of the final measure. The Safety Behaviors in Test Anxiety Questionnaire consists of a 9-item "Superstitious Behaviors" scale and a 10-item "Reassurance Seeking." The measure shows good content validity, factorial validity, internal consistency, and convergent and discriminant validity. Only the Reassurance Seeking scale showed good incremental criterion validity. Overall, these findings suggest that reassurance seeking may be a neglected target for interventions that might increase performance on high stakes tests.
