Factorial Integrity of the Sport Anxiety Scale: A Methodological Note and Revised Scoring Recommendations

Results of several recent studies have called into question the factorial integrity of the Sport Anxiety Scale (SAS; Smith, Smoll, & Schutz, 1990) because two items (14 and 20) that loaded on the Concentration Disruption subscale in our validation and crossvalidation samples have loaded or cross-loaded on the Worry subscale in other samples. We agree that this is a serious problem, and we evaluate proposed modifications of the SAS on both conceptual and empirical grounds. For researchers currently using the SAS, we recommend a new scoring system that deletes two troublesome Concentration Disruption items and one Somatic Anxiety item, preserving separate and factorially consistent Somatic, Worry, and Concentration Disruption subscales. We present evidence that our original Worry scale provides a better CFA fit than a suggested "cognitive anxiety" alternative that combines the two concentration disruption items with the Worry scale items. We also describe nonreplication of the SAS three-factor structure in child samples and caution researchers against computing subscale scores for child samples.