A psychometric study of the State-Trait Anxiety Inventory

This study tested the hypothesis that the correlation between the STAI A-State and A-Trait scales is spuriously high because some of the items in these scales have low content saturation. As measured by Jackson's (1970) Differential Reliability Index, content saturation was found to be high for only 7 A-State and A-Trait items. It was demonstrated that new A-State and A-Trait scales based on these items with high content saturation had high internal consistency reliability and low intercorrelation, and were also very highly correlated with the original STAI scales. These results provided strong support for the tested hypothesis. Implications of these results were discussed for revising the STAI.