Internal Consistency, Factor Structure, and Concurrent Validity of the Treatment Evaluation Inventory

Abstract

The Treatment Evaluation Inventory (TEI) is a commonly used 19-item measure of treatment acceptability. It yields a total score and Patient Progress and Acceptability subscales. The internal consistency and factor structure of the TEI were assessed in two samples, 218 members of the general public and 131 dental students. The total and both subscales had very high internal consistency (alpha for total scale =.96 and .94 in the two samples; alphas for Patient Progress .95 and .92; alphas for Acceptability .90 and .86). Factor analysis yielded a two-factor solution that accounted for 64.4% and 54.0% of the variance, but which did not correspond to the two existing subscales. A second analysis examined the distribution of TEI scores and correlations with two other measures of treatment acceptability (a thermometer rating scale and a 5-point Likert-type scale) in a sample of 222 undergraduate students. All measures were negatively skewed. There were large, significant correlations between the full scale TEI and the other measures, so the full TEI has concurrent validity. However, the TEI did not meet Anastasi's (1976) criterion for independent and useful subscales, therefore only the full scale score should be used.