Psychometric properties of the perceived therapeutic efficacy scale for adhering to a cholesterol-lowering diet

Background: Outcome expectancy may play an important role in behavior change. Previous studies tested the validity and the reliability of the Perceived Therapeutic Efficacy Scale (PTES), a scale that measures outcome expectancy related to adhering to a cholesterol-lowering diet. Further study was needed to examine its psychometric properties in a larger sample. Objective: The aim of this study was to test the psychometric properties of the 10-item PTES in a large sample. Methods: The PTES and the Connor Diet Habit Survey were administered to adults enrolled in a cardiac rehabilitation program. Results: The final sample for the analysis (N = 224) was, on average, 69.35 years old and was predominantly men (66.50%) and white (92.40%); nearly all (96.00%) completed high school. The inter-item correlation matrix revealed that correlation coefficients were greater than 0.80 between 4 pairs of items, suggesting that the 4 items were redundant. After consulting with a content expert and an examination of item content, we removed the 4 redundant items (items 2, 3, 4, and 10) and reduced the scale to 6 items. Principal component analysis revealed a 1-factor scale with high loadings for the 6 items, each greater than 0.70. The reliability of the scale, measured by Cronbach's α, was 0.91. The total PTES score had a moderate correlation with self-reported behaviors of adhering to a cholesterol-lowering diet, as measured by the Connor Diet Habit Survey subscale for cholesterol and fat intake (r = 0.36, P < 0.05). Conclusions: This study demonstrated that the revised and shortened PTES scale is reliable and valid to measure outcome expectancy related to adhering to a cholesterol-lowering diet.