Validated scales to assess adult decisional balance to eat more fruits and vegetables.

Abstract
A scale to measure adult decisional balance to eat more fruits and vegetables was developed and confirmed, and its psychometric properties were assessed. Two simple random samples of adults ages 25-60 years were selected from a nationally representative sampling frame. The development survey had a 72% response rate (n=231). The confirmation survey had a 67.4% response rate (n=2132). In both surveys, a self-administered questionnaire assessed demographics, fruit and vegetable intakes, stages of change, and decisional balance. Principal components analysis with varimax rotation and confirmatory factor analysis were performed. The decisional balance scale had three reliable subscales: « health pros, » « non-health pros, » and « cons. » Model fit was adequate for a « pros » and « cons » hierarchical structure. For both fruits and vegetables, health pros increased significantly between precontemplation and contemplation stages, surpassing the cons. Non-health pros increased significantly between precontemplation and contemplation fruit stages, surpassing the cons in preparation stage. Between precontemplation and action stages, health pros increased (mean effect size=0.90 [fruit] and 0.80 [vegetables]) and cons decreased (mean effect size=0.27 [fruit] and 0.35 [vegetables]). Heterogeneity in this sample may have diluted these effect sizes. This decisional balance scale is valid and reliable.