BACKGROUND: Understanding factors influencing food choices is likely to enhance the effectiveness of strategies to promote healthy eating patterns. This article describes the process used to develop measures of psychosocial factors related to eating patterns in the Working Well project. METHODS: Working Well is a multicenter controlled trial of worksite health promotion interventions, including a nutrition intervention aimed at promoting low-fat, high-fiber eating. The process for developing measures included several steps. First, we defined three domains of psychosocial factors influencing dietary behavior: predisposing, enabling, and change-related factors. We then reviewed large-scale survey findings, compiled a catalogue of items, and developed and refined a 65-item questionnaire for pretesting in a working population. RESULTS: Based on frequency distributions and interitem correlations, the item pool was reduced to 24 items. The 24 remaining items were included in a pilot survey of 652 employees (response rate = 80%). On the basis of pilot data analyses, we eliminated one item and made minor modifications to other items. Factors most strongly associated with dietary intake were self-rated diet, past success at change, and motivation to eat low-fat foods. CONCLUSIONS: Analyses of the cross-sectional pilot data suggest directions for analyses of the final survey. The measures and the development process yielded an instrument and process that can be useful to other researchers.
RCT Entities:
BACKGROUND: Understanding factors influencing food choices is likely to enhance the effectiveness of strategies to promote healthy eating patterns. This article describes the process used to develop measures of psychosocial factors related to eating patterns in the Working Well project. METHODS: Working Well is a multicenter controlled trial of worksite health promotion interventions, including a nutrition intervention aimed at promoting low-fat, high-fiber eating. The process for developing measures included several steps. First, we defined three domains of psychosocial factors influencing dietary behavior: predisposing, enabling, and change-related factors. We then reviewed large-scale survey findings, compiled a catalogue of items, and developed and refined a 65-item questionnaire for pretesting in a working population. RESULTS: Based on frequency distributions and interitem correlations, the item pool was reduced to 24 items. The 24 remaining items were included in a pilot survey of 652 employees (response rate = 80%). On the basis of pilot data analyses, we eliminated one item and made minor modifications to other items. Factors most strongly associated with dietary intake were self-rated diet, past success at change, and motivation to eat low-fat foods. CONCLUSIONS: Analyses of the cross-sectional pilot data suggest directions for analyses of the final survey. The measures and the development process yielded an instrument and process that can be useful to other researchers.