Development of Reliable and Valid Questions to Assess Food Resource Management Behaviors in Adults With Limited Income

Objective: Develop reliable, valid questions to assess changes in food resource management (FRM) behav-iors in adults with limited incomes. Design: Questionnaire development using a mixed-methods approach: content validity (subject matter and curricula), face validity, temporal reliability (test-retest), sensitivity to change, and exploratory factor analy-sis (EFA). Setting: Community settings in 12 states. Participants: Convenience samples of English-speaking Expanded Food and Nutrition Education Program (EFNEP) or EFNEP-eligible adults: 105 (cognitive interviews), 181 (test-retest), 185 (sensitivity), and 389 (EFA) adults. Variables Measured: Behaviors related to FRM skills: planning, shopping, and budgeting. Analysis: Consistency and agreement in cognitive interviews and temporal reliability; sensitivity at posttest (paired t tests, Wilcoxon signed-rank tests); internal consistency of scales identified in EFA (Cronbach a). P < 0.5. Results: All questions had acceptable temporal reliability ranges for the intraclass correlation coefficient (0.48-0.74) and Spearman rank-order correlation (0.48-0.73). All questions were sensitive to change at posttest (P < 0.001). Planning and saving scales, revealed by EFA, demonstrated internal consistency (> 0.80 Cronbach a). Conclusions and Implications: The 9 FRM behavior questions have acceptable temporal reliability and content and face validity and can be used nationally by EFNEP to assess participants' self-reported behavior changes. Other nutrition programs with similar audiences and content could use these questions to measure changes in FRM behaviors.

Automated summary

This study aimed to develop reliable and valid questions to assess changes in food resource management behaviors in adults with limited incomes. The questionnaire development used a mixed-methods approach, including content validity, face validity, temporal reliability, sensitivity to change, and exploratory factor analysis. The results showed that the questions had acceptable temporal reliability and sensitivity to change, and can be used nationally to assess participants' self-reported behavior changes.