Reproducibility of the School-Based Nutrition Monitoring Questionnaire among fourth-grade students in Texas.

ABSTRACT

OBJECTIVE: To assess reproducibility of a School-Based Nutrition Monitoring (SBNM) questionnaire for fourth-grade students. DESIGN: Test-retest. SETTING: Fourth-grade elementary school classrooms PARTICIPANTS: Multiethnic fourth-grade students from 2 area school districts (N = 322). MAIN OUTCOME MEASURES: Reproducibility coefficients with time intervals between questionnaire administrations of 2 hours for the yesterday food choice and physical activity questions or 2 weeks for the nutrition knowledge and attitude questions. ANALYSIS: Spearman rank order correlation, percent agreement, and Kappa statistic calculated for each individual questionnaire item. RESULTS: Of 43 questions that were included in the reproducibility analyses, test-retest kappa values were moderate to high (greater than 0.40) for more than 90% of the questions (39 questions). Spearman coefficients were greater than 0.70 for 24 questions, whereas percent agreement exceeded 75% for 28 questions. CONCLUSIONS AND IMPLICATIONS The SBNM questionnaire showed good to excellent reproducibility for nutrition behavior questions, physical activity, weight behavior, and food selection skills overall. However, questions that assessed nutrition knowledge and attitudes were not as reproducible, probably resulting from learning effects over time. Results from the study indicate that this questionnaire can be easily administered in an elementary school classroom and can be used to measure nutrition behaviors among 9- to 10-year-old school children.