Objective: To evaluate the reproducibility and validity of the School-Based Nutrition Monitoring (SBNM) secondary level student questionnaire.
Design: Reproducibility was evaluated using a test-retest study design by administering the questionnaire on the same day (morning and afternoon, n=254) and, when appropriate, 9 to 14 days apart (n=259). Validity was assessed by comparing foods selected on the questionnaire with foods reported from a single 24-hour recall covering the same referent period (yesterday) in 209 students.
Subjects/setting: Eighth grade students in middle schools from central Texas were used: male and female, approximately 75% white, for the reproducibility study, with 15% Hispanic, 6% African American; for the validation study, approximately 38% white, 41% Hispanic, and 17% African American.
Statistical analysis: Spearman rank order correlation, kappa statistic, and percentage agreement were used for both reproducibility and the validation.
Results: For the reproducibility study, agreement for questions about food intake "yesterday" were 70% to 98%, with kappa statistics ranging from 0.54 to 0.93 and correlations between 0.66 and 0.97. Questions on recent physical activity had high agreement (66% to 89%) as did "usual" physical activity items, weight loss, and food selection behaviors. Nutrition knowledge items showed relatively weaker reliability: agreements ranged from 47% to 92%, with kappa statistics between 0.30 and 0.56. Attitude questions had weaker agreement (50% to 87%), kappa statistics (0.27 to 0.52), and correlations (0.33 to 0.63). For the validation study, correlation coefficients ranged from 0.32 for breads to 0.68 for milk and beans. Percentage agreement ranged from 38% for breads to 89% for gravy.
Applications/conclusions: Most questions on the SBNM secondary questionnaire were found to have acceptable reproducibility, whereas validation of food choice behaviors showed results similar to or better than other dietary assessment instruments for this age group. This questionnaire is a useful epidemiologic tool for surveillance, assessing broad intervention effects among groups or providing needs assessment data on selected nutrition and physical activity-related constructs.