Objective: To develop a self-administered risk questionnaire for common nutrition-related diseases in middle school students.
Methods: Two phases were conducted to develop the questionnaire: scale development and validation. Phase 1 included 7 steps: (1) determining the objective, theoretical framework, principles and format for indicator generation; (2) setting up the preliminary indicator pool; (3) selecting indicators and forming pilot questionnaire through focus groups; (4) testing the pilot questionnaire; (5) further correcting the questionnaire using expert consultation; (6) choosing indicators again using good-poor analysis; and (7) shaping the final questionnaire. Phase 2 consisted of: (1) using the Pearson correlation coefficient to assess test-retest reliability; (2) using the Cronbach's alpha coefficient to assess the internal consistency reliability; (3) using the feedback from field investigation to assess face validity; and (4) using explanatory factor analysis to assess construct validity. Students from 96 classes were selected at random in Hunan Province as the field test samples using stratified sampling and cluster sampling. And the students from 4 out of the 96 classes were chosen again to serve as the test-retest samples. We used Epidata 3.0 to build the database and SPSS 11.0 to analyze the data.
Results: A brief self-administered risk questionnaire for common nutrition-related diseases in middle school students with 12 items being formed after Phase 1. Good-poor analysis showed results from t tests for each item were statistically significant (P<0.05). The Pearson correlation coefficient was 0.76 (P<0.05) and the Cronbach's alpha coefficient was 0.56. The questionnaire was accepted by the students participating in the field test. Four common factors were extracted using explanatory factor analysis, accounting for 50.18% of the total variation.
Conclusion: The brief self-administered risk questionnaire for common nutrition-related diseases in middle school students is reliable and valid.