Study on nutritional monitoring of Persian malnourished school students: A check on reproducibility and validity of questionnaire.

AIM: To develop a nutritional screening tool and examine its reliability and validity in elementary school children with malnutrition based on PRECEDE-PROCEED model, in Shiraz, Iran.

METHODS: A questionnaire was designed and its reproducibility was analysed via test-retest design among students with malnutrition. The students were in the fourth to sixth grades. Validity was assessed using face validity, content validity, construct validity, item difficulty and item discrimination. Reliability was measured by Cronbach's alpha (internal consistency), the intra-class correlation coefficient (ICC) and κ statistics (test-retest reliability).

RESULTS: A total of 165 primary school students (mean age of 11.21 ± 1.08 years, 65.5% females) participated in the study. The PRECEDE-PROCEED theoretical model for health education was adapted for this questionnaire. The questionnaire was divided into four main sections: predisposing, reinforcing, enabling factors and behaviour. Coefficient α reliability estimate was above 0.9 for most scales. The ICCs for different scales of the questionnaire were within the range of 0.402-0.800. Values for κ ranged from 0.412 to 0.785 for most scales.

CONCLUSIONS: The results showed reasonable validity, excellent reliability and moderate to substantial reproducibility for all scales of the questionnaire except two questions for κ statistics. The present study is of great importance, as there are no psychometrically evaluated tools (based on health behaviour models) with accepted validity and reliability for the evaluation of such factors in children with malnutrition. The questionnaire developed in the present study can be used to assess the effect of intervention among primary school students with malnutrition.