Test-retest reliability and validity of the feeding your preschooler questionnaire for low-income Hispanic populations

Access full-text files

Date

2009-08

Authors

Loyo, Jennifer Joleen

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

This methods-oriented dissertation focuses on the psychometric evaluation of the Feeding Your Preschooler Questionnaire (FYPQ) designed to assess the eating habits and diet quality of young children. Parental proxy reports (n=135) were obtained through pen and pencil administration of the FYPQ and an in-person interview using a 24- hour food recall (24HR).Test-retest reliability (n=82) was determined using a repeated measures design with Wilcoxon signed rank tests and Spearman correlations for the food frequency questionnaire (FFQ) portion of the survey, parental self-efficacy, parental role modeling, parental practices, and parental perception of child food preference regarding fruits and vegetables scales. Test-retest reliabilities ranged from r=.53 for water to r=.84 for vegetables for the FFQ and from r=.64 for role modeling to r=.71 for parental perception of child preference for the psychosocial measures. Concurrent construct validity (n=107) was examined with a cross-sectional study design using the Wilcoxon signed rank test, Spearman correlations, and cross-classification analysis into quartiles of food group intakes. Spearman’s correlations between the FFQ and the 24HR were .46 for milk, .22 for fruit, .22 for vegetables, .11 for grains and .07 for protein. Cross-classification analysis revealed that 29% of children were classified in the same quartile and 69% in the same or within one quartile, and gross misclassification ranged from 2% to 10%. Nomological validity was examined using weighted least squares regression. Two regression analyses with fruit and vegetable intake on first the FFQ and second the 24HR as the dependent variable examined the influences of psychosocial environmental predictors and food insecurity. The FFQ regression model explained 28% (p<.05) of the variance in fruit and vegetable intake, with the significant predictors of parental role modeling and food insecurity. The 24HR regression analysis predicting fruit and vegetable intake explained 11% (p<.05), with parental perception of child preference and parental role modeling as significant predictors. In summary, the FYPQ demonstrated good test-retest reliability. The study provides evidence of concurrent validity for the FFQ for assessment of milk consumption and fruit and vegetable variety in preschool children's diets and of nomological validity in the prediction of fruit and vegetable consumption.

Description

text

LCSH Subject Headings

Citation