Parental versus child reporting of fruit and vegetable consumption

BACKGROUND The purpose of this study was to (1) compare parental and child recording of children's fruit and vegetable (F&V) consumption, including family-related factors, and (2) investigate the potential differences in the relation of children's and parents' perceptions of family-related factors. METHODS Children were recruited from Dutch seventh and eighth grade classrooms. Each child and one of their parents completed parallel questionnaires. A total of 371 matched child-parent surveys were included in the analyses. To compare parental and child reports of consumption and family-related factors regarding F&V intake several techniques were used such as paired sample t-test, chi-square tests, Pearson's correlations and Cohens's kappa as measurement of agreement. To investigate potential differences between the parent's and children's perceptions of family-related factors, linear regression analyses were conducted. RESULTS The results indicated weak agreement for F&V consumption (Cohen's kappa coefficients of .31 and .20, respectively) but no differences in mean consumption of fruit at the group level. Regarding the family-environmental factors related to fruit consumption, significant differences were found between the perceptions of subjective norm, and the availability and accessibility of fruit. Perceptions of subjective norm, parental modelling and exposure regarding vegetable consumption were also viewed differently by the two groups. The family-environmental factors reported by the children were similarly associated with F&V consumption compared to those reported by their respective parents. However, parents rated these factors more favourably than their children did. CONCLUSION The results indicated a low level of agreement between parental and child reporting of F&V intake and their assessment of family-environmental factors on individual level. This has important implications for the development and evaluation of interventions and we recommend that researchers clearly indicate which source of information they use in their studies as it remains unclear which source is more valid.However, when the effects of interventions are studied at the group level, our results suggest that it makes no difference whether children or parents report the child's fruit consumption. The same holds for determinant studies; both parental and child reports can be used. However, perceptions of these factors differ significantly.