Validation of a survey instrument to assess home environments for physical activity and healthy eating in overweight children

Int J Behav Nutr Phys Act. 2008 Jan 11:5:3. doi: 10.1186/1479-5868-5-3.

Abstract

Background: Few measures exist to measure the overall home environment for its ability to support physical activity (PA) and healthy eating in overweight children. The purpose of this study was to develop and test the reliability and validity of such a measure.

Methods: The Home Environment Survey (HES) was developed to reflect availability, accessibility, parental role modelling, and parental policies related to PA resources, fruits and vegetables (F&V), and sugar sweetened drinks and snacks (SS). Parents of overweight children (n = 219) completed the HES and concurrent behavioural assessments. Children completed the Block Kids survey and wore an accelerometer for one week. A subset of parents (n = 156) completed the HES a second time to determine test-retest reliability. Finally, 41 parent dyads living in the same home (n = 41) completed the survey to determine inter-rater reliability. Initial psychometric analyses were completed to trim items from the measure based on lack of variability in responses, moderate or higher item to scale correlation, or contribution to strong internal consistency. Inter-rater and test-retest reliability were completed using intraclass correlation coefficients. Validity was assessed using Pearson correlations between the HES scores and child and parent nutrition and PA.

Results: Eight items were removed and acceptable internal consistency was documented for all scales (alpha = .66-84) with the exception of the F&V accessibility. The F&V accessibility was reduced to a single item because the other two items did not meet reliability standards. Test-retest reliability was high (r > .75) for all scales. Inter-rater reliability varied across scales (r = .22-.89). PA accessibility, parent role modelling, and parental policies were all related significantly to child (r = .14-.21) and parent (r = .15-.31) PA. Similarly, availability of F&V and SS, parental role modelling, and parental policies were related to child (r = .14-36) and parent (r = .15-26) eating habits.

Conclusion: The HES shows promise as a potentially valid and reliable assessment of the physical and social home environment related to a child's physical activity and eating habits.