Invariance of the Household Food Insecurity Access Scale Across Different Groups of Adolescents and Young Adults

Abstract

Background: Cross-group comparisons of household food insecurity and its associations using multiple-item scales assume that scale scores can be interpreted as identical across groups. However, scores should not be interpreted as identical across groups without evidence of measurement invariance. Noninvariant measures indicate that the underlying construct may be different across groups. Objective: To determine whether the Household Food Insecurity Access Scale (HFIAS) is invariant across different groups of Ghanaian and South African youth aged 15 to 24. Methods: We analyzed cross-sectional quantitative data from 1437 and 4165 young South Africans and Ghanaians, respectively. Multi-group confirmatory factor analysis was used to examine whether the HFIAS was invariant across different groups of youth, including sex (male or female), age group (middle adolescence, late adolescence, or emerging adulthood), and receipt of child support grant (yes or no). We assessed 3 levels of invariance: configural, metric, and scalar. The model fit between nested models was compared using χ2 difference testing. Results: Invariance tests indicated that the HFIAS had configural, metric, and scalar invariance across different groups of Ghanaian and South African youth. Model fit statistics across all invariance levels indicated good fit of our hypothesized model with the observed data. χ2 difference testing results were not statistically significant across all nested models. Conclusions: Food insecurity, as measured by the HFIAS, meant the same thing for different groups of Ghanaian and South African youth. Evidence of invariance means that the HFIAS scores could be interpreted as identical across youth groups in our study.