A pilot examination of the inter-rater reliability of the 18-item Household Food Security Module between cohabiting mothers and fathers

Abstract

Food insecurity, defined as the inability to access sufficient food for an active, healthy life, affects 11.1% of the US population and is primarily assessed using the United States Department of Agriculture's (USDA) 18-item Household Food Security Survey Module (HFSSM). The HFSSM is a self-report measure presumed to represent all household members, but it is unknown if cohabiting parents report food security differently. This pilot study aimed to determine (i) the inter-rater reliability of the HFSSM; (ii) the direction of any difference between responses to the HFSSM; and (iii) the item-level response similarities and differences among mothers and fathers of young children. Twenty-five pairs of low-income, food-insecure cohabiting parents of 2.5- to 10-year-old children participated in cognitive interviews assessing their level of food security using the USDA's HFSSM and completed related questionnaires. Intraclass correlations were computed to compare the responses of each dyad on the HFSSM overall and by item. Results revealed that overall report of food security was significantly but weakly correlated (r = .40, p = .02) within dyads. The majority of fathers (60%) reported higher food security scores on the HFSSM than their respective female partners. Furthermore, item-level intraclass correlations revealed that some HFSSM questions had poor reliability between cohabiting parents. This research identifies that a potential shortcoming of the HFSSM is the under-reporting of food insecurity by fathers compared with mothers within the household. These findings have implications for the utility of this measure used in national monitoring of the nation's food security.

Keywords: assessment; food security; gender; household; low-income population; parents.