Examining the Invariance of a Common Household Food Insecurity Measure Across Various Adult and Child Characteristics

Area of Science:

• Public Health
• Sociology
• Nutrition Science

Background:

• Food insecurity is a significant public health concern affecting various population segments.
• Accurate measurement of food insecurity is crucial for effective policy and intervention.
• The 10-item Household Food Security Survey Module (HFSSM) is a widely used tool for assessing food insecurity.

Purpose of the Study:

• To evaluate the construct validity of the HFSSM.
• To determine if the HFSSM yields equivalent measurements across different demographic groups (child sex, race/ethnicity, adult sex, relative status).
• To assess the suitability of the HFSSM for cross-group comparisons in food insecurity research.

Main Methods:

• Cross-sectional study utilizing data from the 2019 National Health Interview Survey child survey (n=1,575).
• Confirmatory factor analysis (CFA) was employed to assess the HFSSM's factor structure.
• Multiple group CFA was conducted to test for measurement invariance across specified respondent characteristics.

Main Results:

• A one-factor model demonstrated the best fit for the HFSSM data.
• The HFSSM exhibited measurement invariance across child sex, racial and ethnic groups, and relative status.
• Noninvariance was detected for the adult sex variable, indicating differential item functioning.

Conclusions:

• The HFSSM is a valid measure for cross-group comparisons concerning child sex, race/ethnicity, and relative status.
• Differences in interpretation of HFSSM items by adult males and females necessitate cautious use and advanced statistical modeling for sex-based comparisons.
• Future research should incorporate methods to address the observed noninvariance related to adult sex when utilizing the HFSSM.