Skip to main content
ARS Home » Research » Publications at this Location » Publication #148937
Title: VALIDATION OF THE US HOUSEHOLD FOOD SECURITY MODULE IN THE LOWER MISSISSIPPI DELTA (LMD) REGION OF ARKANSAS, MISSISSIPPI, AND LOUISIANA

Author
item SIMPSON, PIPPA - ACHRI - DAC
item GOSSETT, JEFFERY - ACHRI - DAC
item HALL, RENEE - ACHRI - DAC
item CASEY, PATRICK - ACHRI
item Bogle, Margaret
item NORD, MARK - ERS,USDA

Submitted to: Journal of Federation of American Societies for Experimental Biology
Publication Type: Abstract Only
Publication Acceptance Date: 1/14/2003
Publication Date: 4/13/2003
Citation: SIMPSON, P.M., GOSSETT, J.M., HALL, R.A., CASEY, P.H., BOGLE, M.L., NORD, M. VALIDATION OF THE US HOUSEHOLD FOOD SECURITY MODULE IN THE LOWER MISSISSIPPI DELTA (LMD) REGION OF ARKANSAS, MISSISSIPPI, AND LOUISIANA. JOURNAL OF FEDERATION OF AMERICAN SOCIETIES FOR EXPERIMENTAL BIOLOGY. 2003. v. 17(4). p. A708.

Interpretive Summary:

Technical Abstract: Introduction: Food security (FS) has been shown to be an important factor in nutrition and health. In a predominantly low income area, such as the LMD, where food insecurity is twice as high as nationally, it is important to validate that the instrument used is indeed measuring the same phenomena as nationally and that information about FS has been reliably collected. Objective: To validate the 18-question US Household FS Module, using statistical methods based on the Rasch measurement model for use in the LMD. Methods: In a representative sample of 1659 households, 655 with children, who live in the LMD, we calculate average item discrimination, item severity scores and fit statistics, and compare these with corresponding statistics from the Current Population Survey (CPS) 1998-1999. Results: On average, the items discriminate at least as well as the CPS, with the severity scores agreeing well. This indicates that all items can be used. One question about cutting the size of a child¿s meal does not discriminate well, as shown by a large infit statistic (1.40). The large outfit statistic (7.6) for the least severe question about being worried that there would not be enough to eat shows that few people did not affirm this question while affirming other questions. Conclusion: The FS scale can be used with confidence in rural, low income populations such as the LMD. Supported by ARS/USDA Project #6251-53000-003-00D