University of Massachusetts Amherst
	Department of Mathematics and Statistics
	Statistics and Probability Seminar

Abstract: Food insecurity is a status of "limited or uncertain availability of nutritionally adequate and safe foods or limited or uncertain ability to acquire acceptable foods in socially acceptable ways" (USDA 1990). And a number of national surveys include questionnaires to measure food insecurity: the Current Population Survey, the Panel Study of Income Dynamics, the National Health and Nutrition Survey, the Early Childhood Longitudinal Survey, and more. I will briefly describe important concepts to understand food insecurity in US and current research progress in the area. Then I will discuss several statistical challenges in studying food insecurity in US using complex survey samples. These challenges include 1) ambiguity in what food insecurity measures compared to more traditional measures in poverty, hunger, or malnutrition, 2) selection bias by neglecting transitional housings or emergency shelters in household surveys, 3) lack of generalization of inference theories for statistical models to complex survey settings. For example, inference results of mixed effects model is not well established for complex surveys. I will then present ideas to address these challenges. I will demonstrate these ideas using aforementioned survey data sets. Except for some parts, the talk is for general audience and is not intended to be technical.