TWO ESSAYS ON THE APPLICATION OF ORDERED PROBABILITY MODEL

This thesis includes two essays on health issues of mental health and household food insecurity, and how socioeconomic, demographic, and other factors may influence these health issues are explored. The first essay investigates the effects of regular physical activity and socio-demographic factors on depressive symptoms for both men and women. Data for this study come from the 2011 Behavioral Risk Factor Surveillance System (BRFSS) and an ordered probability model with binary endogenous physical activity is developed to accommodate the ordinal nature of depression outcomes. Results suggest that physical activity is most beneficial for mild and moderate depressed individuals and the effect of regular physical activity is most notable on mild depressed females. In addition, socio-demographic factors are found to vary significantly between gender, and factors of age, income, race, education, employment status and recent mental health condition play important roles in affecting depressive symptoms. With data from the 2010 and 2011 Current Population Survey (CPS), the second essay investigates the effectiveness of the Supplemental Nutrition Assistance Program (SNAP) in reducing household food insecurity with a simultaneous equation model among husband-wife families with children (HW-C). Parental resource variables are used to better explore the effects of HW-C’s SNAP participation on FI. Our results suggest the participation of SNAP can reduce the probability of being food insecure among adults only (FIA) by 4.2%, but increases the probability of being low food security among children (LFSC) by 3% and increases the probability of being very low food security among children (VLFSC) by 1.2%. Parental resource variables and socio-demographic variables are also found to play important role in determining household food insecurity.