Analysis of Data for Measuring Food Availability, Access and Nutritional Status (Topic 17)

Policy choice depends crucially on an understanding of the causes of different food and nutrition insecurity profiles. It is important to know not only the causes but also their differential impacts in order to facilitate priorities in policymaking and allocation of funds for different programmes. For instance, it would be important to know the statistical relationship between nutritional determinants and outcomes.
This topic introduces regression analysis as a tool to address such issues. To begin with, it focuses on simple linear regression covering a range of topics like (i) basic concepts of regression analysis; (ii) specification -- assumptions of the simple linear regression model; (iii) estimation – ordinary least squares (OLS) method; (iv) the normality assumption and its implications; (v) prediction in the linear regression model; (vi) estimation of a poverty line; and (vii) how measurement of policy variables matters.