ABSTRACT
This chapter explores logistic regression which is the technique to use when the outcome variable is discrete taking two or more possible values. It utilizes contingency tables to help support and illustrate logistic regression. The chapter illustrates how to interpret and report simple outputs from SPSS and Stata for dichotomous, polychotomous and continuous independent variables. The command ‘margins’ in Stata is a useful way of estimating marginal effects. A marginal effect is defined as changes in responses for a change in a covariate. The chapter provides a range of additional examples to illustrate how to perform and report multinomial logistic regression with some dependent variable categories. The research literature suggests that aspirations are a progressive developmental process that builds from childhood into adulthood. The chapter reviews logistic regression output as B-values, EXP(B)-values and standard errors.
