A dichotomous variable, a special case of categorical variable, consists of two categories. The particular values of a dichotomous variable have no numerical meaning. A dichotomous variable can either be naturally existing or constructed by a researcher through recoding a variable with more variation into two categories. A dichotomous variable may be either an independent or a dependent variable, depending on its role in the research design. The role of the dichotomous variable in the research design has implications for the selection of appropriate statistical analyses. This entry focuses on how a dichotomous variable may be defined or coded and then outlines the implications of its construction for data analysis.