Dichotomous Dependent Variable
Next we consider studies in which the dependent variable is dichotomous. In such cases there are at least two important issues that lead us away from using ordinary least squares estimation and a linear model. First, because Y can take on only two values (such as 0 and 1) it is illogical to expect that a line provides a good model. A linear model will produce estimates that reach beyond the 0-1 limits. Second, because all observed values of Y will always be either 0 or 1, it is not possible that the random error terms in the model will have a normal distribution or a constant variance.
With a dichotomous dependent variable, one approach is to use logistic regression.[2]. Estimating a logistic model is straightforward ...
Get Practical Data Analysis with JMP® now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
