Probit analysis or probit regression models are used to analyze data with binary response variables. The reason that I have included probit models in this course is that there are several multivariate extensions to probit analysis that make it appropriate for a multivariate course.

The intention here is to present a series units for binary response variables that parallels the the units on regression for continuous response variables.

Binary response varaibles can also be analyzed with logistic regression. However, there are no commonly available multivariate logistic software. Afifi and Clark include logistic regression in their Computer-Aided Multivariate Analysis book. In the univariate case both the logistic regression and probit regression solutions are very similar to one another.

Multivariate Course Page