Discriminant analysis can predict group membership based on independent variables and identify characteristics that classify individuals. It works when the dependent variable is dichotomous or multicategorical and independents are metric. For example, it could classify athletes as basketballers or volleyballers based on anthropometric measures, or students as high or low performers based on skills. For accurate results, group sizes should not differ greatly, there should be at least five times as many cases as independents, variables must be normally distributed, and have linear and homoscedastic relationships. SPSS can be used to conduct a discriminant analysis to identify players' categories in a sports selection process.