ABSTRACT
Continuous random variables include random variables which are referred to as possessing exponential, Gamma, Weibull, normal, and log-normal distributions. Rather than possessing a probability mass function, they distinguish themselves from discrete random variables by possessing a density. The point of balance then matches the center of gravity. The discrimination function plays a similar role to the regression function in binary regression. Anderson provides a medical diagnosis example in the bivariate case. This approach is particularly useful for investigating multiplicative interaction effects between two variables. The three-directional approach would best describe advocated methodology for many three-way contingency tables, since, by repeated applications of Goodman’s full-rank analysis for two-way tables, this approach frequently helps the statistician to infer the key real-life conclusions from the data.
