Abstract

Probit and logit models are two commonly used techniques for regressing certain independent variables on a dichotomous dependent variable. Through this paper we propose an alternative regression model to the probit and logit models, based on the generalized logistic distribution of Balakrishnan and Leung (Communications in Statistics – Computation and Simulation, 1988). We discuss the maximum likelihood estimation of the parameters of the model and illustrate its usefulness with the help of a real life data set.

Author: C. Satheesh Kumar and L. Manju

Received on: January, 2015

Accepted on: December, 2018