A COMPARISON OF BINARY LOGIT AND PROBIT MODELS WITH A SIMULATION STUDY

Logit and probit models which widely used are members of the family of generalized linear models. Particularly, when the dependent variable is binary, both models may be used for the estimation of the functional relationship between dependent and independent variables. Since those models are utilized for the same purposes, the question of which model performs better comes to the mind. For this intention, a Monte Carlo simulation was carried out to compare both the binary probit and logit models under different conditions. In data generation stage, by employing latent variable approach, different sample sizes, different cut points, and different correlations between dependent variable and independent variables were taken into account. To make a comparison between logit and probit models, residuals, deviations and different Pseudo-R squares which are used for qualitative data analysis, were calculated and the results were interpreted.