BAYESIAN INFERENCE OF BINARY LOGISTIC REGRESSION MODEL FOR ASSESSING ERYTHROCYTES SEDIMENTATION RATE

Abstract

Binary logistic regression model allowed the prediction of a discrete outcome such as group membership, from a set of explanatory variables that may be continuous, discrete, dichotomous or a mix of any of these. Binary logistic regression model has many applications in various fields such as agriculture, transport, education, health and so on. In this study, binary logistic regression model was employed to determine the health status of an individual with the levels of the plasma proteins: fibrinogen and globulin considered. An informative prior probability distribution, a natural conjugate prior, was adopted to estimate the parameters of the binary logistic regression model using Bayesian approach. The model’s fit was assessed with the posterior predictive p-value approach in which the model with the highest percentage of correct classification of 87.5% was selected as the best model. The model’s adequacy was evaluated with the hit rate goodness of fit test in which the observed hit rate was greater than the criterion hit rate. Therefore, the model was significant since the model’s estimates fitted the data at an acceptable level.