Bayesian Bisector Regression Method and an application

The purpose of this study is to examine the performance of Bayesian Type II regression Analysis. With this purpose, simple linear regression and bisector regression equations are calculated by the help of Bayesian approach on a real data set. Because in the earlier studies its mentioned that the best technique among Type II regression techniques is the Bisector regression technique, regression equations are obtained by considering the X and Y variables as the dependent variable respectively and then the Bayesian Bisector equation is calculated by bisecting these two regression lines. Different sample sizes are considered to obtain the posterior distribution based on prior and likelihood information and then the performances of Bayesian regression equations are compared according to MSE and RE criteria. The results of the study indicates that performance of Bayesian bisector technique has the minimum MSE for the sample sizes n=100, n=75 and n=50 which means that the performance of Bayesian bisector technique is the best for these sample sizes for the related data set.