Jackknife Kibria-Lukman estimator for the beta regression model

Abstract

The beta regression model is a flexible model, which widely used when the dependent variable is in ratios and percentages in the range of (0.1). The coefficients of the beta regression model are estimated using the maximum likelihood method. In cases where there is a multicollinearity problem, the use of maximum likelihood (ML) leads to problems such as inconsistent parameter estimates and inflated variance.In the presence of multicollinearity, the use of maximum likelihood (ML) leads to problems such as inconsistent parameter estimates and inflated variance. In this study, KL estimator and its jackknifed version are proposed to reduce the effects of multicollinearity in the beta regression model. The performance of the proposed jackknifed KL beta regression estimator is compared with ridge, Liu and KL estimators through simulation studies and real data applications. The results show that the proposed estimators mostly outperform ML, ridge, Liu and KL estimators.