اختبارمشكلة الارتباط الذاتي في نموذج الانحدار الخطي المتعدد باستخدام تقنية حصينة

Linear regression is one of the important and commonly used methods in analyzing, estimating, and predicting the future values of the phenomena studied in many economic, medical, psychological, and other branches. The Ordinary Least Square method is one of the commonly used methods for estimating the linear regression equation. To apply the least squares (OLS) method, a set of assumptions must be verified. However, violating one of these assumptions leads to a group of problems, the most important of which are the problem of violating the normal distribution, the problem of multicollinearity, the problem of heterogeneity of variance, and the problem of autocorrelation between random errors. In this study, the focus was on the autocorrelation problem due to its responsibility for violating important properties of ordinary least squares (OLS) estimates. The Brosch-Godfrey test is the most widely used method to detect autocorrelation. However, recent studies have shown that this test is easily affected by high leverage points. In this paper, we proposed a new method for testing the hippocampal Broch-Godfrey that is resistant to high leverage points. The performance of the proposed method was compared with existing methods using a set of real data as well as a simulation study. The results of the study showed that the proposed Brosch-Godfrey test is very powerful in detecting the problem of autocorrelation with and without the presence of high leverage points.