Approaches for series correlation test and parameter estimation
Qiuhai Zhong · Journal of Shenzhen Institute of Information Technology · 2006
This paper focuses on the problem of series correlation in the econometric models. Firstly, it explains reasons why series correlation exists and the influences on the models' estimation parameters when using ordinary least squares (OLS) . Secondly, test methods are illustrated for series correlation, which fall into two types:false-correlation test and true-correlation test. And the merits and demerits of the above-mentioned methods are presented respectively. Thirdly, this paper discusses the estimation method, generalized least squares (GLS) , and shows how to get correlation coefficient. Finally, methods above are compared and summarized, and research trends are indicated.