作者: Hamid Habibagahi , John L. Pratschke
DOI: 10.2307/1926279
关键词:
摘要: where X1 represents a series of T observations (t = 1, . .. , T). This ratio is not appropriate for testing the independence residuals from least squares regression; its purpose to test autocorrelation among observed variables. The Durbin-Watson development (1950, 1951) von Neumann was concerned with regression serial correlation. They null hypothesis H,, that there no correlation between (Ut) against alternative H1 Ut are positively correlated. statistic