Multiple Regression Approaches

This book provide the practical ways for solving The mulltiple regression model for normality of errors, homogeneity of error variances and independence of serial correlation (no auto correlation). Normality of error can be tested through histogram of residual, normal probability plot and Jarqua-Bera (JB) test and the homogeneity of errors will be tested through Spearman‟s Rank Correlation Test and “Durbin- Watson d statistics” was used for detecting serial correlation .