Applied Linear Regression ModelsProviding a blend of theory and application, this text emphasizes understanding concepts, which are demonstrated by means of worked-out examples. The theoretical side enables applications of regression analysis to be carried out successfully. Computer and graphic plots focus the text on analysis and models, and discussion of regression diagnostics includes the DFBETAS, DFFITS, and PRESS measures. Model building is examined in order to allow students to see how the model-building process integrates many of the elements considered in earlier chapters. This edition includes new topics, such as: robust tests for constancy of the error variance, smoothing techniques to explore the shape of the regression function, robust regression and non-parametric regression techniques. |
Contents
Inferences in Regression Analysis | 44 |
2 | 83 |
Diagnostics and Remedial Measures | 95 |
Copyright | |
18 other sections not shown
Common terms and phrases
appropriate b₁ B₁X B₂ Bonferroni calculations column confidence interval Cook's distance data set decision rule degrees of freedom denoted diagnostic error sum error terms error variance estimated regression coefficients estimated regression function estimated standard deviations expected values explanatory variables extra sum family confidence coefficient Figure fitted regression function fitted values Ŷ Hence interaction effects least squares estimates linear regression function linear regression model logistic regression mean response mean square measure multicollinearity nonlinear regression normal error normal probability plot Note observations ordinary least squares outlier outlying P-value parameters percent confidence interval prediction interval Problem procedure random Refer regression analysis regression coefficients regression line regression model 2.1 regression relation residual plot response function response variable robust regression simple linear regression ẞ₁ SSTO standardized regression subset sum of squares Table test statistic transformation variance-covariance matrix vector weighted least squares X₁ X₂ Y₁