Probability & Mathematical Statistics S.
2 total works
This study integrates old and new procedures for analyzing the problems of ill-conditioning (which include collinearity) that attend statistical data. It aids in understanding how these problems affect statistical estimation and provides suggestions and examples of corrective action. The approach emphasizes diagnostics, drawing on elements of statistics, econometrics, data and numerical analysis. Any of the analytical and geometric tools that are needed are developed and explained. Numerous illustrations and data sets are included as a source for user experiments. The diagnostics featured in the text deal with situations where two explanatory variables vary in much the same way, so it is not possible to separate their effects.
Provides practicing statisticians and econometricians with new tools for assessing quality and reliability of regression estimates. Diagnostic techniques are developed that aid in the systematic location of data points that are unusual or inordinately influential, and measure the presence and intensity of collinear relations among the regression data and help to identify variables involved in each and pinpoint estimated coefficients potentially most adversely affected. Emphasizes diagnostics and includes suggestions for remedial action.