ABSTRACT
Regression analysis, a common technique, is used in many scientific disci-
plines. These may be linear or nonlinear models or a complex set of models
such as the structural equation models, two-stage least squares, censoring, etc.
The multiple imputation analysis framework is the same: perform the analy-
sis on each completed data set, extract the meaningful parameter estimates,
their covariance matrix, test statistics and combine them using the methods
described in Chapter 4. This chapter covers some key issues about imputa-
tion and analysis in the context of a regression analysis and provides some
additional combining rules. These are specific, to regression analysis such as
ANOVA, Partial F-test and R2 and Adjusted R2 analysis.