Regression with Imputed Covariates: A Generalized Missing Indicator Approach

47 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Weighted‐Average Least Squares (WALS): A Survey (2016, background): “Journal of Economic Surveys (2016) Vol. 30, No. 1, pp. 117–148 C© 2014 John Wiley & Sons Ltd Dardanoni et al. (2011, 2012) exploited WALS for handling the bias-precision trade-off that arises in the estimation of linear regression models with missing covariate values replaced by imputations.”
Bayesian Model Averaging and Weighted-Average Least Squares: Equivariance, Stability, and Numerical Issues (2011, unknown): “Dardanoni et al. (2011b) show that this bias-precision trade-oﬀ is equivalent to that arising in an extended regression model with two subsets of regressors: the focus regressors corresponding to the observed and imputed covariates, and the auxiliary regressors corresponding to all possible…”
Tuning multiple imputation by predictive mean matching and local residual draws (2014, background): “For β = 10 this initially appears surprising, but occurs because the strong association between y and x comes close to the assumption required for complete cases analysis to be valid, that the probability of xi being missing is conditionally independent of yi given xi [40].”