TY  - JOUR
KW  - st0387
KW  -  smcfcs
KW  -  multiple imputation
KW  -  substantive model compatible
KW  -  congenial
KW  -  interactions
KW  -  nonlinearities
ID  - discovery1470042
N2  - Multiple imputation (MI) is a practical, principled approach to handling missing data. When used to impute missing values in covariates of regression models, imputation models may be mis-specified if they are not compatible with the substantive model of interest for the outcome. In this article we introduce the smcfcs
command, which imputes covariates by substantive model compatible fully conditional specification (SMC{FCS). This modifies the popular FCS or chained equations approach to MI by imputing each covariate compatibly with a
user-specified substantive model. The smcfcs command is compared to standard FCS imputation using mi impute chained
in a simulation study and illustrative analysis of data from a study investigating time to tumour recurrence in breast cancer.
TI  - Multiple imputation of covariates by substantive-model compatible fully conditional specification
EP  - 456
AV  - public
Y1  - 2015/06//
JF  - The Stata Journal
A1  - Bartlett, JW
A1  - Morris, TP
SN  - 1536-867X
UR  - http://www.stata-journal.com/article.html?article=st0387
N1  - © Copyright 2001?2015 StataCorp LP
IS  - 2
SP  - 437
VL  - 15
ER  -