Skip to main content
. 2008 Nov 27;10(2):335–351. doi: 10.1093/biostatistics/kxn041

Fig. 1.

Fig. 1.

General model for regression of y on x using a combination of data sets with different observed covariates. Circles represent unknown quantities and squares represent observed data. Covariates x(M1) missing in data set 1 are predicted from a regression fitted using the observed values of x(M1) in data set 2 and variables x(C) common to both. Covariates x(M2) missing in data set 2 are predicted in a similar way using information from data set 1.