The use of background data in the analysis of covariance
Eryl A. C. Shirley · Statistics in Medicine · 1982
The advantages and disadvantages of using extraneous information about the relationship between variate and covariate in analysis of covariance are examined. The problem is considered both theoretically and by a simulation study. It is found that the usefulness of extraneous information in increasing the precision of analysis of covariance can be assessed using simple formulae.