Abstract:
|
Calibration weighting is widely used to decrease variance and reduce nonresponse bias. In the purely sampling context, Deville & Särndal (1992) demonstrate that many alternative forms of calibration weighting are asymptotically equivalent. It is unclear whether this conclusion holds when nonresponse exists and single-step calibration weighting is used to reduce nonresponse bias. In this paper, we examine three widely used calibration estimators, the GREG with only main effect covariates (GREG_Main), poststratification, and raking. We first demonstrate that with nonresponse, GREG_Main, poststratification, and raking may perform differently and survey practitioners should examine the outcome model and the response pattern when choosing between these estimators. Then we examine several alternative variance estimators for raking with nonresponse and show that when raking is model-biased, none of the linearization variance estimators under evaluation is unbiased. In contrast, the jackknife replication method performs well in variance estimation, although the confidence interval may still be centered in the wrong place if the point estimate is biased.
|