TY - JOUR
AU - Smith, T. M. F.
AB - Broadly defined, post-stratification embraces most methods involving the rewieghting of survey results after selection. Narrowly defined, as in the sampling literature including this paper, it refers to a method for increasing precision. As such it is just one of a wide class of methods including ratio and regression estimation. Some practical problems, such as small sample sizes in post-strata, can be solved by following Bethlehem & Keller and treating post-stratification as an example of regression estimation. It is argued that inferences should be made conditional on the selection of post-strata, and problems are identified with this approach for complex sampling schemes.
TI - Post-Stratification
JF - Journal of the Royal Statistical Society Series D: The Statistician
DO - 10.2307/2348284
DA - 2018-12-05
UR - https://www.deepdyve.com/lp/oxford-university-press/post-stratification-GBcBzX2O2a
SP - 315
EP - 323
VL - 40
IS - 3
DP - DeepDyve
ER -