Online Program Home
My Program

Abstract Details

Activity Number: 579 - Sampling, Variance Estimation, and Advancements with Auxiliary Data
Type: Contributed
Date/Time: Wednesday, July 31, 2019 : 2:00 PM to 3:50 PM
Sponsor: Survey Research Methods Section
Abstract #305122
Title: TRUMP: Tuned Regression Unbiased Mean Predictor
Author(s): Sarjinder Singh* and Stephen Sedory
Companies: Texas A&M University-Kingsville and Texas A & M University-Kingsville
Keywords: Calibration; TRUMP Cuts; TRUMP Care Coefficient; Chain Type TRUMP Cuts; First Basic Information (FBI)

In this paper, we introduce a new Tuned Regression Unbiased Mean Predictor (TRUMP) which we show that can be adjusted for smaller variance than the linear regression predictor due to Hansen, Hurwitz and Madow (1953) when there is Heteroscedasticity, which we call here Hillary Campaign Coefficient(H). Thus the proposed new TRUMP model can be made more efficient than the Best Linear Unbiased Predictor (BLUP) based on the choice of a TRUMP Care coefficient (g). R-codes to find values of the TRUMP Care Coefficient (g) for beating the Hillary Campaign Coefficient(H) are also included. At the end, use of multi-auxiliary variables case has been discussed.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2019 program