Sensitivity Analysis for Survey Weights

-
Cunz 160 and Zoom

Erin Hartman, PhD, Berkeley

Erin Hartmann

Erin Hartman, PhD 
Assistant Professor 
University of California, Berkley Department of Political Science 
https://erinhartman.com/

 

 

 

 

Survey weighting allows researchers to account for bias in survey samples, due to unit nonresponse or  convenience sampling, using measured demographic covariates. Unfortunately, in practice, it is impossible to  know whether the estimated survey weights are sufficient to alleviate concerns about bias due to  unobserved confounders or incorrect functional forms used in weighting. In the following paper, we propose  two sensitivity analyses for the exclusion of important covariates: (1) a sensitivity analysis for partially  observed confounders (i.e., variables measured across the survey sample, but not the target population), and  (2) a sensitivity analysis for fully unobserved confounders (i.e., variables not measured in either the survey or  the target population). We provide graphical and numerical summaries of the potential bias that arises from  such confounders, and introduce a benchmarking approach that allows researchers to quantitatively reason  about the sensitivity of their results. We demonstrate our proposed sensitivity analyses using state-level  2020 U.S. Presidential Election polls. 

Zoom link: https://osu.zoom.us/j/94300135379?pwd=aElUeitFRDd6WTZhRHlicyttMkMwdz09