Strengthening Instrumental Variables Through Weighting
- 229 Downloads
Instrumental variable (IV) methods are widely used to deal with the issue of unmeasured confounding and are becoming popular in health and medical research. IV models are able to obtain consistent estimates in the presence of unmeasured confounding, but rely on assumptions that are hard to verify and often criticized. An instrument is a variable that influences or encourages individuals toward a particular treatment without directly affecting the outcome. Estimates obtained using instruments with a weak influence over the treatment are known to have larger small-sample bias and to be less robust to the critical IV assumption that the instrument is randomly assigned. In this work, we propose a weighting procedure for strengthening the instrument while matching. Through simulations, weighting is shown to strengthen the instrument and improve robustness of resulting estimates. Unlike existing methods, weighting is shown to increase instrument strength without compromising match quality. We illustrate the method in a study comparing mortality between kidney dialysis patients receiving hemodialysis or peritoneal dialysis as treatment for end-stage renal disease.
KeywordsCausal inference End-stage renal disease Instrumental variables Unmeasured confounding Weak instruments
The authors have benefited from conversations with Michael Biaocchi. This project has been funded in whole or in part through funds from the National Institute of Diabetes and Digestive and Kidney Diseases, the National Institutes of Health, and the Department of Health and Human Services, under Contract No. HHSN276201400001C. Yun Li is funded by the National Institutes of Health (R01-DK070869).
- 6.Belloni A, Chernozhukov V, Hansen C (2010) Lasso methods for gaussian instrumental variables models. arXiv:1012.1297
- 7.Bound J, Jaeger DA, Baker RM (1995) Problems with instrumental variables estimation when the correlation between the instruments and the endogenous explanatory variable is weak. J Am Stat Assoc 90(430):443–450Google Scholar
- 9.Caner M, Fan Q (2010) The adaptive lasso method for instrumental variable selection. Technical report, working paper, North Carolina State UniversityGoogle Scholar
- 16.Kim H, Kim KH, Park K, Kang SW, Yoo TH, Ahn SV, Ahn HS, Hann HJ, Lee S, Ryu JH, Kim SJ, Kang DH, Choi KB, Ryu DR (2014) A population-based approach indicates an overall higher patient mortality with peritoneal dialysis compared to hemodialysis in korea. Kidney Int 86(5):991–1000CrossRefGoogle Scholar
- 25.Neyman J (1923) On the application of probability theory to agricultural experiments. Stat Sci 5:463–480Google Scholar
- 34.Tam P (2009) Peritoneal dialysis and preservation of residual renal function. Perit Dial Int 29(Supplement 2):S108–S110Google Scholar