Likelihood analysis of the binary instrumental variable model
Research output: Contribution to journal › Journal article › Research › peer-review
Standard
Likelihood analysis of the binary instrumental variable model. / Ramsahai, R. R.; Lauritzen, Steffen L.
In: Biometrika, Vol. 98, No. 4, 2011, p. 987-994.Research output: Contribution to journal › Journal article › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - JOUR
T1 - Likelihood analysis of the binary instrumental variable model
AU - Ramsahai, R. R.
AU - Lauritzen, Steffen L.
PY - 2011
Y1 - 2011
N2 - Instrumental variables are widely used for the identification of the causal effect of one random variable on another under unobserved confounding. The distribution of the observable variables for a discrete instrumental variable model satisfies certain inequalities but no conditional independence relations. Such models are usually tested by checking whether the relative frequency estimators of the parameters satisfy the constraints. This ignores sampling uncertainty in the data. Using the observable constraints for the instrumental variable model, a likelihood analysis is conducted. A significance test for its validity is developed, and a bootstrap algorithm for computing confidence intervals for the causal effect is proposed. Applications are given to illustrate the advantage of the suggested approach.
AB - Instrumental variables are widely used for the identification of the causal effect of one random variable on another under unobserved confounding. The distribution of the observable variables for a discrete instrumental variable model satisfies certain inequalities but no conditional independence relations. Such models are usually tested by checking whether the relative frequency estimators of the parameters satisfy the constraints. This ignores sampling uncertainty in the data. Using the observable constraints for the instrumental variable model, a likelihood analysis is conducted. A significance test for its validity is developed, and a bootstrap algorithm for computing confidence intervals for the causal effect is proposed. Applications are given to illustrate the advantage of the suggested approach.
U2 - 10.1093/biomet/asr040
DO - 10.1093/biomet/asr040
M3 - Journal article
VL - 98
SP - 987
EP - 994
JO - Biometrika
JF - Biometrika
SN - 0006-3444
IS - 4
ER -
ID: 128006787