Publication Date: December 2014
Revision Date: October 2016
>Kernel-based estimators are often evaluated at multiple bandwidths as a form of sensitivity analysis. However, if in the reported results, a researcher selects the bandwidth based on this analysis, the associated conﬁdence intervals may not have correct coverage, even if the estimator is unbiased. This paper proposes a simple adjustment that gives correct coverage in such situations: replace the normal quantile with a critical value that depends only on the kernel and ratio of the maximum and minimum bandwidths the researcher has entertained. We tabulate these critical values and quantify the loss in coverage for conventional conﬁdence intervals. For a range of relevant cases, a conventional 95% conﬁdence interval has coverage between 70% and 90%, and our adjustment amounts to replacing the conventional critical value 1.96 with a number between 2.2 and 2.8. Our results also apply to other settings involving trimmed data, such as trimming to ensure overlap in treatment eﬀect estimation. A Monte Carlo study conﬁrms that our approach gives accurate coverage in ﬁnite samples. We illustrate our approach with three empirical applications.
Supplemental Material: Supplemental material
Supplement pages: 41
Nonparametric estimation, Multiple testing, Regression discontinuity
JEL Classification Codes: C01, C14