Hybrid Confidence Intervals for Informative Uniform Asymptotic Inference After Model Selection, accepted at Biometrika
Short and Simple Confidence Intervals when the Directions of Some Effects are Known (with Philipp Ketz), accepted at Review of Economics and Statistics
Stata code available from SSC archive: type "ssc install ssci"
Inference for Losers (with Isaiah Andrews, Dillon Bowen and Toru Kitagawa), American Economic Association Papers and Proceedings, 112 (2022), 635-640.
Inference After Estimation of Breaks (with Isaiah Andrews and Toru Kitagawa), Journal of Econometrics, 224 (2021), 39-59.
Asymptotically Uniform Tests After Consistent Model Selection in the Linear Regression Model, Journal of Business and Economic Statistics, 38 (2020), 810-825.
Estimation and Inference with a (Nearly) Singular Jacobian (with Sukjin Han), Quantitative Economics, 10 (2019), 1019-1068.
Bonferroni-Based Size-Correction for Nonstandard Testing Problems, Journal of Econometrics, 200 (2017), 17-35.
Parameter Estimation Robust to Low-Frequency Contamination (with Jonathan B. Hill), Journal of Business and Economic Statistics, 35 (2017), 598-610.
Memory Parameter Estimation in the Presence of Level Shifts and Deterministic Trends (with Pierre Perron), Econometric Theory, 29 (2013), 1196-1237.
Estimation of the Long-Memory Stochastic Volatility Model Parameters that is Robust to Level Shifts and Deterministic Trends, Journal of Time Series Analysis, 34 (2013), 285-301.
Critical Values Robust to P-hacking (with Pascal Michaillat)
(previously titled "Incentive-Compatible Critical Values")
Despite its prevalence, p-hacking is not taken into account in classical hypothesis testing theory. To address this problem, we build a model of p-hacking and use it to construct critical values such that, if these values are used to determine significance, and if scientists adjust their p-hacking behavior to these new significance standards, then significant results occur with the desired frequency. Because such robust critical values allow for p-hacking, they are larger than classical critical values. For instance, in the model calibrated to the social and medical sciences, the robust critical value for any test statistic is the classical critical value for the same test statistic with one fifth of the significance level.
Inference on Winners (with Isaiah Andrews and Toru Kitagawa), revise and resubmit at Quarterly Journal of Economics
2019 Version (referenced in "Inference After Estimation of Breaks")
Many empirical questions concern target parameters selected through optimization. For example, researchers may be interested in the effectiveness of the best policy found in a randomized trial, or the best-performing investment strategy based on historical data. Such settings give rise to a winner's curse, where conventional estimates are biased and conventional confidence intervals are unreliable. This paper develops optimal confidence intervals and median-unbiased estimators that are valid conditional on the target selected and so overcome this winner's curse. If one requires validity only on average over targets that might have been selected, we develop hybrid procedures that combine conditional and projection confidence intervals to offer further performance gains relative to existing alternatives.
Adam McCloskey
Department of Economics
University of Colorado at Boulder
256 UCB
Boulder, CO 80309
Phone: (303) 735-7908 Fax: (303) 492-8960 E-mail: adam.mccloskey@colorado.edu