Probabilistic Pairwise Model Comparisons Based on Discrepancy Measures
and a Reconceptualization of the P-Value
Presented by: Joseph Cavanaugh, Ph.D.,
Professor of Statistics and Actuarial Science, College of Liberal Arts and Sciences,
Professor of Statistics, College of Public Health, University of Iowa, Fellow of the ASA
Date: Keynote Address
Thursday, April 9, 2020
1:15 p.m. – 2:15 p.m. Central time
Location: University of Kansas, Edwards Campus
Regnier Hall Auditorium Rm 164
12600 Quivira Rd.,
Overland Park, KS 66213
Keynote Address Abstract:
In problems involving the selection and assessment of statistical models, discrepancy measures are often employed. A discrepancy gauges the separation between a fitted candidate model and the underlying generating model. In this work, we consider pairwise comparisons of fitted models based on a probabilistic evaluation of the ordering of the constituent discrepancies. An estimate of the probability is derived using the bootstrap.
In the framework of hypothesis testing, nested models are often compared on the basis of the p-value. Specifically, the simpler null model is favored unless the p-value is sufficiently small, in which case the null model is rejected and the more general alternative model is retained. We argue that in certain settings, the p-value and the bootstrapped discrepancy comparison probability (BDCP) are equivalent. In particular, we have established this equivalence for the Wald, score, and likelihood ratio tests by employing suitably defined discrepancy measures.
We contend that the connection between the p-value and the BDCP not only leads to potentially new insights regarding the utility and limitations of the p-value, yet also facilitates discrepancy-based inferences in settings beyond the limited confines of nested model hypothesis testing. We explore the behavior of the BDCP in a simulation study, and illustrate its utility in a biomedical application pertaining to the post-AMI treatment efficacy of ACE (angiotensin converting enzyme) inhibitors and ARBs (angiotensin receptor blockers).
This work is joint with Ben Riedle and Andrew Neath.
Presenter: Joseph Cavanaugh Ph.D.
Dr Cavanaugh is a Professor of Statistics and Actuarial Science in the College of Liberal Arts and Sciences and a Professor of Statistics in the College of Public Health at the University of Iowa and he is a Fellow of the ASA.
Matched Pairs and Longitudinal Binary Observations
Presented by: Bernhard Klingenberg, Ph.D.,
Professor of Statistics, Department of Mathematics and Statistics, Williams College
Date: Short Course
Thursday, April 9, 2020
8:15 a.m. – 12:15 p.m. Central time
Location: University of Kansas, Edwards Campus
Regnier Hall Auditorium Rm 164
12600 Quivira Rd.,
Overland Park, KS 66213
Course Abstract:
TBD
Presenter: Bernhard Klingenberg, Ph.D
Dr Klingenberg is a Professor of Statistics in the Department of Mathematics and Statistics at Williams College