Practical valid inferences for the two-sample binomial problem
From MaRDI portal
Abstract: Our interest is whether two binomial parameters differ, which parameter is larger, and by how much. This apparently simple problem was addressed by Fisher in the 1930's, and has been the subject of many review papers since then. Yet there continues to be new work on this issue and no consensus solution. Previous reviews have focused primarily on testing and the properties of validity and power, or primarily on confidence intervals, their coverage, and expected length. Here we evaluate both. For example, we consider whether a p-value and its matching confidence interval are compatible, meaning that the p-value rejects at level if and only if the confidence interval excludes all null parameter values. For focus, we only examine non-asymptotic inferences, so that most of the p-values and confidence intervals are valid (i.e., exact) by construction. Within this focus, we review different methods emphasizing many of the properties and interpretational aspects we desire from applied frequentist inference: validity, accuracy, good power, equivariance, compatibility, coherence, and parameterization and direction of effect. We show that no one method can meet all the desirable properties and give recommendations based on which properties are given more importance.
This paper is a study on Fisher's exact test. The question is whether two binomial parameters differ, which parameter is larger, and by how much. The authors define a frequentist triple as an estimator of the parameter of interest, a confidence interval, and a \(p\)-value function with computational issues. They evaluate both, the properties of validity and power \textit{and} confidence intervals, their coverage, and expected length. They limit the scope of this paper by considering only frequentist approaches, leaving out asymptotic methods or other approximations. They work out that no method can meet all the desirable properties and give recommendations based on which properties are given more importance. They examine a real application to discuss the issues. They show properties of frequentist triples. They develop methods for creating one-sided exact unconditional testing procedures, one-sided conditional exact tests, melded confidence intervals, non-central confidence intervals and associated tests, and mid-\(p\) methods. Finally, they give some recommendations how to perform frequentist inferences on the two-sample binomial problem.
Recommendations
- A cautionary note on exact unconditional inference for a difference between two independent binomial proportions
- Resolving the conflict over fisher's exact test
- Contemporary frequentist views of the 2 2 binomial trial
- The exact unconditional z-pooled test for equality of two binomial probabilities: optimal choice of the Berger and Boos confidence coefficient
- Exact tests for equality of two proportions: fisher v. bayes
Cites work
- UMP(U)-tests for a binomial parameter: A paradox
- A cautionary note on exact unconditional inference for a difference between two independent binomial proportions
- A New Test for 2 × 2 Tables
- An optimality theory for mid p-values in 2 2 contingency tables
- Asymptotical tests in 2 2 comparative trials: (unconditional approach).
- Choosing the optimal unconditioned test for comparing two independent proportions
- Combining one-sample confidence procedures for inference in the two-sample case
- Computation of exact confidence limits from discrete data
- Computing an Exact Confidence Interval for the Common Odds Ratio in Several 2 × 2 Contingency Tables
- Confidence curves and improved exact confidence intervals for discrete distributions
- Confidence distribution, the frequentist distribution estimator of a parameter: a review
- Confidence Intervals for Proportions and Related Measures of Effect Size
- Confidence intervals that match Fisher's exact or Blaker's exact tests
- Contemporary frequentist views of the 2 2 binomial trial
- Exact P-values for discrete models obtained by estimation and maximization
- Exact Analysis of Discrete Data
- Exact confidence intervals for the relative risk and the odds ratio
- Exact solutions to the Behrens - Fisher problem: asymptotically optimal and finite sample efficient choice among
- Improved Buehler limits based on refined designated statistics
- Information, for testing the equality of two probabilities, from the margins of the \(2 \times 2\) table.
- On construction of the smallest one-sided confidence interval for the difference of two proportions
- On small-sample confidence intervals for parameters in discrete distributions
- On the Optimality and Limitations of Buehler Bounds
- P Values Maximized Over a Confidence Set for the Nuisance Parameter
- Problems with existing procedures to calculate exact unconditional P-values for non-inferiority/superiority and confidence intervals for two binomials and how to resolve them
- Proving non-inferiority or equivalence of two treatments with dichotomous endpoints using exact methods
- Raised conditional level of significance for the 2 × 2‐table when testing the equality of two probabilities
- Sampling Methods for Wallenius' and Fisher's Noncentral Hypergeometric Distributions
- Small sample confidence intervals for the difference,ratio and odds ratio of two success probabilities
- Small-sample comparisons of confidence intervals for the difference of two independent binomial proportions
- Small-Sample Confidence Intervals for p 1 - p 2 and p 1 /p 2 in 2 × 2 Contingency Tables
- Statistics in Epidemiology: The Case-Control Study
- Table of percentage points of the behrens-fisher distribution
- Test of Significance for 2 × 2 Contingency Tables
- Testing Statistical Hypotheses
- Test‐Based Exact Confidence Intervals for the Difference of Two Binomial Proportions
- The efficiency of Buehler confidence limits.
- The equivalence of the mid p-value and the expected p-value for testing equality of two balanced binomial proportions
- THE MEANING OF A SIGNIFICANCE LEVEL
- The minimaxity of the mid P-value under linear and squared loss functions
- Unconditional Non-Asymptotic One-Sided Tests for Independent Binomial Proportions When the Interest Lies in Showing Non-Inferiority and/or Superiority
- Unconditional small-sample confidence intervals for the odds ratio
- Univariate Discrete Distributions
Cited in
(11)- Interpreting \(p\)-values and confidence intervals using well-calibrated null preference priors
- The two-sample problem via relative belief ratio
- exact2x2
- Plausibility functions and exact frequentist inference
- Optimal confidence interval for the difference between proportions
- Design of randomized clinical trials with a binary endpoint: conditional versus unconditional analyses of a two-by-two table
- Are the tests overpowered or underpowered? A unified solution to correctly specify type I errors in design of clinical trials for two sample proportions
- Analyzing environmental bioassays with spatial odds, risk, and survival probability ratio regressions
- Melded Confidence Intervals Do Not Provide Guaranteed Coverage
- Exact samples sizes for clinical trials subject to size and power constraints
- A boundary-optimized rejection region test for the two-sample binomial problem
This page was built for publication: Practical valid inferences for the two-sample binomial problem
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q76638)
