Confidence-based decision theory

D. R. Bickel, “Coherent frequentism: A decision theory based on confidence sets,” Communications in Statistics – Theory and Methods 41, 1478-1496 (2012). Full article (open access) | 2009 version

This paper proposes a framework of inference based on a confidence posterior, a parameter probability distribution that does not require any prior distribution. While the Bayesian posterior is defined in terms of a conditional distribution given the observed data, the confidence posterior is instead defined such that the probability that the parameter value lies in any fixed subset of parameter space, given the observed data, is equal to the coverage rate of the corresponding confidence interval. Inferences based on the confidence posterior are reliable in the sense that the certainty level of a composite hypothesis is a weakly consistent estimate of the 0-1 indicator of hypothesis truth. At the same time, the confidence posterior is as non-contradictory as the Bayesian posterior since both satisfy the same coherence axioms. Using the theory of coherent upper and lower probabilities, the confidence posterior is generalized for situations in which no approximate or exact confidence set is available. Examples of hypothesis testing and estimation illustrate the range of applications of the proposed framework.

Additional summaries appear in the abstract and in Section 1.3 of the paper.

undefined

How to use priors with caution

D. R. Bickel, “Controlling the degree of caution in statistical inference with the Bayesian and frequentist approaches as opposite extremes,” Electronic Journal of Statistics 6, 686-709 (2012). Full text (open access) | 2011 preprint

This paper’s framework of statistical inference is intended to facilitate the development of new methods to bridge the gap between the frequentist and Bayesian approaches. Four concrete examples illustrate how such intermediate methods can leverage strengths of the two extreme approaches.

undefined

How to combine statistical methods

D. R. Bickel, “Game-theoretic probability combination with applications to resolving conflicts between statistical methods,” International Journal of Approximate Reasoning DOI:10.1016/j.ijar.2012.04.002 (2012). Online ahead of print | 2011 preprint

This paper proposes both a novel solution to the problem of combining probability distributions and a framework for using the new method to combine the results of differing statistical methods that may legitimately be used to analyze the same data set. While the paper emphasizes theoretical development, it is motivated by the need to combine two conflicting estimators of the probability of differential gene expression.

Effect-size estimates from hypothesis probabilities

D. R. Bickel, “Empirical Bayes interval estimates that are conditionally equal to unadjusted confidence intervals or to default prior credibility intervals,” Statistical Applications in Genetics and Molecular Biology 11 (3), art. 7 (2012). Full article | 2010 preprint

Estimating probabilities of enrichment

Z. Yang, Z. Li, and D. R. Bickel, “Empirical Bayes estimation of posterior probabilities of enrichment,” Technical Report, Ottawa Institute of Systems Biology, Technical Report, Ottawa Institute of Systems Biology, arXiv:1201.0153 (2011). Full preprint | 2010 seed

This paper adapts novel empirical Bayes methods for the problem of detecting enrichment in the form of differential representation of genes associated with a biological category with respect to a list of genes identified as differentially expressed. A microarray case study illustrates the methods using Gene Ontology (GO) terms, and a simulation study compares their performance. We report that which enrichment methods work best depends strongly on how many GO terms or other biological categories are of interest.

Combining inferences from different methods

D. R. Bickel, “Resolving conflicts between statistical methods by probability combination: Application to empirical Bayes analyses of genomic data,” Technical Report, Ottawa Institute of Systems Biology, arXiv:1111.6174 (2011). Full preprint

This paper proposes a solution to the problem of combining the results of differing statistical methods that may legitimately be used to analyze the same data set. The motivating application is the combination of two estimators of the probability of differential gene expression: one uses an empirical null distribution, and the other uses the theoretical null distribution. Since there is usually not any reliable way to predict which null distribution will perform better for a given data set and since the choice between them often has a large impact on the conclusions, the proposed hedging strategy addresses a pressing need in statistical genomics. Many other applications are also mentioned in the abstract and described in the introduction.

Minimax strength of statistical evidence

D. R. Bickel, “A predictive approach to measuring the strength of statistical evidence for single and multiple comparisons,” Canadian Journal of Statistics 39, 610–631 (2011). Full text | Revised preprint | 2010 draft

This paper introduces a novel approach to the multiple comparisons problem by generalizing a promising method of model selection developed by information theorists. The first two sections present that method and its main advantages over conventional approaches without burdening statisticians with unfamiliar terms from coding theory. A quantitative proteomics case study facilitates application of the new method to the analysis of data sets involving multiple biological features. The theorems describe its operating characteristics.

The cited medium-scale paper presented previous minimum description length (MDL) methods. Unlike those methods, the new MDL methods of the current paper are based on a conflation of the normalized maximum likelihood (NML) with the weighted likelihood (WL). The previous MDL methods are used in the CJS article for comparison with its NML/WL methods.

Degree of caution in inference

D. R. Bickel, “Controlling the degree of caution in statistical inference with the Bayesian and frequentist approaches as opposite extremes,” Technical Report, Ottawa Institute of Systems Biology, arXiv:1109.5278 (2011). Full preprint

This paper’s framework of statistical inference is intended to facilitate the development of new methods to bridge the gap between the frequentist and Bayesian approaches. Three concrete examples illustrate how such intermediate methods can leverage strengths of the two extreme approaches.

Software for local false discovery rate estimation

LFDR-MLE is a suite of R functions for the estimation of local false discovery rates by maximum likelihood under a two-group parametric mixture model of test statistics.

Extending the likelihood paradigm

D. R. Bickel, “The strength of statistical evidence for composite hypotheses: Inference to the best explanation,” to appear in Statistica Sinica (2011). 2010 Preprint