Karl Pearson’s meta analysis revisited

  title={Karl Pearson’s meta analysis revisited},
  author={Art B. Owen},
  journal={Quality Engineering},
  • A. Owen
  • Published 18 November 2009
  • Mathematics
  • Quality Engineering
This paper revisits a meta-analysis method proposed by Pearson [Biometrika 26 (1934) 425-442] and first used by David [Biometrika 26 (1934) 1-11 ]. It was thought to be inadmissible for over fifty years, dating back to a paper of Birnbaum [J. Amer. Statist. Assoc. 49 (1954) 559-574]. It turns out that the method Birnbaum analyzed is not the one that Pearson proposed. We show that Pearson's proposal is admissible. Because it is admissible, it has better power than the standard test of Fisher… 

Figures from this paper

Meta-analysis based on weighted ordered P-values for genomic data with heterogeneity

This paper proposes a class of meta-analysis methods based on summaries of weighted ordered p-values (WOP) that aim at detecting significance in a majority of studies and investigates the properties of the methods and demonstrates their strengths through simulations studies, comparing to existing procedures.


An order statistic of p-values across combined studies as the test statistic is proposed and applied to three microarray meta-analysis examples including major depressive disorder, brain cancer and diabetes to demonstrate rOP as a more generalizable, robust and sensitive statistical framework to detect disease-related markers.

A new association test based on Chi‐square partition for case‐control GWA studies

A new test based on the partition of Pearson's Chi‐square test statistic is proposed, which utilizes the information of the monotonic trend of relative risks and therefore is more powerful than the Chi‐ square test; furthermore, it reserves the robustness.


A genome-wide association study identifies alleles in FGFR2 associated with risk of sporadic postmenpausal breast cancer and a method of combining measures of association and Hardy-Weinberg disequilibrium for fine mapping in case-control studies.

Comprehensive literature review and statistical considerations for GWAS meta-analysis

This article systematically reviews the GWAS meta-analysis literature, highlighting methodology and software options and reviewing methods that have been used in real studies, and illustrating differences among methods using a case study.

J an 2 01 4 Replicability analysis for genome-wide association studies

A formal empirical Bayes approach for discovering whether results have been replicated across studies, in which the optimal rejection region for discovering replicated results is estimated, is suggested.

Replicability Across Multiple Studies

A selective overview of analyses that can be carried out towards establishing replicability of the scientific findings is provided and methods for the setting where a single outcome is examined in multiple studies are described.

Imputation of Truncated p-Values For Meta-Analysis Methods and Its Genomic Application.

The proposed methods were applied to several genomic applications in colorectal cancer, pain and liquid association analysis of major depressive disorder and showed that imputation methods outperformed existing naïve approaches and are recommended for future applications.

Stouffer’s Test in a Large Scale Simultaneous Hypothesis Testing

Stouffer’s test based on z-scores is revisited and showed its advantage over the two aforementioned methods in the analysis of large-scale microarray data and differentially expressed genes by non-genotoxic and genotoxic carcinogen compounds are found.



Combining probability from independent tests: the weighted Z‐method is superior to Fisher's approach

  • M. Whitlock
  • Mathematics
    Journal of evolutionary biology
  • 2005
The results in this note show that, when combining P‐values from multiple tests of the same hypothesis, the weighted Z‐method should be preferred.


y is a non-decreasing function of x, having values confined to the interval (0, 1). Further d p(y) = p(x)Iy = I for < y < l. . (3) ~!dx In other words the probability law for the integral, y, is

Screening for Partial Conjunction Hypotheses

This work addresses the problem of testing many partial conjunction hypotheses simultaneously using the false discovery rate (FDR) approach and proves that if the FDR controlling procedure in Benjamini and Hochberg is used for this purpose the FDR is controlled under various dependency structures.

Numerical Methods of Statistics

This is a very useful textbook for undergratuate students of statistics or a graduate level course for industrial engineers and should be on the reference book shelf of everyone who works in manufacturing quality control.

An improved Bonferroni procedure for multiple tests of significance

SUMMARY A modification of the Bonferroni procedure for testing multiple hypotheses is presented. The method, based on the ordered p-values of the individual tests, is less conservative than the

Multiple Testing Procedures with Applications to Genomics

This chapter discusses single-Step Multiple Testing Procedures for Controlling General Type I Error Rates, as well asmentation and resampling-Based Empirical Bayes multiple testing procedures forcontrolling Generalized Tail Probability Error Rates.

Categorical Data Analysis

Modeling and inferential procedures (estimation and precision assessment) are discussed for a single proportion, thereafter in the context of cross-classified data, i.e., contingency tables, is discussed, with particular emphasis on testing the null hypothesis of no association (independence) between the row and column classification.

Combining Independent Tests of Significance

It is shown that no single method of combining independent tests of significance is optimal in general, and hence that the kinds of tests to be combined should be considered in selecting a method of combination.

AGEMAP: A Gene Expression Database for Aging in Mice

The AGEMAP (Atlas of Gene Expression in Mouse Aging Project) gene expression database is presented, which is a resource that catalogs changes in gene expression as a function of age in mice and finds great heterogeneity in the amount of transcriptional changes with age in different tissues.

Statistical methods for research workers

The prime object of this book is to put into the hands of research workers, and especially of biologists, the means of applying statistical tests accurately to numerical data accumulated in their own