Multiple comparisons and association selection in general epidemiology
International Journal of EpidemiologyPublished 3 May 2008Open access
Sander Greenland
Citations86
SJR quartileQ1
SJR score2.69
SNIP2.48
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
Comments here are intended to clarify some points and outline connections of his discussion to broader problems, describing how methods such as Wakefield’s can be appropriate in epidemiology beyond genetic research.
Abstract
In this issue of the journal, Prof. Jon Wakefield provides a contribution to the complex topic of screening genetic associations. My comments here are intended to clarify some points and outline connections of his discussion to broader problems, describing how methods such as Wakefield’s can be appropriate in epidemiology beyond genetic research. I will also comment on a few aspects of his presentation related to technical issues. I will assume (as does Wakefield) that the reader is familiar with the terminology of Bayesian statistics as well that of conventional (frequentist) statistics. I will start with a comment on interpretation of P-values, and then turn to the issue of multiple comparisons.
Keywords
MathematicsBiochemistry, Genetics and Molecular Biology
TechnometricsApplied Regression Analysis
18,042 Citations2005
This tutorial discusses simple and multiple linear regression, diagnostics, model selection, models with categorical variables, and nonlinear models; logistic regression.
EpidemiologyNo Adjustments Are Needed for Multiple Comparisons
5,844 Citations1990Kenneth J. Rothman
A policy of not making adjustments for multiple comparisons is preferable because it will lead to fewer errors of interpretation when the data under evaluation are not random numbers but actual observations on nature.
The Annals of StatisticsBayesianly Justifiable and Relevant Frequency Calculations for the Applied Statistician
1,349 Citations1984Donald B. Rubin
Journal of the American Statistical AssociationPublication Decisions and their Possible Effects on Inferences Drawn from Tests of Significance—or Vice Versa
936 Citations1959T. D. Sterling
The American StatisticianCalibration of<i>ρ</i>Values for Testing Precise Null Hypotheses
857 Citations2001Thomas Sellke, M. J. Bayarri +1 more
Journal of the American Statistical AssociationData Analysis Using Stein's Estimator and its Generalizations
842 Citations1975Bradley Efron, Carl N. Morris
American Journal of EpidemiologyMultiple Comparisons and Related Issues in the Interpretation of Epidemiologic Data
469 Citations1995David A. Savitz, Andrew F. Olshan
International Journal of EpidemiologyBayesian perspectives for epidemiological research: I. Foundations and basic methods
384 Citations2006Sander Greenland
It is argued that a criterion for scientific acceptability of a prior distribution is that it be expressible as prior data, so that the strength of prior assumptions can be gauged by how much data they represent.
American Journal of Epidemiologyp Values, Hypothesis Tests, and Likelihood: Implications for Epidemiology of a Neglected Historical Debate
305 Citations1993Steven N. Goodman
An analysis using another method promoted by Fisher, mathematical likelihood, shows that the p value substantially overstates the evidence against the null hypothesis.
American Journal of EpidemiologyInvited Commentary: Variable Selection versus Shrinkage in the Control of Multiple Confounders
270 Citations2007Sander Greenland
International Journal of EpidemiologyBayesian perspectives for epidemiological research. II. Regression analysis
206 Citations2007Sander Greenland
Extensions of the basic Bayesian methods using data priors to regression modelling, including hierarchical (multilevel) models, provide an alternative to the parsimony-oriented approach of frequentist regression analysis and facilitate realistic use of imprecise but important prior information.
Journal of the American Statistical AssociationBayesians, Frequentists, and Scientists
198 Citations2005Bradley Efron
American Journal of EpidemiologyTHE PROBLEM OF MULTIPLE INFERENCE IN STUDIES DESIGNED TO GENERATE HYPOTHESES
193 Citations1985Duncan C. Thomas, Jack Siemiatycki +4 more
Methods are illustrated by application to preliminary data from a study aimed at identifying hitherto unsuspected occupational carcinogens, and prefer an approach in which all associations in the data are reported, whether significant or not, followed by a ranking in order of priority for investigation using empirical Bayes techniques.
BiometricsWhen Should Epidemiologic Regressions Use Random Coefficients?
163 Citations2000Sander Greenland
It is argued that regression models with random coefficients offer a more scientifically defensible framework for epidemiologic analysis than the fixed‐effects models now prevalent in epidemiology.
Statistics in MedicineMethods for epidemiologic analyses of multiple exposures: A review and comparative study of maximum‐likelihood, preliminary‐testing, and empirical‐bayes regression
143 Citations1993Sander Greenland
This work reviews two hierarchical methods and compares their performance in a problem of predicting neonatal mortality rates and suggests that hierarchical methods should become part of the standard approaches to multiple-exposure studies.
EpidemiologyEmpirical-Bayes Adjustments for Multiple Comparisons Are Sometimes Useful
138 Citations1991Sander Greenland, James M. Robins
In situations in which an additional or even primary goal of analysis is to reach a set of decisions based on the data, Bayes and empirical-Bayes adjustments can provide a better basis for the decisions than conventional procedures.
Journal of the American Statistical AssociationSmoking and Lung Cancer: Some Observations on Two Recent Reports
123 Citations1958Joseph Berkson
If smoking increases the risk of death, it should affect heavy smokers most; the first thing to do is to see whether the death rates of the heavy smokers are higher than the rates of those who are less than heavy smokers, which suggests itself as a sensitive method.
American Journal of EpidemiologyMultiple Comparisons, Explained
97 Citations1998Steven N. Goodman
This essay tries to show how the issue here is at the core of what epidemiology, and indeed science, is all about, and identifies the real issue for which the methodological arguments are a proxy.
BMC Medical Research MethodologyPublication bias in situ
93 Citations2004Carl V. Phillips
This research examines whether choices about how to analyze the data and which results to report create a publication bias within the published results, a bias I label "publication bias in situ" (PBIS).
International Statistical ReviewOn the Use of Wald's Test in Exponential Families
81 Citations1985Michael Væth, Michael Væth
Archives of Environmental Health An International JournalEmpirical-Bayes and Semi-Bayes Approaches to Occupational and Environmental Hazard Surveillance
70 Citations1994Sander Greenland, Charles Poole
This work provides an elementary introduction to empirical-Bayes analysis of occupational and environmental hazard surveillance data, using a study of occupational exposures and cancer mortality to illustrate key concepts.
International Journal of EpidemiologyReporting and interpretation in genome-wide association studies
67 Citations2008Jon Wakefield
The Bayes factor and the q-value provide complementary information and when used in addition to the P-value may be used to reduce the number of reported findings that are subsequently not reproduced.
The American StatisticianFirst (?) Occurrence of Common Terms in Mathematical Statistics
63 Citations1995H. A. David
A list of over 300 terms commonly used in mathematical statistics is presented with their apparent first occurrence in print, with some of the more interesting problems encountered in preparing the list.
PubMedEmpirical Bayes adjustments for multiple results in hypothesis-generating or surveillance studies.
63 Citations2000Kyle Steenland, Isabelle Bray +2 more
Empirical Bayes (EB) adjustments are used to evaluate standardized incidence ratios (SIRs) for cancer and occupation among craftsmen and laborers in Nordic countries to reduce the number of false positives otherwise flagged for further investigation.
Statistics in MedicineOn effect‐measure modification: Relationships among changes in the relative risk, odds ratio, and risk difference
38 Citations2008Babette Brumback, Arthur Berg
Investigating relationships among changes in the relative risk, odds ratio, and risk difference from one stratum to another finds circumstances under which the measures change in opposite directions.
International Journal of EpidemiologyCommentary: On 'Quality in epidemiological research: should we be submitting papers before we have the results and submitting more hypothesis generating research?'
18 Citations2007Debbie A. Lawlor
Lawlor’s excellent proposals to mitigate publication bias complements other ideas, such as (truly) blind review of completed studies by blanking-out study results, but certain conventional presumptions in the editorial are in error and have blinded discussants to important sources of publication bias.
American Journal of EpidemiologyA COUNTEREXAMPLE TO THE TEST-BASED PRINCIPLE OF SETTING CONFIDENCE LIMITS
16 Citations1984Sander Greenland