Evaluation of various estimators for standardized mean difference in meta‐analysis

Statistics in Medicine - Tập 40 Số 2 - Trang 403-426 - 2021
Lifeng Lin1, Ariel M. Aloe2
1Department of Statistics, Florida State University Tallahassee, Florida, USA
2Educational Measurement and Statistics, University of Iowa, Iowa City, Iowa, USA

Tóm tắt

Meta‐analyses of a treatment's effect compared with a control frequently calculate the meta‐effect from standardized mean differences (SMDs). SMDs are usually estimated by Cohen's d or Hedges' g. Cohen's d divides the difference between sample means of a continuous response by the pooled standard deviation, but is subject to nonnegligible bias for small sample sizes. Hedges' g removes this bias with a correction factor. The current literature (including meta‐analysis books and software packages) is confusingly inconsistent about methods for synthesizing SMDs, potentially making reproducibility a problem. Using conventional methods, the variance estimate of SMD is associated with the point estimate of SMD, so Hedges' g is not guaranteed to be unbiased in meta‐analyses. This article comprehensively reviews and evaluates available methods for synthesizing SMDs. Their performance is compared using extensive simulation studies and analyses of actual datasets. We find that because of the intrinsic association between point estimates and standard errors, the usual version of Hedges' g can result in more biased meta‐estimation than Cohen's d. We recommend using average‐adjusted variance estimators to obtain an unbiased meta‐estimate, and the Hartung‐Knapp‐Sidik‐Jonkman method for accurate estimation of its confidence interval.

Từ khóa


Tài liệu tham khảo

10.1038/nature25753

10.1001/jama.2014.8167

10.1001/jama.2014.5559

10.1002/jrsm.1146

10.1002/bimj.201800071

10.1016/j.jclinepi.2019.11.015

10.1007/s10654-018-0449-x

10.1002/sim.6632

10.1111/2041-210X.12927

10.1136/bmj.k4817

10.1001/jama.286.10.1218

10.1371/journal.pmed.0050045

10.1186/1471-2288-14-30

10.1056/NEJMsa065779

Hedges LV, 1985, Statistical Methods for Meta‐Analysis

10.1093/ije/dys124

10.1001/jama.298.4.430

Tendal B, 2009, Disagreements in meta‐analyses using outcomes measured on continuous or rating scales: observer agreement study, BMJ, 3128, 339

10.1371/journal.pone.0204056

10.1002/ecs2.2419

10.3102/10769986006002107

10.1037/0033-2909.92.2.490

10.1037/0033-2909.93.2.388

Cohen J, 1977, Statistical Power Analysis for the Behavioral Sciencies

Casella G, 2001, Statistical Inference

10.1191/1740774505cn081oa

10.1002/9780470693926

10.1002/9780470386347

CooperH HedgesLV ValentineJC.The Handbook of Research Synthesis and Meta‐Analysis 2nd ed.New York NY:Russell Sage Foundation;2009.

10.1002/9780470743386

10.1201/9780203909935

10.1515/9781400846184

10.1007/978-1-4614-2278-5

Lipsey MW, 2001, Practical Meta‐Analysis

10.1007/978-3-319-21416-0

Schlattmann P, 2009, Medical Applications of Finite Mixture Models

10.3102/0013189X005010003

Glass GV, 1981, Meta‐Analysis in Social Research

10.3389/fpsyg.2020.00660

Cohen J, 1988, Statistical Power Analysis for the Behavioral Sciences

10.1016/0140-6736(91)91975-Z

10.1136/bmj.309.6957.789

10.1136/bmj.309.6965.1351

10.1001/jama.1996.03540160054033

10.1056/NEJM199708213370806

10.1016/S0140-6736(99)04149-5

10.1002/jrsm.12

10.1016/0197-2456(86)90046-2

10.6028/jres.087.022

10.1002/(SICI)1097-0258(19990215)18:3<321::AID-SIM28>3.0.CO;2-P

10.7326/M13-2886

10.1002/jrsm.1140

10.1002/jrsm.1164

10.1002/sim.7665

10.1002/jrsm.1316

10.1080/01621459.1978.10480117

10.1007/978-1-4614-4818-1

10.1177/0013164498058002005

10.1002/sim.4780140406

10.1002/sim.1009

10.1002/sim.1482

10.1002/sim.1262

10.1081/SAC-120023885

10.1186/1471-2288-14-25

10.1002/sim.4029

10.1093/ije/dyn204

10.1016/S2468-1253(16)30206-0

10.1001/jamapsychiatry.2014.1611

10.18637/jss.v036.i03

Schwarzer G, 2007, Meta: an R package for meta‐analysis, R News, 7, 40

10.1002/9781119536604

Palmer TM, 2016, Meta‐Analysis in Stata: An Updated Collection from the Stata Journal

10.3758/BF03195529

10.1002/sim.7186

10.1002/sim.8422

10.1136/bmj.c221

10.1002/sim.918

10.1002/sim.3165

10.1093/biomet/asv011

10.1371/journal.pmed.1003019

Hunter JE, 1990, Methods of Meta‐Analysis: Correcting Error and Bias in Research Findings

Card NA, 2012, Applied Meta‐Analysis for Social Science Research