A re‐evaluation of fixed effect(s) meta‐analysis
Open accessAbstract
Summary Meta-analysis is a common tool for synthesizing results of multiple studies. Among methods for performing meta-analysis, the approach known as ‘fixed effects’ or ‘inverse variance weighting’ is popular and widely used. A common interpretation of this method is that it assumes that the underlying effects in contributing studies are identical, and for this reason it is sometimes dismissed by practitioners. However, other interpretations of fixed effects analyses do not make this assumption, yet appear to be little known in the literature. We review these alternative interpretations, describing both their strengths and their limitations. We also describe how heterogeneity of the underlying effects can be addressed, with the same minimal assumptions, through either testing or meta-regression. Recommendations for the practice of meta-analysis are given; it is hoped that these will foster more direct connection of the questions that meta-analysts wish to answer with the statistical methods they choose.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Inference from a fixed-effect model is appropriate for pooling a single replication effect with a single original effect when the target is the mean effect size for those two effects.
“For this purpose, inference from a fixed-effect model is appropriate (Rice et al., 2017)”
From Investigating the replicability of preclinical cancer biology (2021) · cites this paper as Rice et al., 2017
The fixed effects model for meta-analysis does not require the assumption of homogeneity.
“Even though the fixed effects model for meta-analysis does not require the assumption of homogeneity [34]”
From Sex-specific DNA methylation differences in Alzheimer’s disease pathology (2021) · cites this paper as [34]
Cites
Statistical Aspects of the Analysis of Data From Retrospective Studies of Disease
(1959) · JNCI: Journal of the National Cancer Institute · doi:10.1093/jnci/22.4.719 · Cited by 14,689 · cited 1× · see citing works
“In this notation, the classical Mantel–Haenszel estimator (Mantel and Haenszel, 1959; Kuritz et al., 1988) is ψ̂= k∑ i=1 AiDi ni / k∑ i′=1 Bi′Ci′ ni′ : It provides a measure of the observed association between the ‘row’ and ‘column’ variables, across a series of 2×2 tables.”
Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2009) · doi:10.1002/9780470743386 · Cited by 14,225 · see citing works
Self-citationMethods of Meta-Analysis: Correcting Error and Bias in Research Findings
Schmidt, F. L., & Hunter, J. E. (2015) · doi:10.4135/9781483398105 · Cited by 5,542 · see citing works
The Handbook of Research Synthesis and Meta-Analysis
(2019) · doi:10.7758/9781610448864 · Cited by 4,553 · see citing works
Hunter, J., & Schmidt, F. (2004) · doi:10.4135/9781412985031 · Cited by 3,023 · see citing works
Fixed-and random-effects models in meta-analysis.
Hedges, L. V., & Vevea, J. L. (1998) · Psychological Methods · doi:10.1037/1082-989x.3.4.486 · Cited by 2,977 · see citing works
A Re-Evaluation of Random-Effects Meta-Analysis
Higgins, J. P. T., Thompson, S. G., & Spiegelhalter, D. J. (2009) · Journal of the Royal Statistical Society Series A: Statistics in Society · doi:10.1111/j.1467-985x.2008.00552.x · Cited by 2,569 · cited 5× · see citing works
“The random-effects model is discussed at length elsewhere (Higgins et al., 2009); we particularly note here that its assumptions are, in general, different from both the common effect and fixed effects model.”
BackgroundResultInfluentialSelf-citationControlling the risk of spurious findings from meta-regression
Higgins, J. P. T., & Thompson, S. G. (2004) · Statistics in Medicine · doi:10.1002/sim.1752 · Cited by 1,267 · cited 2× · see citing works
“…through weighted linear regression and is SE.β̂MR/= √{ 1 / k∑ i=1 wi.xi − x̂/2 } : .6/ When true linearity of the βi with xi does not hold, this is not in general a valid estimate of the variability of β̂MR, and statistical inference based on it will not be valid (Higgins and Thompson, 2004).”
BackgroundSelf-citationThe power of statistical tests in meta-analysis
Hedges, L. V., & Pigott, T. D. (2001) · Psychological Methods · doi:10.1037/1082-989x.6.3.203 · Cited by 716 · cited 2× · see citing works
“Under heterogeneity Q has a non-central χ2k−1-distribution with non-centrality parameter Σki=1wi.βi −β/2 (Hedges and Pigott, 2001); in many practical settings this value will be small (Higgins et al., 2003) so the power of the test will be small.”
BackgroundComplex Surveys: A Guide to Analysis Using R
Lumley, T. (2010) · doi:10.1002/9780470580066 · Cited by 647 · see citing works
Self-citationPractical aspects of imputation-driven meta-analysis of genome-wide association studies
de Bakker, P. I. W., Ferreira, M. A. R., Jia, X., Neale, B. M., Raychaudhuri, S., & Voight, B. F. (2008) · Human Molecular Genetics · doi:10.1093/hmg/ddn288 · Cited by 531 · see citing works
Meta-analysis: Neither quick nor easy
Berman, N. G., & Parker, R. A. (2002) · BMC Medical Research Methodology · doi:10.1186/1471-2288-2-10 · Cited by 270 · see citing works
Zhong, H., & Prentice, R. L. (2010) · Genetic Epidemiology · doi:10.1002/gepi.20437 · Cited by 94 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
Page, M. J., Moher, D., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., et al. (2021) · BMJ · doi:10.1136/bmj.n160 · Cited by 11,883 · see citing works
Analysing data and undertaking meta‐analyses
Deeks, J. J., Higgins, J. P., Altman, D. G., & on behalf of the Cochrane Statistical Methods Group (2019) · Cochrane Handbook for Systematic Reviews of Interventions · doi:10.1002/9781119536604.ch10 · Cited by 2,672 · see citing works
Self-citationAnalysing Data and Undertaking Meta‐Analyses
Deeks, J. J., Higgins, J. P., & Altman, D. G. (2008) · Cochrane Handbook for Systematic Reviews of Interventions · doi:10.1002/9780470712184.ch9 · Cited by 1,800 · see citing works
Self-citationPage, M. J., Moher, D., Bossuyt, P., Boutron, I., Hoffmann, T., mulrow, C., Shamseer, L., Tetzlaff, J., Akl, E., Brennan, S. E., et al. (2020) · doi:10.31222/osf.io/gwdhk · Cited by 916 · see citing works
Correcting for Bias in Psychology: A Comparison of Meta-Analytic Methods
Carter, E. C., Schönbrodt, F. D., Gervais, W. M., & Hilgard, J. (2019) · Advances in Methods and Practices in Psychological Science · doi:10.1177/2515245919847196 · Cited by 588 · see citing works
Investigating the replicability of preclinical cancer biology
Errington, T. M., Mathur, M., Soderberg, C. K., Denis, A., Perfito, N., Iorns, E., & Nosek, B. A. (2021) · eLife · doi:10.7554/elife.71601 · Cited by 393 · see citing works
“For this purpose, inference from a fixed- effect model is appropriate (Rice et al., 2017).”
MethodOne-stage dose–response meta-analysis for aggregated data
Crippa, A., Discacciati, A., Bottai, M., Spiegelman, D., & Orsini, N. (2019) · Statistical Methods in Medical Research · doi:10.1177/0962280218773122 · Cited by 365 · see citing works
Methods for evidence synthesis in the case of very few studies
Bender, R., Friede, T., Koch, A., Kuss, O., Schlattmann, P., Schwarzer, G., & Skipka, G. (2018) · Research Synthesis Methods · doi:10.1002/jrsm.1297 · Cited by 250 · see citing works
Qu, H., Guo, M., Chai, H., Wang, W., Gao, Z., & Shi, D. (2018) · Journal of the American Heart Association · doi:10.1161/jaha.118.009835 · Cited by 178 · see citing works
“The fixed-effect model and randomeffect model were used in the meta-analysis.(20) We performed subgroup analysis and meta-regression to detect the potential sources of heterogeneity in the condition of I(2)≥50%.”
MethodCongenital heart disease and risk of cardiovascular disease: a meta-analysis of cohort studies
Wang, T., Chen, L., Yang, T., Huang, P., Wang, L., Zhao, L., Zhang, S., Ye, Z., Chen, L., Zheng, Z., et al. (2019) · Journal of the American Heart Association · doi:10.1161/jaha.119.012030 · Cited by 177 · see citing works
A historical review of publication bias
Marks‐Anglin, A., & Chen, Y. (2020) · Research Synthesis Methods · doi:10.1002/jrsm.1452 · Cited by 165 · see citing works
Bayesian Meta-Analysis with Weakly Informative Prior Distributions
Williams, D. R., Rast, P., & Bürkner, P. C. (2018) · doi:10.31234/osf.io/7tbrm · Cited by 124 · see citing works
Jenkins, D. J. A., Willett, W. C., Yusuf, S., Hu, F. B., Glenn, A. J., Liu, S., Mente, A., Miller, V., Bangdiwala, S. I., Gerstein, H. C., et al. (2024) · The Lancet Diabetes & Endocrinology · doi:10.1016/s2213-8587(23)00344-3 · Cited by 117 · see citing works
Liu, Q., Ayoub‐Charette, S., Khan, T. A., Au‐Yeung, F., Blanco Mejia, S., de Souza, R. J., Wolever, T. M. S., Leiter, L. A., Kendall, C. W. C., & Sievenpiper, J. L. (2019) · Journal of the American Heart Association · doi:10.1161/jaha.118.010977 · Cited by 65 · see citing works
Meta-Analysis of Complex Interventions
Tanner-Smith, E. E., & Grant, S. (2018) · Annual Review of Public Health · doi:10.1146/annurev-publhealth-040617-014112 · Cited by 62 · see citing works
Sex-specific DNA methylation differences in Alzheimer’s disease pathology
Zhang, L., Young, J. I., Gomez, L., Silva, T. C., Schmidt, M. A., Cai, J., Chen, X., Martin, E. R., & Wang, L. (2021) · Acta Neuropathologica Communications · doi:10.1186/s40478-021-01177-8 · Cited by 58 · see citing works
Tura, F., Buil-Gil, D., & Adeniyi, O. (2026) · Evidence Base · doi:10.1080/30679125.2025.2605330 · Cited by 2 · see citing works
- Authors
- Rice, K., Higgins, J. P. T., & Lumley, T.
- Affiliations
- University of Washington (US)
- University of Bristol (GB)
- University of Auckland (NZ)
- Corresponding author
- Rice, Kenneth
- Venue
- Journal of the Royal Statistical Society Series A: Statistics in Society
- Publisher
- Oxford University Press (OUP)
- Published
- 2017-03-16
- Volume
- 181
- Issue
- 1
- Pages
- 205-227
- ISSN
- 0964-1998 (ISSN-L)
- 1467-985X
- OpenAlex
- W2596968510
- Cited by
- 287
- References
- 13
- Citations by year
- 2026: 31 · 2025: 37 · 2024: 37 · 2023: 32 · 2022: 28 · 2021: 38 · 2020: 33 · 2019: 30 · 2018: 13 · 2017: 5 · 2016: 1
- FWCI
- 19.81
- Type
- journal article
- Language
- en
- Open access
- Licence
- Locations
- Journal of the Royal Statistical Society Series A (Statistics in Society) (publishedVersion)
- Bristol Research (University of Bristol) (submittedVersion, other-oa)
- Topics
- Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Bayesian Inference
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Epidemiology
- Keywords
- fixed-effects meta-analysis, inverse variance weighting, meta-regression, heterogeneity testing
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
