Meta-Analysis

Meta-analysis
A meta-analysis is a statistical analysis that

combines the results of multiple scientific
studies. Meta-analyses can be performed when
there are multiple scientific studies addressing
the same question, with each individual study
reporting measurements that are expected to
have some degree of error. The aim then is to
use approaches from statistics to derive a pooled
estimate closest to the unknown common truth
based on how this error is perceived. It is thus a
basic methodology of Metascience. Meta-
analytic results are considered the most
trustworthy source of evidence by the evidence- A meta-analysis is the highest form of
based medicine literature.[1][2][3] knowledge in science
Not only can meta-analyses provide an estimate

of the unknown effect size, it also has the capacity to contrast results from different
studies and identify patterns among study results, sources of disagreement among those
results, or other interesting relationships that may come to light with multiple studies.[4]
However, there are some methodological problems with meta-analysis. If individual

studies are systematically biased due to questionable research practices (e.g., data
dredging, data peeking, dropping studies) or the publication bias at the journal level, the
meta-analytic estimate of the overall treatment effect may not reflect the actual efficacy of
a treatment.[5][6] Meta-analysis has also been criticized for averaging differences among
heterogeneous studies because these differences could potentially inform clinical
decisions.[7] For example, if there are two groups of patients experiencing different
treatment effects studies in two randomised control trials (RCTs) reporting conflicting
results, the meta-analytic average is representative of neither group, similarly to
averaging the weight of apples and oranges, which is neither accurate for apples nor
oranges.[8] In performing a meta-analysis, an investigator must make choices which can
affect the results, including deciding how to search for studies, selecting studies based on
a set of objective criteria, dealing with incomplete data, analyzing the data, and
accounting for or choosing not to account for publication bias.[9] This makes meta-
analysis malleable in the sense that these methodological choices made in completing a
meta-analysis are not determined but may affect the results.[10] For example, Wanous
and colleagues examined four pairs of meta-analyses on the four topics of (a) job
performance and satisfaction relationship, (b) realistic job previews, (c) correlates of role
conflict and ambiguity, and (d) the job
satisfaction and absenteeism relationship,
and illustrated how various judgement
calls made by the researchers produced
different results.[11]
Meta-analyses are often, but not always,

important components of a systematic
review procedure. For instance, a meta-
analysis may be conducted on several
clinical trials of a medical treatment, in an
effort to obtain a better understanding of
how well the treatment works. Here it is
convenient to follow the terminology used
by the Cochrane Collaboration,[12] and use
"meta-analysis" to refer to statistical
methods of combining evidence, leaving Graphical summary of a meta-analysis of over
other aspects of 'research synthesis' or 1,000 cases of diffuse intrinsic pontine glioma and
'evidence synthesis', such as combining other pediatric gliomas, in which information about
information from qualitative studies, for the mutations involved as well as generic
the more general context of systematic outcomes were distilled from the underlying
reviews. A meta-analysis is a secondary primary literature.
[13][14]
source. In addition, meta-analysis
may also be applied to a single study in
cases where there are many cohorts which have not gone through identical selection
criteria or to which the same investigational methodologies have not been applied to all in
the same manner or under the same exacting conditions. Under these circumstances each
cohort is treated as an individual study and meta-analysis is used to draw study-wide
conclusions.[15]
History
The historical roots of meta-analysis can be traced back to 17th century studies of
astronomy,[16] while a paper published in 1904 by the statistician Karl Pearson in the
British Medical Journal[17] which collated data from several studies of typhoid
inoculation is seen as the first time a meta-analytic approach was used to aggregate the
outcomes of multiple clinical studies.[18][19] The first meta-analysis of all conceptually
identical experiments concerning a particular research issue, and conducted by
independent researchers, has been identified as the 1940 book-length publication
Extrasensory Perception After Sixty Years, authored by Duke University psychologists J.
G. Pratt, J. B. Rhine, and associates.[20] This encompassed a review of 145 reports on ESP
experiments published from 1882 to 1939, and included an estimate of the influence of
unpublished papers on the overall effect (the file-drawer problem). The term "meta-
analysis" was coined in 1976 by the statistician Gene V. Glass,[21][22][23] who stated "my
major interest currently is in what we have come to call ...the meta-analysis of research.
The term is a bit grand, but it is precise and apt ... Meta-analysis refers to the analysis
of analyses". Although this led to him being widely recognized as the modern founder of
the method, the methodology behind what he termed "meta-analysis" predates his work
by several decades.[24][25] But unlike earlier methods which aimed at aggregating study
results to achieve higher levels of statistical significance, Glass's work aimed at describing
aggregated measures of relationships and effects. The statistical theory surrounding
meta-analysis was greatly advanced by the work of Nambury S. Raju, Larry V. Hedges,
Harris Cooper, Ingram Olkin, John E. Hunter, Jacob Cohen, Thomas C. Chalmers, Robert
Rosenthal, Frank L. Schmidt, John E. Hunter, and Douglas G. Claurett.[23][26] In 1992,
meta-analysis was first applied to ecological questions[27] by Jessica Gurevitch who used
meta-analysis to study competition in field experiments.[28][29] The field of meta-analysis
has expanded greatly since the 1970s and touches multiple disciplines including
psychology, medicine, and ecology.[22] Further the more recent creation of evidence
synthesis communities has increased the cross pollination of ideas, methods, and the
creation of software tools across disciplines.[30][31][32]
Steps in a meta-analysis
A meta-analysis is usually preceded by a systematic review, as this allows identification
and critical appraisal of all the relevant evidence (thereby limiting the risk of bias in
summary estimates). The general steps are then as follows:[1]
1. Formulation of the research question, e.g. using the PICO model (Population,
Intervention, Comparison, Outcome).
2. Search of literature
3. Selection of studies ('incorporation criteria')
Based on quality criteria, e.g. the requirement of randomization and blinding in a
clinical trial
Selection of specific studies on a well-specified subject, e.g. the treatment of
breast cancer.
Decide whether unpublished studies are included to avoid publication bias (file
drawer problem)
4. Decide which dependent variables or summary measures are allowed. For instance,
when considering a meta-analysis of published (aggregate) data:
Differences (discrete data)

Means (continuous data)
Hedges' g is a popular summary measure for continuous data that is standardized
in order to eliminate scale differences, but it incorporates an index of variation
between groups:
in which is the treatment mean, is the control mean, the pooled variance.
5. Selection of a meta-analysis model, e.g. fixed effect or random effects meta-analysis.
6. Examine sources of between-study heterogeneity, e.g. using subgroup analysis or
meta-regression.
Formal guidance for the conduct and reporting of meta-analyses is provided by the
Cochrane Handbook (http://training.cochrane.org/handbook).
For reporting guidelines, see the Preferred Reporting Items for Systematic Reviews and
Meta-Analyses (PRISMA) statement.[33]
Methods and assumptions
Approaches
In general, two types of evidence can be distinguished when performing a meta-analysis:
individual participant data (IPD), and aggregate data (AD). The aggregate data can be
direct or indirect.
AD is more commonly available (e.g. from the literature) and typically represents
summary estimates such as odds ratios or relative risks. This can be directly synthesized
across conceptually similar studies using several approaches (see below). On the other
hand, indirect aggregate data measures the effect of two treatments that were each
compared against a similar control group in a meta-analysis. For example, if treatment A
and treatment B were directly compared vs placebo in separate meta-analyses, we can use
these two pooled results to get an estimate of the effects of A vs B in an indirect
comparison as effect A vs Placebo minus effect B vs Placebo.
IPD evidence represents raw data as collected by the study centers. This distinction has
raised the need for different meta-analytic methods when evidence synthesis is desired,
and has led to the development of one-stage and two-stage methods.[34] In one-stage
methods the IPD from all studies are modeled simultaneously whilst accounting for the
clustering of participants within studies. Two-stage methods first compute summary
statistics for AD from each study and then calculate overall statistics as a weighted
average of the study statistics. By reducing IPD to AD, two-stage methods can also be
applied when IPD is available; this makes them an appealing choice when performing a
meta-analysis. Although it is conventionally believed that one-stage and two-stage
methods yield similar results, recent studies have shown that they may occasionally lead
to different conclusions.[35][36]
Statistical models for aggregate data
Direct evidence: Models incorporating study effects only

Fixed effect model
The fixed effect model provides a weighted average of a series of study estimates. The
inverse of the estimates' variance is commonly used as study weight, so that larger studies
tend to contribute more than smaller studies to the weighted average. Consequently,
when studies within a meta-analysis are dominated by a very large study, the findings
from smaller studies are practically ignored.[37] Most importantly, the fixed effects model
assumes that all included studies investigate the same population, use the same variable
and outcome definitions, etc. This assumption is typically unrealistic as research is often
prone to several sources of heterogeneity; e.g. treatment effects may differ according to
locale, dosage levels, study conditions, ...
Random effects model
A common model used to synthesize heterogeneous research is the random effects model
of meta-analysis. This is simply the weighted average of the effect sizes of a group of
studies. The weight that is applied in this process of weighted averaging with a random
effects meta-analysis is achieved in two steps:[38]
1. Step 1: Inverse variance weighting

2. Step 2: Un-weighting of this inverse variance weighting by applying a random effects
variance component (REVC) that is simply derived from the extent of variability of the
effect sizes of the underlying studies.
This means that the greater this variability in effect sizes (otherwise known as
heterogeneity), the greater the un-weighting and this can reach a point when the random
effects meta-analysis result becomes simply the un-weighted average effect size across the
studies. At the other extreme, when all effect sizes are similar (or variability does not
exceed sampling error), no REVC is applied and the random effects meta-analysis
defaults to simply a fixed effect meta-analysis (only inverse variance weighting).
The extent of this reversal is solely dependent on two factors:[39]
1. Heterogeneity of precision
2. Heterogeneity of effect size
Since neither of these factors automatically indicates a faulty larger study or more reliable
smaller studies, the re-distribution of weights under this model will not bear a
relationship to what these studies actually might offer. Indeed, it has been demonstrated
that redistribution of weights is simply in one direction from larger to smaller studies as
heterogeneity increases until eventually all studies have equal weight and no more
redistribution is possible.[39] Another issue with the random effects model is that the
most commonly used confidence intervals generally do not retain their coverage
probability above the specified nominal level and thus substantially underestimate the
statistical error and are potentially overconfident in their conclusions.[40][41] Several fixes
have been suggested[42][43] but the debate continues on.[41][44] A further concern is that
the average treatment effect can sometimes be even less conservative compared to the
fixed effect model[45] and therefore misleading in practice. One interpretational fix that
has been suggested is to create a prediction interval around the random effects estimate
to portray the range of possible effects in practice.[46] However, an assumption behind
the calculation of such a prediction interval is that trials are considered more or less
homogeneous entities and that included patient populations and comparator treatments
should be considered exchangeable[47] and this is usually unattainable in practice.
There are many methods used to estimate between studies variance with restricted
maximum likelihood estimator being the least prone to bias and one of the most
commonly used.[48] Several advanced iterative techniques for computing the between
studies variance exist including both maximum likelihood and restricted maximum
likelihood method and random effects models using these methods can be run with
multiples software platforms including in Excel,[49] Stata,[50] SPSS,[51] and R.[52]
Most meta-analyses include between 2 and 4 studies and such a sample is more often
than not inadequate to accurately estimate heterogeneity. Thus it appears that in small
meta-analyses, an incorrect zero between study variance estimate is obtained, leading to a
false homogeneity assumption. Overall, it appears that heterogeneity is being consistently
underestimated in meta-analyses and sensitivity analyses in which high heterogeneity
levels are assumed could be informative.[53] These random effects models and software
packages mentioned above relate to study-aggregate meta-analyses and researchers
wishing to conduct individual patient data (IPD) meta-analyses need to consider mixed-
effects modelling approaches.[54]
IVhet model
Doi & Barendregt working in collaboration with Khan, Thalib and Williams (from the
University of Queensland, University of Southern Queensland and Kuwait University),
have created an inverse variance quasi likelihood based alternative (IVhet) to the random
effects (RE) model for which details are available online.[49] This was incorporated into
MetaXL version 2.0,[55] a free Microsoft excel add-in for meta-analysis produced by
Epigear International Pty Ltd, and made available on 5 April 2014. The authors state that
a clear advantage of this model is that it resolves the two main problems of the random
effects model. The first advantage of the IVhet model is that coverage remains at the
nominal (usually 95%) level for the confidence interval unlike the random effects model
which drops in coverage with increasing heterogeneity.[40][41] The second advantage is
that the IVhet model maintains the inverse variance weights of individual studies, unlike
the RE model which gives small studies more weight (and therefore larger studies less)
with increasing heterogeneity. When heterogeneity becomes large, the individual study
weights under the RE model become equal and thus the RE model returns an arithmetic
mean rather than a weighted average. This side-effect of the RE model does not occur
with the IVhet model which thus differs from the RE model estimate in two
perspectives:[49] Pooled estimates will favor larger trials (as opposed to penalizing larger
trials in the RE model) and will have a confidence interval that remains within the
nominal coverage under uncertainty (heterogeneity). Doi & Barendregt suggest that while
the RE model provides an alternative method of pooling the study data, their simulation
results[56] demonstrate that using a more specified probability model with untenable
assumptions, as with the RE model, does not necessarily provide better results. The latter
study also reports that the IVhet model resolves the problems related to underestimation
of the statistical error, poor coverage of the confidence interval and increased MSE seen
with the random effects model and the authors conclude that researchers should
henceforth abandon use of the random effects model in meta-analysis. While their data is
compelling, the ramifications (in terms of the magnitude of spuriously positive results
within the Cochrane database) are huge and thus accepting this conclusion requires
careful independent confirmation. The availability of a free software (MetaXL)[55] that
runs the IVhet model (and all other models for comparison) facilitates this for the
research community.
Direct evidence: Models incorporating additional information
Quality effects model
Doi and Thalib originally introduced the quality effects model.[57] They[58] introduced a
new approach to adjustment for inter-study variability by incorporating the contribution
of variance due to a relevant component (quality) in addition to the contribution of
variance due to random error that is used in any fixed effects meta-analysis model to
generate weights for each study. The strength of the quality effects meta-analysis is that it
allows available methodological evidence to be used over subjective random effects, and
thereby helps to close the damaging gap which has opened up between methodology and
statistics in clinical research. To do this a synthetic bias variance is computed based on
quality information to adjust inverse variance weights and the quality adjusted weight of
the ith study is introduced.[57] These adjusted weights are then used in meta-analysis. In
other words, if study i is of good quality and other studies are of poor quality, a
proportion of their quality adjusted weights is mathematically redistributed to study i
giving it more weight towards the overall effect size. As studies become increasingly
similar in terms of quality, re-distribution becomes progressively less and ceases when all
studies are of equal quality (in the case of equal quality, the quality effects model defaults
to the IVhet model – see previous section). A recent evaluation of the quality effects
model (with some updates) demonstrates that despite the subjectivity of quality
assessment, the performance (MSE and true variance under simulation) is superior to
that achievable with the random effects model.[59][60] This model thus replaces the
untenable interpretations that abound in the literature and a software is available to
explore this method further.[55]
Indirect evidence: Network meta-analysis methods

Indirect comparison meta-analysis methods (also called network meta-analyses, in
particular when multiple treatments are assessed simultaneously) generally use two main
methodologies. First, is the Bucher method[61] which is a single or repeated comparison
of a closed loop of three-treatments such
that one of them is common to the two
studies and forms the node where the loop
begins and ends. Therefore, multiple two-
by-two comparisons (3-treatment loops)
are needed to compare multiple
treatments. This methodology requires
that trials with more than two arms have
two arms only selected as independent
pair-wise comparisons are required. The A network meta-analysis looks at indirect
alternative methodology uses complex comparisons. In the image, A has been analyzed in
statistical modelling to include the relation to C and C has been analyzed in relation to
multiple arm trials and comparisons b. However the relation between A and B is only
simultaneously between all competing known indirectly, and a network meta-analysis
treatments. These have been executed looks at such indirect evidence of differences
using Bayesian methods, mixed linear between methods and interventions using
models and meta-regression approaches. statistical method.
Bayesian framework
Specifying a Bayesian network meta-analysis model involves writing a directed acyclic

graph (DAG) model for general-purpose Markov chain Monte Carlo (MCMC) software
such as WinBUGS.[62] In addition, prior distributions have to be specified for a number of
the parameters, and the data have to be supplied in a specific format.[62] Together, the
DAG, priors, and data form a Bayesian hierarchical model. To complicate matters further,
because of the nature of MCMC estimation, overdispersed starting values have to be
chosen for a number of independent chains so that convergence can be assessed.[63]
Recently, multiple R software packages were developed to simplify the model fitting (e.g.,
metaBMA[64] and RoBMA[65]) and even implemented in statistical software with
graphical user interface (GUI): JASP. Although the complexity of the Bayesian approach
limits usage of this methodology, recent tutorial papers are trying to increase accessibility
of the methods.[66][67] Methodology for automation of this method has been
suggested[62] but requires that arm-level outcome data are available, and this is usually
unavailable. Great claims are sometimes made for the inherent ability of the Bayesian
framework to handle network meta-analysis and its greater flexibility. However, this
choice of implementation of framework for inference, Bayesian or frequentist, may be less
important than other choices regarding the modeling of effects[68] (see discussion on
models above).
Frequentist multivariate framework

On the other hand, the frequentist multivariate methods involve approximations and
assumptions that are not stated explicitly or verified when the methods are applied (see
discussion on meta-analysis models above). For example, the mvmeta package for Stata
enables network meta-analysis in a frequentist framework.[69] However, if there is no
common comparator in the network, then this has to be handled by augmenting the
dataset with fictional arms with high variance, which is not very objective and requires a
decision as to what constitutes a sufficiently high variance.[62] The other issue is use of
the random effects model in both this frequentist framework and the Bayesian
framework. Senn advises analysts to be cautious about interpreting the 'random effects'
analysis since only one random effect is allowed for but one could envisage many.[68]
Senn goes on to say that it is rather naıve, even in the case where only two treatments are
being compared to assume that random-effects analysis accounts for all uncertainty about
the way effects can vary from trial to trial. Newer models of meta-analysis such as those
discussed above would certainly help alleviate this situation and have been implemented
in the next framework.
Generalized pairwise modelling framework
An approach that has been tried since the late 1990s is the implementation of the
multiple three-treatment closed-loop analysis. This has not been popular because the
process rapidly becomes overwhelming as network complexity increases. Development in
this area was then abandoned in favor of the Bayesian and multivariate frequentist
methods which emerged as alternatives. Very recently, automation of the three-treatment
closed loop method has been developed for complex networks by some researchers[49] as
a way to make this methodology available to the mainstream research community. This
proposal does restrict each trial to two interventions, but also introduces a workaround
for multiple arm trials: a different fixed control node can be selected in different runs. It
also utilizes robust meta-analysis methods so that many of the problems highlighted
above are avoided. Further research around this framework is required to determine if
this is indeed superior to the Bayesian or multivariate frequentist frameworks.
Researchers willing to try this out have access to this framework through a free
software.[55]
Tailored meta-analysis
Another form of additional information comes from the intended setting. If the target
setting for applying the meta-analysis results is known then it may be possible to use data
from the setting to tailor the results thus producing a 'tailored meta-analysis'.,[70][71] This
has been used in test accuracy meta-analyses, where empirical knowledge of the test
positive rate and the prevalence have been used to derive a region in Receiver Operating
Characteristic (ROC) space known as an 'applicable region'. Studies are then selected for
the target setting based on comparison with this region and aggregated to produce a
summary estimate which is tailored to the target setting.
Aggregating IPD and AD

Meta-analysis can also be applied to combine IPD and AD. This is convenient when the
researchers who conduct the analysis have their own raw data while collecting aggregate
or summary data from the literature. The generalized integration model (GIM)[72] is a
generalization of the meta-analysis. It allows that the model fitted on the individual
participant data (IPD) is different from the ones used to compute the aggregate data
(AD). GIM can be viewed as a model calibration method for integrating information with
more flexibility.
Validation of meta-analysis results
The meta-analysis estimate represents a weighted average across studies and when there
is heterogeneity this may result in the summary estimate not being representative of
individual studies. Qualitative appraisal of the primary studies using established tools can
uncover potential biases,[73][74] but does not quantify the aggregate effect of these biases
on the summary estimate. Although the meta-analysis result could be compared with an
independent prospective primary study, such external validation is often impractical. This
has led to the development of methods that exploit a form of leave-one-out cross
validation, sometimes referred to as internal-external cross validation (IOCV).[75] Here
each of the k included studies in turn is omitted and compared with the summary
estimate derived from aggregating the remaining k- 1 studies. A general validation
statistic, Vn based on IOCV has been developed to measure the statistical validity of
meta-analysis results.[76] For test accuracy and prediction, particularly when there are
multivariate effects, other approaches which seek to estimate the prediction error have
also been proposed.[77]
Challenges
A meta-analysis of several small studies does not always predict the results of a single
large study.[78] Some have argued that a weakness of the method is that sources of bias
are not controlled by the method: a good meta-analysis cannot correct for poor design or
bias in the original studies.[79] This would mean that only methodologically sound studies
should be included in a meta-analysis, a practice called 'best evidence synthesis'.[79]
Other meta-analysts would include weaker studies, and add a study-level predictor
variable that reflects the methodological quality of the studies to examine the effect of
study quality on the effect size.[80] However, others have argued that a better approach is
to preserve information about the variance in the study sample, casting as wide a net as
possible, and that methodological selection criteria introduce unwanted subjectivity,
defeating the purpose of the approach.[81]
Publication bias: the file drawer problem

Another potential pitfall is the reliance on the available body of published studies, which
may create exaggerated outcomes due to publication bias, as studies which show negative
results or insignificant results are less likely to be published.[82] For example,
pharmaceutical companies have been known to hide
negative studies and researchers may have overlooked
unpublished studies such as dissertation studies or
conference abstracts that did not reach publication.
This is not easily solved, as one cannot know how
many studies have gone unreported.[83]
This file drawer problem (characterized by negative or

non-significant results being tucked away in a A funnel plot expected without the file
cabinet), can result in a biased distribution of effect drawer problem. The largest studies
sizes thus creating a serious base rate fallacy, in which converge at the tip while smaller
the significance of the published studies is studies show more or less
overestimated, as other studies were either not symmetrical scatter at the base
submitted for publication or were rejected. This
should be seriously considered when interpreting the
outcomes of a meta-analysis.[83][6]
The distribution of effect sizes can be visualized with a

funnel plot which (in its most common version) is a
scatter plot of standard error versus the effect size. It
makes use of the fact that the smaller studies (thus
larger standard errors) have more scatter of the
magnitude of effect (being less precise) while the A funnel plot expected with the file
larger studies have less scatter and form the tip of the drawer problem. The largest studies
funnel. If many negative studies were not published, still cluster around the tip, but the
the remaining positive studies give rise to a funnel bias against publishing negative
plot in which the base is skewed to one side studies has caused the smaller
(asymmetry of the funnel plot). In contrast, when studies as a whole to have an
unjustifiably favorable result to the
there is no publication bias, the effect of the smaller
hypothesis
studies has no reason to be skewed to one side and so
a symmetric funnel plot results. This also means that
if no publication bias is present, there would be no relationship between standard error
and effect size.[84] A negative or positive relation between standard error and effect size
would imply that smaller studies that found effects in one direction only were more likely
to be published and/or to be submitted for publication.
Apart from the visual funnel plot, statistical methods for detecting publication bias have
also been proposed.[85] These are controversial because they typically have low power for
detection of bias, but also may make false positives under some circumstances.[86] For
instance small study effects (biased smaller studies), wherein methodological differences
between smaller and larger studies exist, may cause asymmetry in effect sizes that
resembles publication bias. However, small study effects may be just as problematic for
the interpretation of meta-analyses, and the imperative is on meta-analytic authors to
investigate potential sources of bias.[87]
A Tandem Method for analyzing publication bias has been suggested for cutting down
false positive error problems.[88] This Tandem method consists of three stages. Firstly,
one calculates Orwin's fail-safe N, to check how many studies should be added in order to
reduce the test statistic to a trivial size. If this number of studies is larger than the number
of studies used in the meta-analysis, it is a sign that there is no publication bias, as in that
case, one needs a lot of studies to reduce the effect size. Secondly, one can do an Egger's
regression test, which tests whether the funnel plot is symmetrical. As mentioned before:
a symmetrical funnel plot is a sign that there is no publication bias, as the effect size and
sample size are not dependent. Thirdly, one can do the trim-and-fill method, which
imputes data if the funnel plot is asymmetrical.
The problem of publication bias is not trivial as it is suggested that 25% of meta-analyses
in the psychological sciences may have suffered from publication bias.[88] However, low
power of existing tests and problems with the visual appearance of the funnel plot remain
an issue, and estimates of publication bias may remain lower than what truly exists.
Most discussions of publication bias focus on journal practices favoring publication of

statistically significant findings. However, questionable research practices, such as
reworking statistical models until significance is achieved, may also favor statistically
significant findings in support of researchers' hypotheses.[89][90]
Problems related to studies not reporting non-statistically significant

effects
Studies often do not report the effects when they do not reach statistical significance. For
example, they may simply say that the groups did not show statistically significant
differences, without reporting any other information (e.g. a statistic or p-value).
Exclusion of these studies would lead to a situation similar to publication bias, but their
inclusion (assuming null effects) would also bias the meta-analysis. MetaNSUE, a method
created by Joaquim Radua, has shown to allow researchers to include unbiasedly these
studies.[91] Its steps are as follows:
Maximum likelihood estimation of the meta-analytic effect and the heterogeneity

between studies.
Multiple imputation of the NSUEs adding noise to the estimate of the effect.
Separate meta-analyses for each imputed dataset.
Pooling of the results of these meta-analyses.
Problems related to the statistical approach
Other weaknesses are that it has not been determined if the statistically most accurate
method for combining results is the fixed, IVhet, random or quality effect models, though
the criticism against the random effects model is mounting because of the perception that
the new random effects (used in meta-analysis) are essentially formal devices to facilitate
smoothing or shrinkage and prediction may be impossible or ill-advised.[92] The main
problem with the random effects approach is that it uses the classic statistical thought of
generating a "compromise estimator" that makes the weights close to the naturally
weighted estimator if heterogeneity across studies is large but close to the inverse
variance weighted estimator if the between study heterogeneity is small. However, what
has been ignored is the distinction between the model we choose to analyze a given
dataset, and the mechanism by which the data came into being.[93] A random effect can
be present in either of these roles, but the two roles are quite distinct. There's no reason
to think the analysis model and data-generation mechanism (model) are similar in form,
but many sub-fields of statistics have developed the habit of assuming, for theory and
simulations, that the data-generation mechanism (model) is identical to the analysis
model we choose (or would like others to choose). As a hypothesized mechanisms for
producing the data, the random effect model for meta-analysis is silly and it is more
appropriate to think of this model as a superficial description and something we choose
as an analytical tool – but this choice for meta-analysis may not work because the study
effects are a fixed feature of the respective meta-analysis and the probability distribution
is only a descriptive tool.[93]
Problems arising from agenda-driven bias
The most severe fault in meta-analysis often occurs when the person or persons doing the
meta-analysis have an economic, social, or political agenda such as the passage or defeat
of legislation. People with these types of agendas may be more likely to abuse meta-
analysis due to personal bias. For example, researchers favorable to the author's agenda
are likely to have their studies cherry-picked while those not favorable will be ignored or
labeled as "not credible". In addition, the favored authors may themselves be biased or
paid to produce results that support their overall political, social, or economic goals in
ways such as selecting small favorable data sets and not incorporating larger unfavorable
data sets. The influence of such biases on the results of a meta-analysis is possible
because the methodology of meta-analysis is highly malleable.[10]
A 2011 study done to disclose possible conflicts of interests in underlying research studies
used for medical meta-analyses reviewed 29 meta-analyses and found that conflicts of
interests in the studies underlying the meta-analyses were rarely disclosed. The 29 meta-
analyses included 11 from general medicine journals, 15 from specialty medicine journals,
and three from the Cochrane Database of Systematic Reviews. The 29 meta-analyses
reviewed a total of 509 randomized controlled trials (RCTs). Of these, 318 RCTs reported
funding sources, with 219 (69%) receiving funding from industry (i.e. one or more
authors having financial ties to the pharmaceutical industry). Of the 509 RCTs, 132
reported author conflict of interest disclosures, with 91 studies (69%) disclosing one or
more authors having financial ties to industry. The information was, however, seldom
reflected in the meta-analyses. Only two (7%) reported RCT funding sources and none
reported RCT author-industry ties. The authors concluded "without acknowledgment of
COI due to industry funding or author industry financial ties from RCTs included in
meta-analyses, readers' understanding and appraisal of the evidence from the meta-
analysis may be compromised."[94]
For example, in 1998, a US federal judge found that the United States Environmental
Protection Agency had abused the meta-analysis process to produce a study claiming
cancer risks to non-smokers from environmental tobacco smoke (ETS) with the intent to
influence policy makers to pass smoke-free–workplace laws. The judge found that:
EPA's study selection is disturbing. First, there is evidence in the record

supporting the accusation that EPA "cherry picked" its data. Without criteria
for pooling studies into a meta-analysis, the court cannot determine whether
the exclusion of studies likely to disprove EPA's a priori hypothesis was
coincidence or intentional. Second, EPA's excluding nearly half of the
available studies directly conflicts with EPA's purported purpose for analyzing
the epidemiological studies and conflicts with EPA's Risk Assessment
Guidelines. See ETS Risk Assessment at 4-29 ("These data should also be
examined in the interest of weighing all the available evidence, as
recommended by EPA's carcinogen risk assessment guidelines (U.S. EPA,
1986a) (emphasis added)). Third, EPA's selective use of data conflicts with the
Radon Research Act. The Act states EPA's program shall "gather data and
information on all aspects of indoor air quality" (Radon Research Act § 403(a)
(1)) (emphasis added).[95]
As a result of the abuse, the court vacated Chapters 1–6 of and the Appendices to EPA's
"Respiratory Health Effects of Passive Smoking: Lung Cancer and other Disorders".[95]
Comparability and validity of included studies
Meta-analysis may often not be a substitute for an adequately powered primary study.[96]
Heterogeneity of methods used may lead to faulty conclusions.[97] For instance,

differences in the forms of an intervention or the cohorts that are thought to be minor or
are unknown to the scientists could lead to substantially different results, including
results that distort the meta-analysis' results or are not adequately considered in its data.
Vice versa, results from meta-analyses may also make certain hypothesis or interventions
seem nonviable and preempt further research or approvals, despite certain modifications
– such as intermittent administration, personalized criteria and combination measures –
leading to substantially different results, including in cases where such have been
successfully identified and applied in small-scale studies that were considered in the
meta-analysis. Standardization, reproduction of experiments, open data and open
protocols may often not mitigate such problems, for instance as relevant factors and
criteria could be unknown or not be recorded.
There is a debate about the appropriate balance between testing with as few animals or
humans as possible and the need to obtain robust, reliable findings. It has been argued
that unreliable research is inefficient and wasteful and that studies are not just wasteful
when they stop too late but also when they stop too early. In large clinical trials, planned,
sequential analyses are sometimes used if there is considerable expense or potential harm
associated with testing participants.[98] In applied behavioural science, "megastudies"
have been proposed to investigate the efficacy of many different interventions designed in
an interdisciplinary manner by separate teams.[99] One such study used a fitness chain to
recruit a large number participants. It has been suggested that behavioural interventions
are often hard to compare [in meta-analyses and reviews], as "different scientists test
different intervention ideas in different samples using different outcomes over different
time intervals", causing a lack of comparability of such individual investigations which
limits "their potential to inform policy".[99]
Weak inclusion standards lead to misleading conclusions

Meta-analyses in education are often not restrictive enough in regards to the
methodological quality of the studies they include. For example, studies that include
small samples or researcher-made measures lead to inflated effect size estimates.[100]
However, this problem also troubles meta-analysis of clinical trials. The use of different
quality assessment tools (QATs) lead to including different studies and obtaining
conflicting estimates of average treatment effects.[101][102]
Applications in modern science

Modern statistical meta-analysis does more than just combine the effect sizes of a set of
studies using a weighted average. It can test if the outcomes of studies show more
variation than the variation that is expected because of the sampling of different numbers
of research participants. Additionally, study characteristics such as measurement
instrument used, population sampled, or aspects of the studies' design can be coded and
used to reduce variance of the estimator (see statistical models above). Thus some
methodological weaknesses in studies can be corrected statistically. Other uses of meta-
analytic methods include the development and validation of clinical prediction models,
where meta-analysis may be used to combine individual participant data from different
research centers and to assess the model's generalisability,[103][104] or even to aggregate
existing prediction models.[105]
Meta-analysis can be done with single-subject design as well as group research

designs.[106] This is important because much research has been done with single-subject
research designs.[107] Considerable dispute exists for the most appropriate meta-analytic
technique for single subject research.[108]
Meta-analysis leads to a shift of emphasis from single studies to multiple studies. It

emphasizes the practical importance of the effect size instead of the statistical significance
of individual studies. This shift in thinking has been termed "meta-analytic thinking". The
results of a meta-analysis are often shown in a forest plot.
Results from studies are combined using different approaches. One approach frequently
used in meta-analysis in health care research is termed 'inverse variance method'. The
average effect size across all studies is computed as a weighted mean, whereby the
weights are equal to the inverse variance of each study's effect estimator. Larger studies
and studies with less random variation are given greater weight than smaller studies.
Other common approaches include the Mantel–Haenszel method[109] and the Peto
method.[110]
Seed-based d mapping (formerly signed differential mapping, SDM) is a statistical

technique for meta-analyzing studies on differences in brain activity or structure which
used neuroimaging techniques such as fMRI, VBM or PET.
Different high throughput techniques such as microarrays have been used to understand
Gene expression. MicroRNA expression profiles have been used to identify differentially
expressed microRNAs in particular cell or tissue type or disease conditions or to check the
effect of a treatment. A meta-analysis of such expression profiles was performed to derive
novel conclusions and to validate the known findings.[111]
Meta-analysis of whole genome sequencing studies provides an attractive solution to the

problem of collecting large sample sizes for discovering rare variants associated with
complex phenotypes. Some methods have been developed to enable functionally
informed rare variant association meta-analysis in biobank-scale cohorts using efficient
approaches for summary statistic storage.[112]
See also
Estimation statistics
Mathematics portal
Metascience
Newcastle–Ottawa scale
Reporting bias
Review journal
Secondary research
Study heterogeneity
Systematic review
Galbraith plot
Data aggregation
References
1. Herrera Ortiz AF., Cadavid Camacho E, Cubillos Rojas J, Cadavid Camacho T, Zoe
Guevara S, Tatiana Rincón Cuenca N, Vásquez Perdomo A, Del Castillo Herazo V, &
Giraldo Malo R. A Practical Guide to Perform a Systematic Literature Review and
Meta-analysis. Principles and Practice of Clinical Research. 2022;7(4):47–57.
https://doi.org/10.21801/ppcrj.2021.74.6
2. "Levels of Evidence" (https://www.cebm.ox.ac.uk/resources/levels-of-evidence/oxford-
centre-for-evidence-based-medicine-levels-of-evidence-march-2009). Centre for
Evidence-Based Medicine (CEBM). University of Oxford. March 2009. Retrieved
21 December 2021.
21 December 2021.
3. Developing NICE guidelines: the manual (https://www.nice.org.uk/media/default/about
/what-we-do/our-programmes/developing-nice-guidelines-the-manual.pdf) (PDF).
London: National Institute for Health and Care Excellence. 2014. Retrieved
21 December 2020.
4. Greenland S, O' Rourke K (2008). "Meta-Analysis". In Rothman KJ, Greenland S,
Lash T (eds.). Modern epidemiology (3rd, thoroughly revised and updated ed.).
Philadelphia: Lippincott Williams and Wilkins. p. 652. ISBN 978-0-7817-5564-1.
5. Tabery J (9 May 2014). Beyond Versus: The Struggle to Understand the Interaction of
Nature and Nurture (https://books.google.com/books?id=aeCOAwAAQBAJ&pg=PR9).
MIT Press. ISBN 978-0-262-02737-3.
6. Hunter JE, Schmidt FL (1990). Methods of Meta-Analysis: Correcting Error and Bias
in Research Findings. Newbury Park, California; London; New Delhi: SAGE
Publications.
7. Maziarz M (February 2022). "Is meta-analysis of RCTs assessing the efficacy of
interventions a reliable source of evidence for therapeutic decisions?" (https://doi.org/
10.1016%2Fj.shpsa.2021.11.007). Studies in History and Philosophy of Science. 91:
159–167. doi:10.1016/j.shpsa.2021.11.007 (https://doi.org/10.1016%2Fj.shpsa.2021.1
1.007). PMID 34922183 (https://pubmed.ncbi.nlm.nih.gov/34922183).
S2CID 245241150 (https://api.semanticscholar.org/CorpusID:245241150).
8. Feinstein AR (January 1995). "Meta-analysis: statistical alchemy for the 21st century"
(https://doi.org/10.1016%2F0895-4356%2894%2900110-C). Journal of Clinical
Epidemiology. The Potsdam International Consultation on Meta-Analysis. 48 (1): 71–
79. doi:10.1016/0895-4356(94)00110-C (https://doi.org/10.1016%2F0895-4356%2894
%2900110-C). PMID 7853050 (https://pubmed.ncbi.nlm.nih.gov/7853050).
9. Walker E, Hernandez AV, Kattan MW (June 2008). "Meta-analysis: Its strengths and
limitations". Cleveland Clinic Journal of Medicine. 75 (6): 431–439.
doi:10.3949/ccjm.75.6.431 (https://doi.org/10.3949%2Fccjm.75.6.431).
PMID 18595551 (https://pubmed.ncbi.nlm.nih.gov/18595551). S2CID 22543097 (https
://api.semanticscholar.org/CorpusID:22543097).
10. Stegenga J (December 2011). "Is meta-analysis the platinum standard of evidence?".
Studies in History and Philosophy of Biological and Biomedical Sciences. 42 (4): 497–
507. doi:10.1016/j.shpsc.2011.07.003 (https://doi.org/10.1016%2Fj.shpsc.2011.07.00
3). PMID 22035723 (https://pubmed.ncbi.nlm.nih.gov/22035723).
11. Wanous JP, Sullivan SE, Malinak J (1989). "The role of judgment calls in meta-
analysis". Journal of Applied Psychology. 74 (2): 259–264. doi:10.1037/0021-
9010.74.2.259 (https://doi.org/10.1037%2F0021-9010.74.2.259). ISSN 0021-9010 (htt
ps://www.worldcat.org/issn/0021-9010).
12. "Glossary: meta-analysis" (https://web.archive.org/web/20150126173543/http://www.c
ochrane.org/glossary/5#letterm). The Cochrane Collaboration. Archived from the
original (http://www.cochrane.org/glossary/5#letterm) on 20 December 2014.
13. Gravetter FJ, Forzano LB (1 January 2018). Research Methods for the Behavioral
Sciences (https://books.google.com/books?id=c69EDwAAQBAJ&pg=PA36). Cengage
Learning. p. 36. ISBN 978-1-337-61331-6. "Some examples of secondary sources are
(1) books and textbooks in which the author describes and summarizes past
research, (2) review articles or meta-analyses..."
14. Adams KA, Lawrence EK (2 February 2018). Research Methods, Statistics, and
Applications (https://books.google.com/books?id=K35ZDwAAQBAJ&pg=PT84).
SAGE Publications. ISBN 978-1-5063-5046-2. "The most common types of
SAGE Publications. ISBN 978-1-5063-5046-2. "The most common types of
secondary sources found in academic journals are literature reviews and meta-
analyses."
15. Rühlemann MC, Hermes BM, Bang C, Doms S, Moitinho-Silva L, Thingholm LB, et al.
(February 2021). "Genome-wide association study in 8,956 German individuals
identifies influence of ABO histo-blood groups on gut microbiome". Nature Genetics.
53 (2): 147–155. doi:10.1038/s41588-020-00747-1 (https://doi.org/10.1038%2Fs4158
8-020-00747-1). PMID 33462482 (https://pubmed.ncbi.nlm.nih.gov/33462482).
16. Plackett RL (1958). "Studies in the History of Probability and Statistics: Vii. The
Principle of the Arithmetic Mean". Biometrika. 45 (1–2): 133. doi:10.1093/biomet/45.1-
2.130 (https://doi.org/10.1093%2Fbiomet%2F45.1-2.130).
17. "Report on Certain Enteric Fever Inoculation Statistics" (https://www.ncbi.nlm.nih.gov/
pmc/articles/PMC2355479). British Medical Journal. 2 (2288): 1243–1246. November
1904. doi:10.1136/bmj.2.2288.1243 (https://doi.org/10.1136%2Fbmj.2.2288.1243).
PMC 2355479 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2355479).
PMID 20761760 (https://pubmed.ncbi.nlm.nih.gov/20761760).
18. Nordmann AJ, Kasenda B, Briel M (9 March 2012). "Meta-analyses: what they can
and cannot do" (https://doi.org/10.4414%2Fsmw.2012.13518). Swiss Medical Weekly.
142: w13518. doi:10.4414/smw.2012.13518 (https://doi.org/10.4414%2Fsmw.2012.13
19. O'Rourke K (December 2007). "An historical perspective on meta-analysis: dealing
quantitatively with varying study results" (https://www.ncbi.nlm.nih.gov/pmc/articles/P
MC2121629). Journal of the Royal Society of Medicine. 100 (12): 579–582.
doi:10.1177/0141076807100012020 (https://doi.org/10.1177%2F01410768071000120
20). PMC 2121629 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2121629).
20. Pratt JG, Rhine JB, Smith BM, Stuart CE, Greenwood JA. Extra-Sensory Perception
after Sixty Years: A Critical Appraisal of the Research in Extra-Sensory Perception.
New York: Henry Holt, 1940
21. Glass GV (1976). "Primary, secondary, and meta-analysis of research". Educational
Researcher. 5 (10): 3–8. doi:10.3102/0013189X005010003 (https://doi.org/10.3102%
2F0013189X005010003). S2CID 3185455 (https://api.semanticscholar.org/CorpusID:
3185455).
22. Shadish WR, Lecy JD (September 2015). "The meta-analytic big bang". Research
Synthesis Methods. 6 (3): 246–264. doi:10.1002/jrsm.1132 (https://doi.org/10.1002%2
Fjrsm.1132). PMID 26212600 (https://pubmed.ncbi.nlm.nih.gov/26212600).
23. Glass GV (September 2015). "Meta-analysis at middle age: a personal history".
Research Synthesis Methods. 6 (3): 221–231. doi:10.1002/jrsm.1133 (https://doi.org/1
0.1002%2Fjrsm.1133). PMID 26355796 (https://pubmed.ncbi.nlm.nih.gov/26355796).
24. Cochran WG (1937). "Problems Arising in the Analysis of a Series of Similar
Experiments". Journal of the Royal Statistical Society. 4 (1): 102–118.
doi:10.2307/2984123 (https://doi.org/10.2307%2F2984123). JSTOR 2984123 (https://
www.jstor.org/stable/2984123).
25. Cochran WG, Carroll SP (1953). "A Sampling Investigation of the Efficiency of
Weighting Inversely as the Estimated Variance". Biometrics. 9 (4): 447–459.
www.jstor.org/stable/3001436).
26. Hedges LV (September 2015). "The early history of meta-analysis". Research
Synthesis Methods. 6 (3): 284–286. doi:10.1002/jrsm.1149 (https://doi.org/10.1002%2
Fjrsm.1149). PMID 26097046 (https://pubmed.ncbi.nlm.nih.gov/26097046).
27. Gurevitch J, Koricheva J, Nakagawa S, Stewart G (March 2018). "Meta-analysis and
the science of research synthesis". Nature. 555 (7695): 175–182.
Bibcode:2018Natur.555..175G (https://ui.adsabs.harvard.edu/abs/2018Natur.555..175
G). doi:10.1038/nature25753 (https://doi.org/10.1038%2Fnature25753).
PMID 29517004 (https://pubmed.ncbi.nlm.nih.gov/29517004). S2CID 3761687 (https:/
/api.semanticscholar.org/CorpusID:3761687).
28. Gurevitch J, Morrow LL, Wallace A, Walsh JS (1992). "A Meta-Analysis of
Competition in Field Experiments". The American Naturalist. 140 (4): 539–572.
doi:10.1086/285428 (https://doi.org/10.1086%2F285428). ISSN 0003-0147 (https://w
ww.worldcat.org/issn/0003-0147). JSTOR 2462913 (https://www.jstor.org/stable/2462
913). S2CID 205982939 (https://api.semanticscholar.org/CorpusID:205982939).
29. Gurevitch K (7 March 2018). "Forty years of meta-analysis: We need evidence-based
answers more than ever" (https://researchdata.springernature.com/posts/30896-forty-
years-of-meta-analysis-we-need-evidence-based-answers-more-than-ever).
Research Data at Springer Nature. Retrieved 24 July 2021.
30. Vandvik PO, Brandt L (July 2020). "Future of Evidence Ecosystem Series: Evidence
ecosystems and learning health systems: why bother?". Journal of Clinical
Epidemiology. 123: 166–170. doi:10.1016/j.jclinepi.2020.02.008 (https://doi.org/10.10
16%2Fj.jclinepi.2020.02.008). PMID 32145365 (https://pubmed.ncbi.nlm.nih.gov/3214
31. Cartabellotta A, Tilson JK (June 2019). "The ecosystem of evidence cannot thrive
without efficiency of knowledge generation, synthesis, and translation". Journal of
Clinical Epidemiology. 110: 90–95. doi:10.1016/j.jclinepi.2019.01.008 (https://doi.org/1
0.1016%2Fj.jclinepi.2019.01.008). PMID 30708174 (https://pubmed.ncbi.nlm.nih.gov/
32. Haddaway NR, Bannach-Brown A, Grainger MJ, Hamilton WK, Hennessy EA,
Keenan C, et al. (June 2022). "The evidence synthesis and meta-analysis in R
conference (ESMARConf): levelling the playing field of conference accessibility and
equitability" (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9164457). Systematic
Reviews. 11 (1): 113. doi:10.1186/s13643-022-01985-6 (https://doi.org/10.1186%2Fs1
3643-022-01985-6). PMC 9164457 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC91
33. "The PRISMA statement" (https://web.archive.org/web/20110727021816/http://www.pr
isma-statement.org/). Prisma-statement.org. 2 February 2012. Archived from the
original (http://www.prisma-statement.org/) on 27 July 2011. Retrieved 2 February
2012.
34. Debray TP, Moons KG, van Valkenhoef G, Efthimiou O, Hummel N, Groenwold RH,
Reitsma JB (December 2015). "Get real in individual participant data (IPD) meta-
analysis: a review of the methodology" (https://www.ncbi.nlm.nih.gov/pmc/articles/PM
C5042043). Research Synthesis Methods. 6 (4): 293–309. doi:10.1002/jrsm.1160 (htt
ps://doi.org/10.1002%2Fjrsm.1160). PMC 5042043 (https://www.ncbi.nlm.nih.gov/pmc
/articles/PMC5042043). PMID 26287812 (https://pubmed.ncbi.nlm.nih.gov/26287812).
35. Debray TP, Moons KG, Abo-Zaid GM, Koffijberg H, Riley RD (2013). "Individual
35. Debray TP, Moons KG, Abo-Zaid GM, Koffijberg H, Riley RD (2013). "Individual
participant data meta-analysis for a binary outcome: one-stage or two-stage?" (https://
www.ncbi.nlm.nih.gov/pmc/articles/PMC3621872). PLOS ONE. 8 (4): e60650.
Bibcode:2013PLoSO...860650D (https://ui.adsabs.harvard.edu/abs/2013PLoSO...860
650D). doi:10.1371/journal.pone.0060650 (https://doi.org/10.1371%2Fjournal.pone.00
36. Burke DL, Ensor J, Riley RD (February 2017). "Meta-analysis using individual
participant data: one-stage and two-stage approaches, and why they may differ" (http
s://www.ncbi.nlm.nih.gov/pmc/articles/PMC5297998). Statistics in Medicine. 36 (5):
855–875. doi:10.1002/sim.7141 (https://doi.org/10.1002%2Fsim.7141). PMC 5297998
(https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5297998). PMID 27747915 (https://pu
bmed.ncbi.nlm.nih.gov/27747915).
37. Helfenstein U (March 2002). "Data and models determine treatment proposals--an
illustration from meta-analysis" (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC17423
01). Postgraduate Medical Journal. 78 (917): 131–134. doi:10.1136/pmj.78.917.131 (h
ttps://doi.org/10.1136%2Fpmj.78.917.131). PMC 1742301 (https://www.ncbi.nlm.nih.g
ov/pmc/articles/PMC1742301). PMID 11884693 (https://pubmed.ncbi.nlm.nih.gov/118
84693).
38. Senn S (March 2007). "Trying to be precise about vagueness". Statistics in Medicine.
26 (7): 1417–1430. doi:10.1002/sim.2639 (https://doi.org/10.1002%2Fsim.2639).
39. Al Khalaf MM, Thalib L, Doi SA (February 2011). "Combining heterogenous studies
using the random-effects model is a mistake and leads to inconclusive meta-
analyses". Journal of Clinical Epidemiology. 64 (2): 119–123.
doi:10.1016/j.jclinepi.2010.01.009 (https://doi.org/10.1016%2Fj.jclinepi.2010.01.009).
40. Brockwell SE, Gordon IR (March 2001). "A comparison of statistical methods for
meta-analysis". Statistics in Medicine. 20 (6): 825–840. doi:10.1002/sim.650 (https://d
oi.org/10.1002%2Fsim.650). PMID 11252006 (https://pubmed.ncbi.nlm.nih.gov/11252
41. Noma H (December 2011). "Confidence intervals for a random-effects meta-analysis
based on Bartlett-type corrections". Statistics in Medicine. 30 (28): 3304–3312.
doi:10.1002/sim.4350 (https://doi.org/10.1002%2Fsim.4350). hdl:2433/152046 (https:/
/hdl.handle.net/2433%2F152046). PMID 21964669 (https://pubmed.ncbi.nlm.nih.gov/
42. Brockwell SE, Gordon IR (November 2007). "A simple method for inference on an
overall effect in meta-analysis". Statistics in Medicine. 26 (25): 4531–4543.
doi:10.1002/sim.2883 (https://doi.org/10.1002%2Fsim.2883). PMID 17397112 (https://
pubmed.ncbi.nlm.nih.gov/17397112). S2CID 887098 (https://api.semanticscholar.org/
CorpusID:887098).
43. Sidik K, Jonkman JN (November 2002). "A simple confidence interval for meta-
analysis". Statistics in Medicine. 21 (21): 3153–3159. doi:10.1002/sim.1262 (https://do
i.org/10.1002%2Fsim.1262). PMID 12375296 (https://pubmed.ncbi.nlm.nih.gov/12375
44. Jackson D, Bowden J (January 2009). "A re-evaluation of the 'quantile approximation
method' for random effects meta-analysis" (https://www.ncbi.nlm.nih.gov/pmc/articles/
PMC2991773). Statistics in Medicine. 28 (2): 338–348. doi:10.1002/sim.3487 (https://
doi.org/10.1002%2Fsim.3487). PMC 2991773 (https://www.ncbi.nlm.nih.gov/pmc/artic
doi.org/10.1002%2Fsim.3487). PMC 2991773 (https://www.ncbi.nlm.nih.gov/pmc/artic
les/PMC2991773). PMID 19016302 (https://pubmed.ncbi.nlm.nih.gov/19016302).
45. Poole C, Greenland S (September 1999). "Random-effects meta-analyses are not
always conservative" (https://doi.org/10.1093%2Foxfordjournals.aje.a010035).
American Journal of Epidemiology. 150 (5): 469–475.
doi:10.1093/oxfordjournals.aje.a010035 (https://doi.org/10.1093%2Foxfordjournals.aj
e.a010035). PMID 10472946 (https://pubmed.ncbi.nlm.nih.gov/10472946).
46. Riley RD, Higgins JP, Deeks JJ (February 2011). "Interpretation of random effects
meta-analyses". BMJ. 342: d549. doi:10.1136/bmj.d549 (https://doi.org/10.1136%2Fb
mj.d549). PMID 21310794 (https://pubmed.ncbi.nlm.nih.gov/21310794).
47. Kriston L (March 2013). "Dealing with clinical heterogeneity in meta-analysis.
Assumptions, methods, interpretation" (https://www.ncbi.nlm.nih.gov/pmc/articles/PM
C6878481). International Journal of Methods in Psychiatric Research. 22 (1): 1–15.
doi:10.1002/mpr.1377 (https://doi.org/10.1002%2Fmpr.1377). PMC 6878481 (https://
www.ncbi.nlm.nih.gov/pmc/articles/PMC6878481). PMID 23494781 (https://pubmed.n
cbi.nlm.nih.gov/23494781).
48. Langan, Dean; Higgins, Julian P.T.; Jackson, Dan; Bowden, Jack; Veroniki, Areti
Angeliki; Kontopantelis, Evangelos; Viechtbauer, Wolfgang; Simmonds, Mark (2019).
"A comparison of heterogeneity variance estimators in simulated random‐effects
meta‐analyses" (https://onlinelibrary.wiley.com/doi/10.1002/jrsm.1316). Research
Synthesis Methods. 10 (1): 83–98. doi:10.1002/jrsm.1316 (https://doi.org/10.1002%2F
jrsm.1316). ISSN 1759-2879 (https://www.worldcat.org/issn/1759-2879).
49. "MetaXL User Guide" (http://www.epigear.com/index_files/MetaXL%20User%20Guide
.pdf) (PDF). Retrieved 18 September 2018.
50. Kontopantelis E, Reeves D (1 August 2010). "Metaan: Random-effects meta-analysis"
(https://www.researchgate.net/publication/227629391). Stata Journal. 10 (3): 395–
407. doi:10.1177/1536867X1001000307 (https://doi.org/10.1177%2F1536867X10010
00307) – via ResearchGate.
51. Field, Andy P.; Gillett, Raphael (2010). "How to do a meta-analysis" (http://doi.wiley.co
m/10.1348/000711010X502733). British Journal of Mathematical and Statistical
Psychology. 63 (3): 665–694. doi:10.1348/000711010X502733 (https://doi.org/10.134
8%2F000711010X502733). PMID 20497626 (https://pubmed.ncbi.nlm.nih.gov/204976
26).
52. Viechtbauer, Wolfgang (2010). "Conducting Meta-Analyses in R with the metafor
Package" (http://www.jstatsoft.org/v36/i03/). Journal of Statistical Software. 36 (3).
doi:10.18637/jss.v036.i03 (https://doi.org/10.18637%2Fjss.v036.i03). ISSN 1548-
7660 (https://www.worldcat.org/issn/1548-7660). S2CID 15798713 (https://api.semant
icscholar.org/CorpusID:15798713).
53. Kontopantelis E, Springate DA, Reeves D (2013). Friede T (ed.). "A re-analysis of the
Cochrane Library data: the dangers of unobserved heterogeneity in meta-analyses" (
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3724681). PLOS ONE. 8 (7): e69930.
Bibcode:2013PLoSO...869930K (https://ui.adsabs.harvard.edu/abs/2013PLoSO...869
930K). doi:10.1371/journal.pone.0069930 (https://doi.org/10.1371%2Fjournal.pone.00
54. Kontopantelis E, Reeves D (27 September 2013). "A short guide and a forest plot
54. Kontopantelis E, Reeves D (27 September 2013). "A short guide and a forest plot
command (ipdforest) for one-stage meta-analysis" (https://www.researchgate.net/publ
ication/257316967). Stata Journal. 13 (3): 574–587.
doi:10.1177/1536867X1301300308 (https://doi.org/10.1177%2F1536867X130130030
8) – via ResearchGate.
55. "MetaXL software page" (http://www.epigear.com/). Epigear.com. 3 June 2017.
Retrieved 18 September 2018.
56. Doi SA, Barendregt JJ, Khan S, Thalib L, Williams GM (November 2015). "Advances
in the meta-analysis of heterogeneous clinical trials I: The inverse variance
heterogeneity model". Contemporary Clinical Trials. 45 (Pt A): 130–138.
doi:10.1016/j.cct.2015.05.009 (https://doi.org/10.1016%2Fj.cct.2015.05.009).
hdl:1885/17083 (https://hdl.handle.net/1885%2F17083). PMID 26003435 (https://pub
med.ncbi.nlm.nih.gov/26003435). S2CID 10792959 (https://api.semanticscholar.org/C
orpusID:10792959).
57. Doi SA, Thalib L (January 2008). "A quality-effects model for meta-analysis".
Epidemiology. 19 (1): 94–100. doi:10.1097/EDE.0b013e31815c24e7 (https://doi.org/1
0.1097%2FEDE.0b013e31815c24e7). PMID 18090860 (https://pubmed.ncbi.nlm.nih.
gov/18090860). S2CID 29723291 (https://api.semanticscholar.org/CorpusID:2972329
1).
58. Doi SA, Barendregt JJ, Mozurkewich EL (March 2011). "Meta-analysis of
heterogeneous clinical trials: an empirical example". Contemporary Clinical Trials. 32
(2): 288–298. doi:10.1016/j.cct.2010.12.006 (https://doi.org/10.1016%2Fj.cct.2010.12.
59. Doi SA, Barendregt JJ, Khan S, Thalib L, Williams GM (July 2015). "Simulation
Comparison of the Quality Effects and Random Effects Methods of Meta-analysis".
Epidemiology. 26 (4): e42–e44. doi:10.1097/EDE.0000000000000289 (https://doi.org/
10.1097%2FEDE.0000000000000289). PMID 25872162 (https://pubmed.ncbi.nlm.nih
.gov/25872162).
60. Doi SA, Barendregt JJ, Khan S, Thalib L, Williams GM (November 2015). "Advances
in the meta-analysis of heterogeneous clinical trials II: The quality effects model".
Contemporary Clinical Trials. 45 (Pt A): 123–129. doi:10.1016/j.cct.2015.05.010 (https
://doi.org/10.1016%2Fj.cct.2015.05.010). PMID 26003432 (https://pubmed.ncbi.nlm.ni
h.gov/26003432).
61. Bucher HC, Guyatt GH, Griffith LE, Walter SD (June 1997). "The results of direct and
indirect treatment comparisons in meta-analysis of randomized controlled trials".
Journal of Clinical Epidemiology. 50 (6): 683–691. doi:10.1016/s0895-4356(97)00049-
8 (https://doi.org/10.1016%2Fs0895-4356%2897%2900049-8). PMID 9250266 (https:
//pubmed.ncbi.nlm.nih.gov/9250266).
62. van Valkenhoef G, Lu G, de Brock B, Hillege H, Ades AE, Welton NJ (December
2012). "Automating network meta-analysis". Research Synthesis Methods. 3 (4):
285–299. doi:10.1002/jrsm.1054 (https://doi.org/10.1002%2Fjrsm.1054).
63. Brooks SP, Gelman A (1998). "General methods for monitoring convergence of
iterative simulations" (http://www.stat.columbia.edu/~gelman/research/published/broo
ksgelman2.pdf) (PDF). Journal of Computational and Graphical Statistics. 7 (4): 434–
455. doi:10.1080/10618600.1998.10474787 (https://doi.org/10.1080%2F10618600.19
98.10474787).
64. Heck DW, Gronau QF, Wagenmakers EJ, Patil I (17 March 2021). "metaBMA:
64. Heck DW, Gronau QF, Wagenmakers EJ, Patil I (17 March 2021). "metaBMA:
Bayesian model averaging for random and fixed effects meta-analysis" (https://CRAN.
R-project.org/package=metaBMA). CRAN. Retrieved 9 May 2022.
65. Bartoš F, Maier M, Wagenmakers EJ, Goosen J, Denwood M, Plummer M (20 April
2022). "RoBMA: An R Package for Robust Bayesian Meta-Analyses" (https://CRAN.R
-project.org/package=RoBMA). Retrieved 9 May 2022.
66. Gronau QF, Heck DW, Berkhout SW, Haaf JM, Wagenmakers EJ (July 2021). "A
Primer on Bayesian Model-Averaged Meta-Analysis" (https://doi.org/10.1177%2F2515
2459211031256). Advances in Methods and Practices in Psychological Science. 4
(3). doi:10.1177/25152459211031256 (https://doi.org/10.1177%2F2515245921103125
6). ISSN 2515-2459 (https://www.worldcat.org/issn/2515-2459). S2CID 237699937 (ht
tps://api.semanticscholar.org/CorpusID:237699937).
67. Bartoš F, Maier M, Quintana D, Wagenmakers EJ (16 October 2020). "Adjusting for
Publication Bias in JASP & R - Selection Models, PET-PEESE, and Robust Bayesian
Meta-Analysis" (https://osf.io/75bqn). Advances in Methods and Practices in
Psychological Science. doi:10.31234/osf.io/75bqn (https://doi.org/10.31234%2Fosf.io
%2F75bqn). S2CID 236826939 (https://api.semanticscholar.org/CorpusID:236826939
).
68. Senn S, Gavini F, Magrez D, Scheen A (April 2013). "Issues in performing a network
meta-analysis". Statistical Methods in Medical Research. 22 (2): 169–189.
doi:10.1177/0962280211432220 (https://doi.org/10.1177%2F0962280211432220).
69. White IR (2011). "Multivariate random-effects meta-regression: updates to mvmeta" (h
ttps://doi.org/10.1177%2F1536867X1101100206). The Stata Journal. 11 (2): 255–270.
doi:10.1177/1536867X1101100206 (https://doi.org/10.1177%2F1536867X1101100206
).
70. Willis BH, Hyde CJ (May 2014). "Estimating a test's accuracy using tailored meta-
analysis-How setting-specific data may aid study selection" (https://research.birmingh
am.ac.uk/portal/en/publications/estimating-a-tests-accuracy-using-tailored-metaanaly
sis--how-settingspecific-data-may-aid-study-selection(ce8277f9-9e95-486d-b4f2-2f18
d3f95510).html). Journal of Clinical Epidemiology. 67 (5): 538–546.
71. Willis BH, Hyde CJ (August 2015). "What is the test's accuracy in my practice
population? Tailored meta-analysis provides a plausible estimate" (https://research.bir
mingham.ac.uk/portal/en/publications/what-is-the-tests-accuracy-in-my-practice-popul
ation-tailored-metaanalysis-provides-a-plausible-estimate(7c8e0824-4715-4fcc-bc5f-3
bc24f4ae2f1).html). Journal of Clinical Epidemiology. 68 (8): 847–854.
72. Zhang H, Deng L, Schiffman M, Qin J, Yu K (2020). "Generalized integration model
for improved statistical inference by leveraging external summary data". Biometrika.
107 (3): 689–703. doi:10.1093/biomet/asaa014 (https://doi.org/10.1093%2Fbiomet%2
Fasaa014).
73. Higgins JP, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD, et al. (October
2011). "The Cochrane Collaboration's tool for assessing risk of bias in randomised
trials" (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3196245). BMJ. 343: d5928.
doi:10.1136/bmj.d5928 (https://doi.org/10.1136%2Fbmj.d5928). PMC 3196245 (https:/
/www.ncbi.nlm.nih.gov/pmc/articles/PMC3196245). PMID 22008217 (https://pubmed.
/www.ncbi.nlm.nih.gov/pmc/articles/PMC3196245). PMID 22008217 (https://pubmed.
ncbi.nlm.nih.gov/22008217).
74. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al.
(October 2011). "QUADAS-2: a revised tool for the quality assessment of diagnostic
accuracy studies" (https://doi.org/10.7326%2F0003-4819-155-8-201110180-00009).
Annals of Internal Medicine. 155 (8): 529–536. doi:10.7326/0003-4819-155-8-
201110180-00009 (https://doi.org/10.7326%2F0003-4819-155-8-201110180-00009).
75. Royston P, Parmar MK, Sylvester R (March 2004). "Construction and validation of a
prognostic model across several studies, with an application in superficial bladder
cancer". Statistics in Medicine. 23 (6): 907–926. doi:10.1002/sim.1691 (https://doi.org/
10.1002%2Fsim.1691). PMID 15027080 (https://pubmed.ncbi.nlm.nih.gov/15027080).
76. Willis BH, Riley RD (September 2017). "Measuring the statistical validity of summary
meta-analysis and meta-regression results for use in clinical practice" (https://www.nc
bi.nlm.nih.gov/pmc/articles/PMC5575530). Statistics in Medicine. 36 (21): 3283–3301.
doi:10.1002/sim.7372 (https://doi.org/10.1002%2Fsim.7372). PMC 5575530 (https://w
ww.ncbi.nlm.nih.gov/pmc/articles/PMC5575530). PMID 28620945 (https://pubmed.nc
bi.nlm.nih.gov/28620945).
77. Riley RD, Ahmed I, Debray TP, Willis BH, Noordzij JP, Higgins JP, Deeks JJ (June
2015). "Summarising and validating test accuracy results across multiple studies for
use in clinical practice" (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4973708).
Statistics in Medicine. 34 (13): 2081–2103. doi:10.1002/sim.6471 (https://doi.org/10.1
002%2Fsim.6471). PMC 4973708 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC49
78. LeLorier J, Grégoire G, Benhaddad A, Lapierre J, Derderian F (August 1997).
"Discrepancies between meta-analyses and subsequent large randomized, controlled
trials" (https://doi.org/10.1056%2FNEJM199708213370806). The New England
Journal of Medicine. 337 (8): 536–542. doi:10.1056/NEJM199708213370806 (https://
doi.org/10.1056%2FNEJM199708213370806). PMID 9262498 (https://pubmed.ncbi.nl
m.nih.gov/9262498).
79. Slavin RE (1986). "Best-Evidence Synthesis: An Alternative to Meta-Analytic and
Traditional Reviews". Educational Researcher. 15 (9): 5–9.
doi:10.3102/0013189X015009005 (https://doi.org/10.3102%2F0013189X015009005).
80. Hunter JE, Schmidt FL, Jackson GB, et al. (American Psychological Association.
Division of Industrial-Organizational Psychology) (1982). Meta-analysis: cumulating
research findings across studies. Beverly Hills, California: Sage. ISBN 978-0-8039-
1864-1.
81. Glass GV, McGaw B, Smith ML (1981). Meta-analysis in social research. Beverly
Hills, California: Sage Publications. ISBN 978-0-8039-1633-3.
82. Polanin JR, Tanner-Smith EE, Hennessy EA (2016). "Estimating the Difference
Between Published and Unpublished Effect Sizes: A Meta-Review" (http://journals.sag
epub.com/doi/10.3102/0034654315582067). Review of Educational Research. 86 (1):
207–236. doi:10.3102/0034654315582067 (https://doi.org/10.3102%2F00346543155
82067). ISSN 0034-6543 (https://www.worldcat.org/issn/0034-6543).
83. Rosenthal R (1979). "The "File Drawer Problem" and the Tolerance for Null Results" (
https://semanticscholar.org/paper/dfa3c3689503600fe0bcc45eb0ad319e313079b9).
https://semanticscholar.org/paper/dfa3c3689503600fe0bcc45eb0ad319e313079b9).
Psychological Bulletin. 86 (3): 638–641. doi:10.1037/0033-2909.86.3.638 (https://doi.
org/10.1037%2F0033-2909.86.3.638). S2CID 36070395 (https://api.semanticscholar.
org/CorpusID:36070395).
84. Light RJ, Pillemer DB (1984). Summing up : the science of reviewing research (https:/
/archive.org/details/summingupscience00ligh). Cambridge, Massachusetts: Harvard
University Press. ISBN 978-0-674-85431-4.
85. Vevea JL, Woods CM (December 2005). "Publication bias in research synthesis:
sensitivity analysis using a priori weight functions". Psychological Methods. 10 (4):
428–443. doi:10.1037/1082-989X.10.4.428 (https://doi.org/10.1037%2F1082-989X.10
.4.428). PMID 16392998 (https://pubmed.ncbi.nlm.nih.gov/16392998).
86. Ioannidis JP, Trikalinos TA (April 2007). "The appropriateness of asymmetry tests for
publication bias in meta-analyses: a large survey" (https://www.ncbi.nlm.nih.gov/pmc/
articles/PMC1839799). CMAJ. 176 (8): 1091–1096. doi:10.1503/cmaj.060410 (https://
doi.org/10.1503%2Fcmaj.060410). PMC 1839799 (https://www.ncbi.nlm.nih.gov/pmc/
articles/PMC1839799). PMID 17420491 (https://pubmed.ncbi.nlm.nih.gov/17420491).
87. Hedges LV, Vevea JL (1996). "Estimating Effect Size Under Publication Bias: Small
Sample Properties and Robustness of a Random Effects Selection Model" (http://jour
nals.sagepub.com/doi/10.3102/10769986021004299). Journal of Educational and
Behavioral Statistics. 21 (4): 299–332. doi:10.3102/10769986021004299 (https://doi.o
rg/10.3102%2F10769986021004299). ISSN 1076-9986 (https://www.worldcat.org/iss
n/1076-9986). S2CID 123680599 (https://api.semanticscholar.org/CorpusID:1236805
99).
88. Ferguson CJ, Brannick MT (March 2012). "Publication bias in psychological science:
prevalence, methods for identifying and controlling, and implications for the use of
meta-analyses". Psychological Methods. 17 (1): 120–128. doi:10.1037/a0024445 (htt
ps://doi.org/10.1037%2Fa0024445). PMID 21787082 (https://pubmed.ncbi.nlm.nih.go
v/21787082).
89. Simmons JP, Nelson LD, Simonsohn U (November 2011). "False-positive psychology:
undisclosed flexibility in data collection and analysis allows presenting anything as
significant" (https://doi.org/10.1177%2F0956797611417632). Psychological Science.
22 (11): 1359–1366. doi:10.1177/0956797611417632 (https://doi.org/10.1177%2F095
90. LeBel E, Peters K (2011). "Fearing the future of empirical psychology: Bem's (2011)
evidence of psi as a case study of deficiencies in modal research practice" (https://we
b.archive.org/web/20121124154834/http://publish.uwo.ca/~elebel/documents/l%26p%
282011%2Crgp%29.pdf) (PDF). Review of General Psychology. 15 (4): 371–379.
doi:10.1037/a0025172 (https://doi.org/10.1037%2Fa0025172). S2CID 51686730 (http
s://api.semanticscholar.org/CorpusID:51686730). Archived from the original (http://pu
blish.uwo.ca/~elebel/documents/l&p(2011,rgp).pdf) (PDF) on 24 November 2012.
91. Radua J, Schmidt A, Borgwardt S, Heinz A, Schlagenhauf F, McGuire P, Fusar-Poli P
(December 2015). "Ventral Striatal Activation During Reward Processing in
Psychosis: A Neurofunctional Meta-Analysis" (https://doi.org/10.1001%2Fjamapsychi
atry.2015.2196). JAMA Psychiatry. 72 (12): 1243–1251.
doi:10.1001/jamapsychiatry.2015.2196 (https://doi.org/10.1001%2Fjamapsychiatry.20
15.2196). PMID 26558708 (https://pubmed.ncbi.nlm.nih.gov/26558708).
92. Hodges JS, Clayton MK (February 2011). "Random effects old and new": 1–23.
CiteSeerX 10.1.1.225.2685 (https://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.
1.225.2685).
1.225.2685).
93. Hodges JS (2014). "Random effects old and new.". Richly parameterized linear
models : additive, time series, and spatial models using random effects. Boca Raton:
CRC Press. pp. 285–302. ISBN 978-1-4398-6683-2.
94. Roseman M, Milette K, Bero LA, Coyne JC, Lexchin J, Turner EH, Thombs BD
(2011), "Reporting of Conflicts of Interest in Meta-analyses of Trials of
Pharmacological Treatments" (https://www.rug.nl/research/portal/en/publications/repo
rting-of-conflicts-of-interest-in-metaanalyses-of-trials-of-pharmacological-treatments(d
4a95ee2-429f-45a4-a917-d794ee954797).html), Journal of the American Medical
Association, 305 (10): 1008–1017, doi:10.1001/jama.2011.257 (https://doi.org/10.100
1%2Fjama.2011.257), hdl:11370/d4a95ee2-429f-45a4-a917-d794ee954797 (https://h
dl.handle.net/11370%2Fd4a95ee2-429f-45a4-a917-d794ee954797), PMID 21386079
(https://pubmed.ncbi.nlm.nih.gov/21386079), S2CID 11270323 (https://api.semanticsc
holar.org/CorpusID:11270323)
95. "The Osteen Decision" (http://archive.tobacco.org/Documents/980717osteen.html).
The United States District Court for the Middle District of North Carolina. 17 July
1998. Retrieved 18 March 2017.
96. Munafò MR, Flint J (September 2004). "Meta-analysis of genetic association studies".
Trends in Genetics. 20 (9): 439–444. doi:10.1016/j.tig.2004.06.014 (https://doi.org/10.
1016%2Fj.tig.2004.06.014). PMID 15313553 (https://pubmed.ncbi.nlm.nih.gov/15313
553).
97. Stone DL, Rosopa PJ (1 March 2017). "The Advantages and Limitations of Using
Meta-analysis in Human Resource Management Research". Human Resource
Management Review. 27 (1): 1–7. doi:10.1016/j.hrmr.2016.09.001 (https://doi.org/10.
1016%2Fj.hrmr.2016.09.001). ISSN 1053-4822 (https://www.worldcat.org/issn/1053-4
822).
98. Button KS, Ioannidis JP, Mokrysz C, Nosek BA, Flint J, Robinson ES, Munafò MR
(May 2013). "Power failure: why small sample size undermines the reliability of
neuroscience" (https://doi.org/10.1038%2Fnrn3475). Nature Reviews. Neuroscience.
14 (5): 365–376. doi:10.1038/nrn3475 (https://doi.org/10.1038%2Fnrn3475).
PMID 23571845 (https://pubmed.ncbi.nlm.nih.gov/23571845). S2CID 455476 (https://
api.semanticscholar.org/CorpusID:455476).
99. Milkman KL, Gromet D, Ho H, Kay JS, Lee TW, Pandiloski P, et al. (December 2021).
"Megastudies improve the impact of applied behavioural science" (https://www.ncbi.nl
m.nih.gov/pmc/articles/PMC8822539). Nature. 600 (7889): 478–483.
Bibcode:2021Natur.600..478M (https://ui.adsabs.harvard.edu/abs/2021Natur.600..478
M). doi:10.1038/s41586-021-04128-4 (https://doi.org/10.1038%2Fs41586-021-04128-
PMID 34880497 (https://pubmed.ncbi.nlm.nih.gov/34880497). S2CID 245047340 (htt
ps://api.semanticscholar.org/CorpusID:245047340).
100. Cheung AC, Slavin RE (1 June 2016). "How Methodological Features Affect Effect
Sizes in Education". Educational Researcher. 45 (5): 283–292.
doi:10.3102/0013189X16656615 (https://doi.org/10.3102%2F0013189X16656615).
ISSN 0013-189X (https://www.worldcat.org/issn/0013-189X). S2CID 148531062 (http
s://api.semanticscholar.org/CorpusID:148531062).
101. Jüni P, Witschi A, Bloch R, Egger M (September 1999). "The hazards of scoring the
quality of clinical trials for meta-analysis" (https://doi.org/10.1001%2Fjama.282.11.105
4). JAMA. 282 (11): 1054–1060. doi:10.1001/jama.282.11.1054 (https://doi.org/10.100
1%2Fjama.282.11.1054). PMID 10493204 (https://pubmed.ncbi.nlm.nih.gov/1049320
4).
4).
102. Armijo-Olivo S, Fuentes J, Ospina M, Saltaji H, Hartling L (September 2013).
"Inconsistency in the items included in tools used in general health research and
physical therapy to evaluate the methodological quality of randomized controlled
trials: a descriptive analysis" (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3848693
). BMC Medical Research Methodology. 13 (1): 116. doi:10.1186/1471-2288-13-116 (h
ttps://doi.org/10.1186%2F1471-2288-13-116). PMC 3848693 (https://www.ncbi.nlm.ni
h.gov/pmc/articles/PMC3848693). PMID 24044807 (https://pubmed.ncbi.nlm.nih.gov/
24044807).
103. Debray TP, Riley RD, Rovers MM, Reitsma JB, Moons KG (October 2015). "Individual
participant data (IPD) meta-analyses of diagnostic and prognostic modeling studies:
guidance on their use" (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4603958).
PLOS Medicine. 12 (10): e1001886. doi:10.1371/journal.pmed.1001886 (https://doi.or
g/10.1371%2Fjournal.pmed.1001886). PMC 4603958 (https://www.ncbi.nlm.nih.gov/p
mc/articles/PMC4603958). PMID 26461078 (https://pubmed.ncbi.nlm.nih.gov/264610
78).
104. Debray TP, Moons KG, Ahmed I, Koffijberg H, Riley RD (August 2013). "A framework
for developing, implementing, and evaluating clinical prediction models in an
individual participant data meta-analysis". Statistics in Medicine. 32 (18): 3158–3180.
pubmed.ncbi.nlm.nih.gov/23307585). S2CID 25308961 (https://api.semanticscholar.or
g/CorpusID:25308961).
105. Debray TP, Koffijberg H, Vergouwe Y, Moons KG, Steyerberg EW (October 2012).
"Aggregating published prediction models with individual participant data: a
comparison of different approaches". Statistics in Medicine. 31 (23): 2697–2712.
pubmed.ncbi.nlm.nih.gov/22733546). S2CID 39439611 (https://api.semanticscholar.or
106. Shadish, William R. (2014). "Analysis and meta-analysis of single-case designs: An
introduction" (https://linkinghub.elsevier.com/retrieve/pii/S0022440513001118).
Journal of School Psychology. 52 (2): 109–122. doi:10.1016/j.jsp.2013.11.009 (https://
doi.org/10.1016%2Fj.jsp.2013.11.009). PMID 24606971 (https://pubmed.ncbi.nlm.nih.
gov/24606971).
107. Zelinsky, Nicole A. M.; Shadish, William (19 May 2018). "A demonstration of how to
do a meta-analysis that combines single-case designs with between-groups
experiments: The effects of choice making on challenging behaviors performed by
people with disabilities" (https://www.tandfonline.com/doi/full/10.3109/17518423.2015.
1100690). Developmental Neurorehabilitation. 21 (4): 266–278.
doi:10.3109/17518423.2015.1100690 (https://doi.org/10.3109%2F17518423.2015.110
0690). ISSN 1751-8423 (https://www.worldcat.org/issn/1751-8423). PMID 26809945 (
https://pubmed.ncbi.nlm.nih.gov/26809945). S2CID 20442353 (https://api.semanticsc
holar.org/CorpusID:20442353).
108. Van den Noortgate W, Onghena P (2007). "Aggregating Single-Case Results" (https://
www.questia.com/read/1G1-170115042/the-aggregation-of-single-case-results-using-
hierarchical). The Behavior Analyst Today. 8 (2): 196–209. doi:10.1037/h0100613 (htt
ps://doi.org/10.1037%2Fh0100613).
109. Mantel N, Haenszel W (April 1959). "Statistical aspects of the analysis of data from
retrospective studies of disease". Journal of the National Cancer Institute. 22 (4):
719–748. doi:10.1093/jnci/22.4.719 (https://doi.org/10.1093%2Fjnci%2F22.4.719).
719–748. doi:10.1093/jnci/22.4.719 (https://doi.org/10.1093%2Fjnci%2F22.4.719).
110. Deeks JJ, Higgins JP, Altman DG, et al. (Cochrane Statistical Methods Group) (2021).
"Chapter 10: Analysing data and undertaking meta-analyses: 10.4.2 Peto odds ratio
method" (https://training.cochrane.org/handbook/current/chapter-10#section-10-4-2).
In Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M, Welch V (eds.).
Cochrane Handbook for Systematic Reviews of Interventions (Version 6.2 ed.). The
Cochrane Collaboration.
111. Bargaje R, Hariharan M, Scaria V, Pillai B (January 2010). "Consensus miRNA
expression profiles derived from interplatform normalization of microarray data" (https:
//www.ncbi.nlm.nih.gov/pmc/articles/PMC2802026). RNA. 16 (1): 16–25.
doi:10.1261/rna.1688110 (https://doi.org/10.1261%2Frna.1688110). PMC 2802026 (ht
tps://www.ncbi.nlm.nih.gov/pmc/articles/PMC2802026). PMID 19948767 (https://pubm
ed.ncbi.nlm.nih.gov/19948767).
112. Li, Xihao; Quick, Corbin; Zhou, Hufeng; Gaynor, Sheila M.; Liu, Yaowu; Chen, Han;
Selvaraj, Margaret Sunitha; Sun, Ryan; Dey, Rounak; Arnett, Donna K.; Bielak,
Lawrence F.; Bis, Joshua C.; Blangero, John; Boerwinkle, Eric; Bowden, Donald W.;
Brody, Jennifer A.; Cade, Brian E.; Correa, Adolfo; Cupples, L. Adrienne; Curran,
Joanne E.; de Vries, Paul S.; Duggirala, Ravindranath; Freedman, Barry I.; Göring,
Harald H. H.; Guo, Xiuqing; Haessler, Jeffrey; Kalyani, Rita R.; Kooperberg, Charles;
Kral, Brian G.; Lange, Leslie A.; Manichaikul, Ani; Martin, Lisa W.; McGarvey,
Stephen T.; Mitchell, Braxton D.; Montasser, May E.; Morrison, Alanna C.; Naseri,
Take; O’Connell, Jeffrey R.; Palmer, Nicholette D.; Peyser, Patricia A.; Psaty, Bruce
M.; Raffield, Laura M.; Redline, Susan; Reiner, Alexander P.; Reupena, Muagututi’a
Sefuiva; Rice, Kenneth M.; Rich, Stephen S.; Sitlani, Colleen M.; Smith, Jennifer A.;
Taylor, Kent D.; Vasan, Ramachandran S.; Willer, Cristen J.; Wilson, James G.;
Yanek, Lisa R.; Zhao, Wei; NHLBI Trans-Omics for Precision Medicine (TOPMed)
Consortium; TOPMed Lipids Working Group; Rotter, Jerome I.; Natarajan, Pradeep;
Peloso, Gina M.; Li, Zilin; Lin, Xihong (January 2023). "Powerful, scalable and
resource-efficient meta-analysis of rare variant associations in large whole genome
sequencing studies". Nature Genetics. 55 (1): 154–164. doi:10.1038/s41588-022-
01225-6 (https://doi.org/10.1038%2Fs41588-022-01225-6). PMID 36564505 (https://p
ubmed.ncbi.nlm.nih.gov/36564505). S2CID 255084231 (https://api.semanticscholar.or
Further reading
Cornell JE, Mulrow CD (1999). "Meta-analysis". In Mellenbergh GJ (ed.). Research
methodology in the life, behavioural, and social sciences. London: SAGE. pp. 285–
323. ISBN 978-0-7619-5883-3.
Ellis PD (2010). The Essential Guide to Effect Sizes: An Introduction to Statistical
Power, Meta-Analysis and the Interpretation of Research Results. Cambridge:
Cambridge University Press. ISBN 978-0-521-14246-5.
Sutton AJ, Jones DR, Abrams KR, Sheldon TA, Song F (2000). Methods for meta-
analysis in medical research. London: John Wiley. ISBN 978-0-471-49066-1.
Wilson DB, Lipsey MW (2001). Practical meta-analysis. Thousand Oaks: Sage
publications. ISBN 978-0-7619-2168-4.
Cooper H, Hedges LV, eds. (1994). The Handbook of Research Synthesis. New York:
Russell Sage Foundation. ISBN 978-0-87154-226-7.
Russell Sage Foundation. ISBN 978-0-87154-226-7.
Bonett DG (December 2010). "Varying coefficient meta-analytic methods for alpha
reliability". Psychological Methods. 15 (4): 368–385. doi:10.1037/a0020142 (https://do
i.org/10.1037%2Fa0020142). PMID 20853952 (https://pubmed.ncbi.nlm.nih.gov/2085
Bonett DG, Price RM (November 2014). "Meta-analysis methods for risk differences".
The British Journal of Mathematical and Statistical Psychology. 67 (3): 371–387.
doi:10.1111/bmsp.12024 (https://doi.org/10.1111%2Fbmsp.12024). PMID 23962020 (h
ttps://pubmed.ncbi.nlm.nih.gov/23962020).
Bonett DG (September 2008). "Meta-analytic interval estimation for bivariate
correlations". Psychological Methods. 13 (3): 173–181. doi:10.1037/a0012868 (https:/
/doi.org/10.1037%2Fa0012868). PMID 18778150 (https://pubmed.ncbi.nlm.nih.gov/18
Bonett DG (September 2009). "Meta-analytic interval estimation for standardized and
unstandardized mean differences". Psychological Methods. 14 (3): 225–238.
doi:10.1037/a0016619 (https://doi.org/10.1037%2Fa0016619). PMID 19719359 (https
://pubmed.ncbi.nlm.nih.gov/19719359).
Bonett DG, Price RM (September 2015). "Varying coefficient meta-analysis methods
for odds ratios and risk ratios". Psychological Methods. 20 (3): 394–406.
doi:10.1037/met0000032 (https://doi.org/10.1037%2Fmet0000032). PMID 25751513 (
https://pubmed.ncbi.nlm.nih.gov/25751513).
Bonett DG (November 2020). "Point-biserial correlation: Interval estimation,
hypothesis testing, meta-analysis, and sample size determination" (https://escholarshi
p.org/uc/item/3h82b18b). The British Journal of Mathematical and Statistical
Psychology. 73 Suppl 1 (Suppl 1): 113–144. doi:10.1111/bmsp.12189 (https://doi.org/1
0.1111%2Fbmsp.12189). PMID 31565811 (https://pubmed.ncbi.nlm.nih.gov/31565811
). S2CID 203607297 (https://api.semanticscholar.org/CorpusID:203607297).
Normand SL (February 1999). "Meta-analysis: formulating, evaluating, combining,
and reporting". Statistics in Medicine. 18 (3): 321–359. doi:10.1002/(SICI)1097-
0258(19990215)18:3<321::AID-SIM28>3.0.CO;2-P (https://doi.org/10.1002%2F%28S
ICI%291097-0258%2819990215%2918%3A3%3C321%3A%3AAID-SIM28%3E3.0.C
O%3B2-P). PMID 10070677 (https://pubmed.ncbi.nlm.nih.gov/10070677).
Owen AB (December 2009). "Karl Pearson's meta-analysis revisited" (https://web.arc
hive.org/web/20110726124334/http://statistics.stanford.edu/~ckirby/techreports/GEN/
2009/2009-06.pdf) (PDF). The Annals of Statistics. 37 (6B): 3867–2892.
arXiv:0911.3531 (https://arxiv.org/abs/0911.3531). doi:10.1214/09-AOS697 (https://doi
.org/10.1214%2F09-AOS697). S2CID 7632667 (https://api.semanticscholar.org/Corpu
sID:7632667). Archived from the original (http://www-stat.stanford.edu/~owen/reports/
AOS697.pdf) (PDF) on 26 July 2011.
Slough, Tara; Tyson, Scott A. (2022). "External Validity and Meta‐Analysis". American
Journal of Political Science. doi:10.1111/ajps.12742. ISSN 0092-5853.
Thompson SG, Pocock SJ (November 1991). "Can meta-analyses be trusted?" (https:
//web.archive.org/web/20111122083418/http://tobaccodocuments.org/pm/2047231315
-1318.pdf) (PDF). Lancet. 338 (8775): 1127–1130. doi:10.1016/0140-6736(91)91975-
Z (https://doi.org/10.1016%2F0140-6736%2891%2991975-Z). PMID 1682553 (https://
pubmed.ncbi.nlm.nih.gov/1682553). S2CID 29743240 (https://api.semanticscholar.org
/CorpusID:29743240). Archived from the original (http://tobaccodocuments.org/pm/20
47231315-1318.pdf) (PDF) on 22 November 2011. Retrieved 17 June 2011.. Explores
two contrasting views: does meta-analysis provide "objective, quantitative methods
two contrasting views: does meta-analysis provide "objective, quantitative methods
for combining evidence from separate but similar studies" or merely "statistical tricks
which make unjustified assumptions in producing oversimplified generalisations out of
a complex of disparate studies"?
O'Rourke K (2007). "Just the history from the combining of information: investigating
and synthesizing what is possibly common in clinical observations or studies via
likelihood" (https://web.archive.org/web/20111102182723/http://andrewgelman.com/m
ovabletype/mlm/JustHistory.pdf) (PDF). Oxford: University of Oxford, Department of
Statistics. Archived from the original (http://andrewgelman.com/movabletype/mlm/Just
History.pdf) (PDF) on 2 November 2011. Gives technical background material and
details on the "An historical perspective on meta-analysis" paper cited in the
references.
External links
Cochrane Handbook for Systematic Reviews of Interventions (http://handbook-5-1.co
chrane.org/)
Meta-Analysis at 25 (Gene V Glass) (http://www.gvglass.info/papers/meta25.html)
Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)
Statement (http://www.prisma-statement.org/) Archived (https://web.archive.org/web/2
0110727021816/http://www.prisma-statement.org/) 27 July 2011 at the Wayback
Machine – "an evidence-based minimum set of items for reporting in systematic
reviews and meta-analyses."
"metansue" R package and graphical interface (https://cran.r-project.org/web/package
s/metansue/index.html)
Best Evidence Encyclopedia (http://www.bestevidence.org)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Meta-analysis&oldid=1153092596"

Meta-Analysis - Wikipedia

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Meta-Analysis - Wikipedia

Uploaded by

Copyright:

Available Formats

A meta-analysis is a statistical analysis that

Not only can meta-analyses provide an estimate

However, there are some methodological problems with meta-analysis. If individual

Meta-analyses are often, but not always,

Differences (discrete data)

Methods and assumptions

Statistical models for aggregate data

Direct evidence: Models incorporating study effects only

Random effects model

1. Step 1: Inverse variance weighting

The extent of this reversal is solely dependent on two factors:[39]

Direct evidence: Models incorporating additional information

Quality effects model

Indirect evidence: Network meta-analysis methods

Specifying a Bayesian network meta-analysis model involves writing a directed acyclic

Frequentist multivariate framework

Generalized pairwise modelling framework

Aggregating IPD and AD

Validation of meta-analysis results

Publication bias: the file drawer problem

This file drawer problem (characterized by negative or

The distribution of effect sizes can be visualized with a

Most discussions of publication bias focus on journal practices favoring publication of

Problems related to studies not reporting non-statistically significant

Maximum likelihood estimation of the meta-analytic effect and the heterogeneity

Problems related to the statistical approach

Problems arising from agenda-driven bias

EPA's study selection is disturbing. First, there is evidence in the record

Comparability and validity of included studies

Heterogeneity of methods used may lead to faulty conclusions.[97] For instance,

Weak inclusion standards lead to misleading conclusions

Applications in modern science

Meta-analysis can be done with single-subject design as well as group research

Meta-analysis leads to a shift of emphasis from single studies to multiple studies. It

Seed-based d mapping (formerly signed differential mapping, SDM) is a statistical

Meta-analysis of whole genome sequencing studies provides an attractive solution to the

Retrieved from "https://en.wikipedia.org/w/index.php?title=Meta-analysis&oldid=1153092596"

You might also like