Did You Overrepresent Evidence From One Country, Institution, or Research Group?

How Diverse Is the Evidence Behind Your Conclusion?

Your review contains 25 studies, which sounds reassuring. Then you examine them more closely. Fifteen come from one country. Nine involve the same research group. Several use the same national dataset. Most participants come from similar institutions.

You still have 25 studies, but perhaps not 25 independent tests of whether the finding travels across settings.

Concentration is not automatically a flaw. One country may legitimately dominate research on a locally specific policy, one institution may maintain a uniquely valuable cohort, and one research group may simply be especially productive. The problem arises when a concentrated evidence base is interpreted as though it represents populations, institutions, or contexts it has barely studied.

Evidence Concentration Matters When Context Could Affect the Answer

A Large Literature Can Still Come From a Narrow Evidence Base

Study count does not tell you how broadly the evidence has been tested

Twenty studies conducted across twenty substantially different settings provide a different kind of information from twenty studies repeatedly conducted in one setting. That does not mean the first set is automatically better. It means the two evidence bases support different judgments about contextual replication and applicability.

When studies cluster around the same institutions, researchers, populations, or data systems, their apparent numerical abundance can obscure a narrower evidence base underneath.

Country can matter, but nationality is not itself the mechanism

Finding that most studies come from one country should prompt questions, not an automatic penalty. The relevant issue is whether features that vary across settings could affect the phenomenon or intervention.

Cochrane's guidance on applicability identifies contextual factors such as cultural and linguistic diversity, socioeconomic position, rural or urban setting, and characteristics of service delivery as potentially relevant to whether evidence transfers from one setting to another. It advises caution when generalizing across contexts.

The implication is not that evidence from one country is intrinsically weak. Rather, broad international claims require justification when the underlying evidence comes from a comparatively narrow set of contexts.

A study can be methodologically rigorous yet provide indirect evidence for a substantially different target setting. Cochrane and GRADE address this issue through the concept of indirectness, which asks how closely the available evidence aligns with the population, intervention, comparator, and outcomes in the question of interest.

Institutional concentration can hide contextual dependencies

Research repeatedly conducted at one university, hospital system, laboratory, school network, or company may share characteristics that are not obvious from study titles. Recruitment practices, infrastructure, staff expertise, socioeconomic context, implementation support, institutional culture, or access to technology may influence the observed results.

This matters particularly for complex interventions. A program that succeeds in a highly resourced research-intensive institution may not perform identically when implemented with different personnel, infrastructure, incentives, or participant populations.

The appropriate response is not to dismiss single-institution research. It is to avoid silently converting evidence from a specific environment into a universal claim.

Repeated use of one dataset can create the appearance of replication

Several papers can analyze the same national survey, cohort, registry, administrative database, or institutional dataset. They may ask different questions and make legitimate contributions, but they do not necessarily provide independent replication across populations.

If two analyses use overlapping participants, there may also be a statistical dependency that requires attention. Before interpreting several papers as several independent confirmations, determine whether you need to account for the same participants appearing more than once.

Research-group concentration creates another kind of dependence

Studies from the same research team may be statistically independent while still sharing conceptual and methodological features. Researchers often reuse instruments, analytic conventions, recruitment networks, implementation procedures, theoretical assumptions, or operational definitions across projects.

Replication by the originating team and replication by independent teams therefore answer somewhat different questions. Consistent findings from one productive group may show that an effect is reproducible within that research program, but independent replication can provide additional information about whether the finding survives different investigators, procedures, and settings.

This should not be turned into suspicion based on authorship alone. Repeated authorship is a signal to inspect the structure of the evidence, not evidence that the research is biased.

Search systems can contribute to geographic concentration

Your final evidence base partly reflects where you searched. Database coverage differs across disciplines, journals, languages, and regions. Restricting a search to familiar international databases can make some research ecosystems easier to discover than others.

If geographic breadth matters to the question, consider whether your choice of databases provides adequate coverage of relevant regional literature. Vocabulary can matter too, particularly when terminology differs across disciplines or linguistic contexts.

Language restrictions can narrow the evidence base

Language eligibility and database coverage are separate but related issues. A search may locate studies in several languages only for the review to exclude most of them because of a language restriction. Conversely, a supposedly language-inclusive review cannot include literature its search sources never expose.

Restrictions may sometimes be necessary because of resources or review scope. They should nevertheless be reported transparently, particularly when they could affect the geographic or cultural composition of the evidence.

Map concentration before deciding whether it is a problem

A simple evidence map can make concentration visible. Record where studies were conducted, which institutions and research groups recur, which datasets are reused, and how participants are distributed across relevant contexts.

Do not manufacture diversity by giving weak evidence extra weight

Suppose most high-quality studies come from one region while a few methodologically weak studies come from elsewhere. Geographic diversity does not require treating those weaker studies as equally credible merely to create balance.

Contextual breadth and methodological strength are separate dimensions. You should examine both. A geographically narrow but rigorous evidence base may warrant cautious generalization, while a geographically broad collection of severely biased studies does not become strong evidence simply because the map looks impressive.

The solution is to give stronger and weaker evidence appropriate weight while separately discussing limitations in applicability.

Sometimes concentration accurately describes the research field

A concentrated evidence base may persist even after a broad, multilingual, multi-database search. If so, the concentration itself is a finding about the state of the literature.

Do not imply that evidence exists elsewhere merely because you wish the literature were more diverse. Instead, distinguish a biased search process from a genuine evidence gap. If most available studies really come from a small number of contexts, say so and calibrate the conclusion accordingly.

When 18 Studies Represent Fewer Contexts Than the Number Suggests

The number of included studies has not changed. What changed is the claim the evidence can reasonably support.

Common Mistakes When Judging the Diversity of an Evidence Base

Match the Breadth of Your Claim to the Breadth of Your Evidence

Once you have assembled the literature, inspect where the evidence actually comes from. Do not wait until the limitations section to discover that a supposedly broad conclusion rests on a handful of recurring contexts.

The aim is not to maximize the number of countries represented. It is to understand where the evidence comes from and what that provenance means for the inference you want to make.

Before You Generalize Beyond the Settings You Reviewed

Questions About Geographic and Institutional Concentration

Twenty Studies Can Still Represent a Narrow Slice of the World

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.