Should You Replicate a Famous Finding or a Finding That Actually Matters to Your Field?

Should You Replicate the Finding Everyone Knows?

You want to conduct a replication and several possibilities are available. One is a famous study: highly cited, widely discussed, perhaps even familiar outside its original discipline. Another is far less recognizable but addresses a claim that researchers in your field actually rely on.

Which should you choose?

Famous studies attract replication for understandable reasons. Their visibility makes the potential importance of a replication easy to explain, and a result that changes confidence in an influential claim may affect a large body of subsequent work. But fame can also pull scarce replication resources toward findings that are already heavily scrutinized while less visible but consequential claims remain largely untested.

The better question is not which paper has the most recognizable name. It is where another well-designed study could reduce uncertainty that is actually worth reducing.

Choose Replication Value, Not Visibility Alone

Replication Priority Is a Resource-Allocation Problem

You Cannot Replicate Everything That Might Benefit From Replication

There are far more empirical claims than researchers have time, funding, participants, laboratory capacity, or other resources to replicate. Choosing a replication target is therefore not simply a matter of identifying a study that could be repeated.

You are allocating scarce research resources among competing possibilities.

Isager and colleagues address this problem through the concept of replication value. Their decision model defines replication value in terms of the expected utility that could be gained from replicating a claim. Two broad considerations are central: the value of becoming more certain about the claim and the uncertainty surrounding the claim given the evidence currently available.

This does not provide a universal formula that automatically identifies the correct study to replicate. The authors explicitly acknowledge that value and uncertainty are multifaceted and that value depends partly on who is making the evaluation. The framework is useful because it forces the selection decision away from visibility alone and toward the expected informational value of the replication.

Fame Can Indicate Importance, but It Is Only an Indicator

A famous finding may genuinely be important. It may influence a major theory, appear in textbooks, motivate substantial subsequent research, shape professional practice, or affect public understanding. If confidence in the claim changed, the consequences could be considerable.

That can make it an excellent replication target.

But fame can also arise for reasons that do not map neatly onto scientific importance. Findings may attract attention because they are surprising, intuitively appealing, easy to communicate, controversial, associated with prominent researchers, or published in highly visible outlets.

Similarly, citation counts can indicate scholarly influence without revealing exactly why a paper is cited. A highly cited study may be cited for its method, theoretical discussion, dataset, historical significance, or even as an example of a disputed finding.

The two can overlap. They should not be treated as interchangeable.

Field Relevance Begins With What Depends on the Claim

A less famous finding may be deeply consequential within a particular research community.

Suppose researchers in your field routinely use a particular relationship to justify intervention design, select variables for models, make assumptions about learners, or recommend a professional practice. The originating study may have modest citation counts outside that specialty. Yet if the claim is unreliable, a meaningful portion of subsequent work could be affected.

That makes the claim a serious replication candidate.

When evaluating field relevance, ask:

• Is the claim used as a premise for subsequent research?
• Does it support an important theoretical proposition?
• Do practitioners make decisions partly because of it?
• Does it influence measurement, intervention, or methodological choices?
• Would researchers in the field revise their work if confidence in the claim changed?
• Could resolving the uncertainty prevent substantial effort from being built on unstable evidence?

A finding does not need to be famous outside the field for the answers to these questions to matter.

Importance Without Uncertainty May Not Create a High-Priority Replication

Imagine a highly important finding supported by numerous rigorous independent studies conducted across relevant populations and conditions. Researchers have reasonably precise estimates, and the evidence is broadly consistent.

The claim is important. But does it urgently need another nearly identical replication?

Perhaps not.

In Isager and colleagues' framework, replication value depends not only on the value of being certain but also on current uncertainty. When uncertainty has already been substantially reduced, the expected benefit of another similar replication may decline.

This is why the existence of previous evidence matters. A famous finding may attract another replication precisely because everyone knows it, even though it has already received extensive scrutiny. Meanwhile, a less visible claim with substantial consequences may depend on one small study and a handful of indirect tests.

Replication resources may be more informative in the second case.

This is closely connected to the question of whether another replication remains worthwhile after previous replication success.

Uncertainty Without Importance Is Also Not Enough

The reverse problem is equally important.

Suppose you discover a poorly supported finding based on a small sample, wide uncertainty, and no independent replication. Clearly, the claim is uncertain.

But almost nobody relies on it. It has little theoretical consequence, no meaningful practical application, and resolving the uncertainty would change very little.

Its uncertainty alone does not necessarily make it a high-priority replication target.

This illustrates why "this finding has never been replicated" is an incomplete rationale. The next question is: why should anyone care whether it replicates?

A strong replication target generally combines meaningful uncertainty with meaningful consequences of resolving that uncertainty.

The Claim Is Usually a Better Unit of Selection Than the Paper

Researchers often say, "I want to replicate this study." But a single article can contain several hypotheses, outcomes, experiments, subgroup analyses, and theoretical claims.

Which one are you actually replicating?

Nosek and Errington emphasize the relationship between replication and prior claims. This claim-centered view is useful for target selection because the importance and uncertainty of different claims within the same paper can vary substantially.

A famous article might contain one central finding that has been repeatedly confirmed and another consequential secondary claim that has received little independent testing. Replicating "the paper" is therefore too imprecise a description of the decision.

Define the target claim explicitly. Then evaluate its importance and uncertainty.

Ask What Would Change If the Replication Agreed With the Existing Evidence

Before choosing a target, imagine that your replication produces evidence broadly consistent with the existing claim.

What changes?

Would researchers become more confident in a relationship that currently rests on fragile evidence? Would a professional recommendation have a stronger empirical basis? Would an important theoretical assumption gain independent support?

If the answer is "not much, because several strong studies already establish this reasonably well," the incremental value of the replication may be limited.

Now apply the same question to the less famous candidate. A successful replication might provide the first serious independent support for a claim that a specialized field has been using for years.

That could be a much larger contribution.

Then Ask What Would Change If the Replication Disagreed

The opposite outcome can be even more revealing.

If your replication produced evidence inconsistent with the existing claim, what would researchers need to reconsider?

For a famous finding, perhaps the answer is substantial. A theoretical literature may require re-evaluation, or the result could reveal that an iconic phenomenon is more context-dependent than assumed.

For a field-specific finding, the consequences could be equally important. Researchers might reconsider an intervention, revise a model, stop treating a relationship as established, or investigate a previously ignored boundary condition.

This counterfactual exercise helps distinguish replication targets with genuine evidential stakes from those selected mainly because they would make an interesting headline.

Do Not Confuse Citation Count With Certainty

A heavily cited claim can still rest on surprisingly little direct evidence.

Citations accumulate for many reasons, and later papers may repeatedly cite the original study without independently testing its central finding. A claim can therefore become socially established within a literature faster than it becomes empirically well established.

The opposite can also occur. A claim may have modest citation visibility but be supported by several rigorous independent studies.

When selecting a replication target, separate two questions:

A citation metric cannot answer both questions for you.

Prestigious Publication Does Not Automatically Raise Replication Priority

A study published in a prestigious journal may be visible and consequential. It may also have already received substantial scrutiny precisely because of that visibility.

Conversely, a finding published in a specialized journal may be highly relevant to a particular field while attracting little attention elsewhere.

Journal prestige is therefore another possible signal, not a decision rule.

The same caution applies to awards, conference attention, social-media discussion, news coverage, and Altmetric-type indicators. Isager and colleagues discuss several such measures as possible indicators related to the value of a claim, but they do not equate any single metric with replication value.

A Finding Can Matter Because It Is Foundational Rather Than Famous

Some findings sit quietly underneath an entire line of research.

They may establish an assumption that later studies rarely question. Perhaps a particular instrument is assumed to predict an important outcome. Perhaps a relationship is routinely used to justify an intervention. Perhaps an effect is treated as the first link in a longer causal argument.

Such claims can have substantial downstream consequences even if the original paper is not a citation celebrity.

One useful literature-review strategy is therefore to look backward from current research. Ask which empirical claims contemporary studies repeatedly rely upon without directly retesting them.

That can reveal replication targets that are scientifically important precisely because everyone in the field has stopped noticing that they remain assumptions.

Practical Consequences Can Make a Specialized Finding a High-Value Target

Scientific importance is not the only kind of value.

A finding may influence educational practice, clinical decisions, organizational procedures, public programs, technology implementation, or other consequential activities. If people act differently depending on whether the claim is true, reducing uncertainty may have substantial practical value.

For example, suppose a modestly cited educational study supports a costly intervention that several institutions in your field are beginning to adopt. If independent evidence is sparse, another rigorous test could be more useful to those institutions than replicating a famous but practically irrelevant laboratory effect.

This does not mean applied research automatically outranks theoretical research. The point is that value depends on the consequences of uncertainty, and those consequences can be scientific, practical, or societal.

Your Own Field Expertise Matters, but It Should Be Made Explicit

Replication value is partly perspective-dependent.

A claim that matters enormously to one specialty may be peripheral to another. A funding agency, policymaker, professional association, researcher, or patient group may also assign different value to reducing uncertainty about the same finding.

This does not make replication target selection arbitrary. It means the evaluative perspective should be transparent.

If you argue that a finding matters to your field, show how. Identify the theory, research program, professional decision, or body of subsequent work that depends on it. Do not rely on the assertion that the topic is "important" as though importance were self-evident.

Feasibility Still Matters After You Identify the Ideal Target

A high-value target is not useful if your proposed study cannot test it adequately.

Perhaps the famous experiment requires equipment you cannot access. Perhaps the field-relevant study depends on proprietary materials, an inaccessible population, or an intervention you cannot implement faithfully. Perhaps the required sample is far beyond your available recruitment pool.

Replication selection therefore also involves asking whether your design can actually reduce the uncertainty that motivated the study.

An underpowered or poorly implemented replication of an extremely important claim may provide less information than a rigorous replication of a somewhat less important claim.

This is one reason the question of when replication is worth the research investment cannot be separated completely from feasibility.

Fame and Field Relevance Are Not Mutually Exclusive

There is no reason a replication target cannot be both famous and highly relevant to your field.

If a widely known finding is central to your specialty, remains meaningfully uncertain, and can be tested informatively with your resources, its fame is not a reason to avoid it. The problem arises only when fame substitutes for evaluating the claim.

Likewise, deliberately choosing an obscure finding simply because it is obscure is no better. Obscurity is not a virtue any more than fame is.

The goal is to identify where another study could produce the greatest useful reduction in uncertainty.

Choosing Between a Famous Finding and a Field-Relevant One

Now change one fact. Suppose Candidate A has become enormously influential but, despite its fame, has never received a sufficiently informative independent replication. Its prominence now increases the consequences of uncertainty rather than merely reflecting attention.

Candidate A could become the stronger target.

The lesson is not "always choose field relevance over fame." It is to determine what the fame or field relevance tells you about the value of resolving uncertainty and then examine how uncertain the claim actually remains.

Common Mistakes When Choosing a Replication Target

Compare Claims on Importance, Uncertainty, and What Your Study Can Add

If you have several possible replication targets, do not begin by ranking the papers by citations. Build a small evidence map for the claims themselves.

For each candidate, identify what is claimed, who or what depends on that claim, what independent evidence already exists, how uncertain the evidence remains, and whether your proposed study can materially improve that situation.

You can make this comparison explicit without pretending to calculate a perfectly objective replication score. A simple table of candidate claims, existing evidence, consequences, remaining uncertainty, and feasibility can expose weak rationales surprisingly quickly.

Once you select a target, justify the claim rather than the fame of the paper. Explain why confidence in the claim matters, what evidence currently exists, what remains uncertain, and how your study addresses that uncertainty.

This also gives you a more defensible answer when someone asks, "Why replicate this?" You do not need to argue that nobody has ever studied the topic. The appropriate task is to justify why another test of an existing claim is scientifically useful.

Before Choosing Which Finding to Replicate

Questions About Choosing Findings for Replication

Replicate the Claim Where Greater Certainty Would Matter

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.