What Should You Do When the Abstract and Full Paper Seem to Describe Different Studies?

What if the abstract says one thing and the paper says another?

You retrieve a full paper because its abstract appears eligible. Then something feels wrong. The abstract says 240 participants, while the methods section reports 198. The abstract describes a randomized trial, but the full text makes the allocation procedure look nonrandom. An outcome presented as primary in the abstract appears secondary in the paper. In more extreme cases, the population, intervention, dates, or study design seem different enough that you begin wondering whether the abstract and paper are even describing the same study.

This is more than an irritating reporting problem. If the discrepancy concerns one of your eligibility criteria, choosing the wrong version can change whether the study enters your review.

The appropriate response is not to assume that the abstract is correct or that every discrepancy proves misconduct. First establish what is inconsistent and whether you are actually looking at the corresponding full report. Then make the eligibility decision using the most complete, verified information available.

Investigate the discrepancy before making the screening decision

Not every abstract–full-text discrepancy means the same thing

First confirm that you have the corresponding full paper

Before diagnosing a reporting inconsistency, rule out the mundane explanation: you may be comparing records that are related but not identical.

Check the article title, authors, journal, year, volume and pages or article number, DOI, trial registration number when available, intervention, recruitment dates, setting, and other identifying information. Database records can link to corrections, supplements, protocols, conference abstracts, secondary analyses, or other papers from the same research project.

This distinction matters because a research study and a research report are not necessarily one-to-one. One study can generate several publications. A protocol, baseline paper, primary outcome report, secondary analysis, follow-up study, and conference abstract may all refer to the same underlying project while legitimately reporting different samples or outcomes.

If the apparent mismatch reflects several reports arising from one underlying investigation, the problem is better handled as multiple papers from the same study or dataset rather than as an abstract contradicting its own article.

Then identify exactly what is different

"The abstract and paper do not match" is too vague to support a screening decision. Locate the discrepancy precisely.

The objective is diagnosis, not accusation. A difference in participant numbers, for example, may be entirely legitimate if the abstract reports randomized participants while one analysis in the full paper uses only participants with complete follow-up.

Abstracts are supposed to reflect the full report, but they do not always do so

Current CONSORT guidance states that the abstract of a randomized trial should accurately reflect the full journal article and should not contain information absent from the body of the paper. It also acknowledges empirical evidence showing inconsistencies and omissions between abstracts and full reports.

This is not merely theoretical. A scoping review of methodological research concluded that abstracts in primary biomedical research are commonly inconsistent with their corresponding full reports. Another systematic review synthesized evidence concerning thousands of randomized trials and systematic reviews and found problems in both abstract reporting and consistency with full texts.

More recent empirical work continues to find the problem in particular fields. A 2025 methodological survey of surgical pilot and feasibility randomized trials reported inconsistencies involving conclusions, omitted negative findings, outcomes, and claims appearing in abstracts but not in the main text. These estimates apply to that specific sample and should not be treated as prevalence estimates for research generally.

For full-text eligibility, use the fuller evidence available

Title and abstract screening is necessarily preliminary. Once you retrieve a report, you have more information with which to evaluate the eligibility criteria.

JBI describes study selection as occurring at title/abstract and full-text stages, with eligibility determined using prespecified criteria. PRISMA 2020 similarly asks review authors to report the methods used to determine whether studies met inclusion criteria and distinguishes records screened from reports retrieved and assessed.

Consequently, an abstract-level decision that a record is potentially eligible does not bind you once the full report provides contradictory information. The purpose of retrieving the full text is precisely to resolve questions that abbreviated records cannot settle.

If the abstract says the participants were undergraduates but the methods section clearly establishes that the sample consisted of secondary-school students, the full report changes the eligibility judgment. You should not include the study merely because it passed abstract screening.

But do not assume the full text automatically resolves every contradiction

The full article is usually richer in methodological detail, but "more text" does not guarantee that every inconsistency disappears. The abstract may give one number, the participant-flow diagram another, a table a third, and the methods section a fourth. Occasionally the report itself contains internal contradictions.

In that situation, trace the information to its methodological context. Ask what each number represents, at what stage of the study it was measured, which participants were included, which analysis was performed, and whether a correction or supplementary document exists.

For registered studies, a trial registry or protocol may help establish planned methods or outcomes. These external records should be used carefully: a protocol can clarify what was planned, while the published report tells you what was actually reported as having occurred. The two answer related but different questions.

Do not resolve contradictions by choosing the version that makes the paper eligible

This is one of the most consequential traps. Suppose the abstract satisfies your population criterion but the full text does not. It can be tempting to retain the study because you already expected it to qualify.

If the full text provides clear and specific information showing that an eligibility criterion is not met, exclude the study for that criterion. If the conflict remains genuinely unresolved, flag it for adjudication or clarification rather than pretending the evidence is unambiguous.

A reporting discrepancy is not automatically evidence of misconduct

Abstract–full-text inconsistencies can be serious, but their existence alone does not establish fabrication, falsification, or intentional deception. They may result from word limits, editing, reporting errors, differing denominators, incomplete updating of the abstract after manuscript revision, or selective emphasis.

Some discrepancies may constitute forms of misleading reporting or "spin," particularly when presentation makes findings appear more favorable or certain than the underlying results support. Research on spin distinguishes misleading reporting, misleading interpretation, and inappropriate extrapolation.

For literature screening, however, your immediate task is narrower: determine what study was conducted and whether it meets the review's criteria. Questions about research integrity require stronger evidence than the observation that two sections do not match.

Distinguish a discrepancy from duplicate or overlapping publication

If two apparently different papers report substantially overlapping participants, interventions, recruitment periods, or outcomes, you may be dealing with duplicate or redundant reporting rather than a poorly matched abstract.

That distinction matters because treating multiple reports of the same underlying data as independent studies can distort a synthesis. The appropriate response is to investigate the relationship among the reports and address potential duplicate publication rather than simply choosing whichever paper is easiest to interpret.

Resolving a population discrepancy between an abstract and full report

Common mistakes when abstracts and full papers conflict

Turn the apparent contradiction into a verification sequence

Do not treat the paper as a puzzle to solve intuitively. Use a reproducible sequence that another reviewer could follow.

When the discrepancy leads to exclusion, record the actual reason consistently. This preserves the audit trail emphasized in tracking why studies were excluded.

For systematic reviews, independent screening can also reduce the risk that one reviewer resolves an ambiguous report idiosyncratically. JBI recommends independent study selection by two or more reviewers with disagreements resolved through consensus or a further reviewer.

When the abstract and full text do not match, check this

Questions about discrepancies between abstracts and full papers

Verify the study before deciding which description to believe

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.