November 9, 2026

The Claim Comes from a Review but You Need the Original Study

You are three weeks from a submission deadline, and one sentence in your draft is doing a lot of work. It says something like "prior work has shown that X improves Y under Z conditions," and the citation behind it is a r

Written byWisPaper TeamAI Research Workflow Team
Editorial illustration for The Claim Comes from a Review but You Need the Original Study

You are three weeks from a submission deadline, and one sentence in your draft is doing a lot of work. It says something like "prior work has shown that X improves Y under Z conditions," and the citation behind it is a review article. Your supervisor's comment in the margin is blunt: which study actually showed that? You open the review, scan the reference list, and find forty entries. The claim could come from any of them, or from none of them in the form you wrote it.

This is one of the most common citation problems in a course paper, thesis chapter, or first manuscript. A review is a useful map, but a map is not the territory. When you need to attribute a specific finding, method, or number, you need the original study that produced it. The work below is a practical repair sequence: diagnose what you actually have, decide which source to chase, retrieve and verify it, then rewrite the sentence so the citation matches the evidence.

Why a review citation is not enough for a specific claim

A review synthesizes many studies. Its authors compress, compare, and interpret. By the time a finding reaches a review sentence, the original conditions, sample, and measurement may be summarized or generalized. If you cite the review for a precise claim, you are attributing to the review something it reported secondhand.

That matters for two reasons. First, identity: you need to know which work produced the finding. Second, support: you need to check that the original text actually says what your sentence says. A DOI can be resolved through doi.org to the current destination registered for that identifier, but resolution only helps locate a work. The bibliographic fields and the source text still need to be checked for identity and claim support. In other words, finding the paper is half the job; confirming the claim is the other half.

Before you open a single database, classify the sentence you are trying to fix. Most problem claims fall into one of a few types, and each type needs a different repair.

Type 1: A specific empirical finding

Your sentence reports a result, effect, or measurement. This is the clearest case for chasing the original study. The review is a pointer; the original is the evidence.

Type 2: A method or procedure

Your sentence says a technique was used or validated. Here the original methods paper is usually the right target, because the review may describe the technique without the procedural detail you need.

Type 3: A broad consensus statement

Your sentence summarizes a field's general agreement. A review may legitimately be the right citation here, but you should still know whether you are citing a synthesis or a single study, and say so in the sentence.

Type 4: A number or statistic

Numbers are the highest-risk case. A figure quoted in a review may be rounded, converted, or drawn from a secondary analysis. Trace it to the source that generated it.

Write the claim on one line, note its type, and mark the exact words that carry the factual weight. That marked phrase is what you will verify against the original.

Decide which reference in the review to chase

A review's reference list is not a ranked shortlist. You have to narrow it deliberately.

Start with the review sentence itself. Which reference numbers sit closest to the claim? Review authors usually attach the most direct sources to the most specific statements. Then read the surrounding paragraph: does the review name an author, a year, a cohort, or a method? Those are anchors.

Next, use the review's own framing. If it says "several studies report X," you may need more than one original. If it says "one influential study," you are looking for a single target. If the review groups references by theme, work within the theme that matches your claim.

Finally, consider recency and centrality. A claim about a current method may trace to a recent primary paper; a foundational claim may trace to an older one. You are not looking for the newest reference — you are looking for the one that actually produced the finding.

Retrieve the original study without losing the thread

Once you have candidate references, retrieval becomes a matching problem. You have partial metadata from the review's reference list, and you need to confirm you have the right work.

A scholarly search interface is the natural place to start. WisPaper's academic paper search lets you look for papers through a scholarly search interface, which is useful when you have a title fragment or author name from the review. If you prefer a public entry point that explains the search offering, the Scholar Search page serves that purpose.

When you search, match on multiple fields rather than one. Crossref's REST API documentation describes publisher-deposited metadata that includes titles, authors, venues, dates, identifiers, licenses, funding, ORCID and ROR identifiers, abstracts when deposited, and post-publication updates from trusted sources. You do not need to query the API directly to benefit from that idea: check title, author, venue, and year together. A title match alone is weak evidence of identity.

Google Scholar's search help documents searching by author and title, date sorting and date limits, Cited by and Related articles, citation export, library access links, email alerts, and ways to locate alternative versions of a work. Those features are useful when the review's reference is incomplete or when you need to find a version you can actually read.

As you retrieve candidates, save them immediately. WisPaper's My Library is where you can save or upload papers and organize them, so the original study and the review stay together while you work.

Verify identity before you verify the claim

It is tempting to jump straight to the results section. Resist that until you have confirmed you are holding the right paper.

Run a short identity checklist:

  • Title and subtitle match the reference, including any correction.
  • Author list matches, and the first author is consistent.
  • Venue and year match, allowing for preprint-versus-published versions.
  • DOI or other identifier resolves to the same work.
  • Version is the one you intend to cite, not an earlier draft.

If a field disagrees, you may have found a different paper, a different version, or a metadata error. Resolve it before proceeding. A DOI that resolves correctly is reassuring, but the documentation is clear that resolution locates a work while bibliographic fields and the source text still need checking. Treat resolution as a step, not a verdict.

Check that the original actually supports your sentence

Now read the original with your marked claim in hand. You are looking for the passage that carries the finding, not the abstract alone.

Ask three questions. First, does the original report this finding at all, or does it report something adjacent that the review reframed? Second, do the conditions match — population, setting, method, time frame? Third, is the strength of the claim proportionate? If the original reports an association in one sample, your sentence should not say the effect is established across contexts.

This is where many repairs succeed or fail. A review may compress a cautious finding into a confident sentence. Your job is to restore the original's scope. If the original supports a narrower claim than you wrote, narrow the sentence rather than dropping the citation.

When you have several papers in your library and want to ask questions across them, WisPaper's Library QA supports asking questions based on papers uploaded to My Library or added from search results. That can help you compare what several originals say about the same claim before you rewrite.

Rewrite the sentence so the citation matches the evidence

With the original verified, rewrite the claim. Good rewrites usually do one of four things:

  1. Attribute precisely. Name the original study as the source of the finding.
  2. Narrow the scope. Add the conditions the original actually studied.
  3. Separate synthesis from evidence. Cite the review for the synthesis and the original for the finding.
  4. Split the claim. If two findings came from two papers, give each its own citation.

A useful pattern is to keep the review as context and add the original as support: the review frames the area, and the primary study carries the specific result. That way your reader can see both the map and the territory.

If your sentence cites a number, quote the original's figure and units exactly, and note any conversion you performed. If your sentence cites a method, cite the paper that introduced or validated it, not a later application.

Know when the repair is actually finished

A repaired citation is ready when a reader could follow your sentence to the original and find the support you promised. Use this readiness check:

  • The sentence's factual claim is traceable to a specific passage in the original.
  • The original's conditions and scope are reflected in your wording.
  • The citation points to the original, not only the review, for the specific claim.
  • Any review citation that remains is doing synthesis work, not evidence work.
  • The reference entry matches the version you read.

If any item fails, you are not done. The most common incomplete repair is a correct citation attached to an overbroad sentence.

Clean up the rest of the draft before the next project starts

The same problem rarely appears once. A winter-break cleanup or a January project start is a good moment to sweep the whole document for review-only citations attached to specific claims.

Work through your reference list and flag every review. For each flagged citation, read the sentence it supports and classify it using the four types above. Specific findings and numbers need originals; broad consensus statements may keep the review. This turns a single fix into a repeatable habit.

If you are checking a BibTeX file for references that may not correspond to real works, WisPaper's TrueCite supports BibTeX citation screening by checking entries against real academic databases to flag hallucinated references. That is a useful complement to claim-level verification: it addresses whether a reference exists, while your reading addresses whether it supports the sentence.

For a larger manuscript or a new project, keeping the review, the originals, and your notes in one place prevents the problem from recurring. WisPaper's Research Projects area is where you can create a new project and keep the work organized as it grows. If your next step involves a longer literature task, the Scholar Agent area is designed for longer research tasks with literature review, inspiration, or reproduction use cases.

A worked example: the thesis chapter in January

Suppose your thesis chapter argues that a particular intervention improves outcomes in a defined setting. Your draft cites a review. Your supervisor asks for the original.

You classify the sentence as a specific empirical finding. You return to the review paragraph and find two candidate references attached to the claim. You search for both by title and author, confirm venue and year, and save them to your library. One turns out to be a different study on a related population; the other is the original.

You read the original's results and find that the effect was reported for a narrower group than your sentence implies. You rewrite: the review frames the broader area, and the original is cited for the specific finding in its actual population. You check the reference entry against the version you read, then move to the next flagged review citation.

By the end of the sweep, your chapter distinguishes synthesis from evidence throughout. That is the corrected deliverable: not just one fixed sentence, but a draft where every specific claim points to the study that produced it.

Transferable lessons for your next manuscript

Three habits carry over to any project. First, classify claims before you cite them; the type of claim tells you what kind of source you need. Second, verify identity before support; a correct-looking reference to the wrong paper is worse than an obvious gap. Third, let the original set the scope of your sentence, not the review's summary.

These habits also make your reading more efficient. When you know you need an original, you read for the specific passage rather than the whole paper. When you know a review is appropriate, you can cite it with confidence and say so.

Your next concrete action

Pick the single sentence in your draft that carries the most factual weight and currently cites a review. Write the claim on one line, classify it, and identify the candidate references in the review. Then retrieve the most likely original, confirm its identity, and read the passage that carries the finding. Rewrite the sentence so the citation matches what the original actually shows. Do that once today, and you will have a template for every other review citation in the document.

FAQs

Match the citation to the job the sentence is doing. If the sentence reports a specific finding, method, or number, cite the original that produced it. If the sentence summarizes a field's general agreement, a review may be the appropriate source. Many strong sentences use both: the review for context and the original for the specific evidence.