How Much Effort Should You Spend Obtaining One Difficult-to-Access Paper?

When Does Persistence Stop Being Good Research Practice?

You have a citation for a paper that appears relevant. The DOI works, but the article is inaccessible. Your library does not have it. Repository searches lead nowhere. The author's old institutional page has disappeared. After several attempts, you are still looking.

Should you keep going?

There is no universal number of minutes, searches, or emails that makes a retrieval effort “enough.” The appropriate effort depends primarily on what the paper could contribute to your research and what would happen if you proceeded without it. A potentially decisive study in a systematic review deserves a different retrieval effort from a tangential citation you hoped might provide one more sentence for a background paragraph.

The goal is therefore not maximal persistence. It is proportionate retrieval: spending enough effort to reduce an important evidential uncertainty without allowing one source to consume resources far beyond its likely contribution.

Let the Paper's Evidential Importance Determine the Effort

Retrieval Effort Should Be Proportionate to Evidential Value

Not every inaccessible paper creates the same problem

A difficult source can matter for very different reasons. Perhaps it appears to be the only study of a particular population. Perhaps it contains the numerical result needed for a meta-analysis. Perhaps it is merely one of several papers supporting a well-established background statement.

Those situations should not receive identical treatment. The rational question is not simply “How difficult is this paper to obtain?” but “How consequential would it be if I never obtained it?”

Systematic reviews require a more structured retrieval process

Formal evidence synthesis changes the calculation. The Cochrane Handbook describes retrieving full-text reports of potentially relevant studies as part of the study-selection process and recommends corresponding with investigators where appropriate when information needed to determine eligibility or obtain missing information is unavailable. If information remains unobtainable, the review process should record that status rather than pretend that the report was assessed normally.

PRISMA 2020 likewise explicitly accommodates reports that were sought but could not be retrieved. Guidance accompanying PRISMA explains that this category can include reports unavailable through local collections or interlibrary loan, failures to obtain material from authors, and broken links.

This does not mean you must search indefinitely. It means that failure to retrieve a potentially relevant report is itself information about the review process and should be made visible.

Start with high-yield, low-cost routes

Before deciding that a paper is genuinely difficult to obtain, use the routes most likely to solve the problem quickly. Resolve the DOI, inspect the publisher page, search the exact title, check your institutional library, look for legitimate repository copies, and inspect author or institutional pages where appropriate.

If the article is subscription-restricted, the relevant options are those used when a research paper is behind a paywall. If you can locate only the bibliographic record, follow a structured process for dealing with a citation whose full text cannot initially be found.

Escalate when the paper matters enough

If ordinary searching fails and the source is important, escalate to routes that require more effort. Depending on your context, these may include interlibrary loan or document delivery, help from a research librarian, searching specialized repositories or catalogs, tracing another legitimate version, and contacting the corresponding author.

These routes are not interchangeable. A library request may be particularly appropriate for an older journal article held elsewhere, while author contact may be efficient for an accepted manuscript or an associated file. If direct author contact is appropriate, make the request specific and concise rather than sending a generic demand for access. The practical considerations involved in requesting a paper from its author apply here.

Ask what another retrieval attempt is likely to change

Suppose you have already checked the publisher, your library, several appropriate repositories, and the author's current institutional profile. You have also submitted an interlibrary-loan request. Running the same title through six additional general search engines may add very little.

This is a marginal-value problem. Each additional retrieval attempt has a cost in time and attention. Continue when another route has a plausible chance of succeeding and the paper matters enough to justify that cost. Stop when you are merely repeating low-yield searches without changing the evidential situation.

Replaceable sources deserve less heroic searching

Sometimes you want a paper because another author cited it for a broad proposition. If several accessible primary studies or authoritative sources establish the same point, obtaining that one paper may add little to your work.

That does not mean replacing sources casually. Two papers that sound similar may study different populations, measures, interventions, or periods. But if your informational need can genuinely be satisfied by stronger accessible evidence, there is little methodological virtue in spending hours pursuing a particular citation merely because you encountered it first.

Unique sources deserve more effort

The calculation changes when the inaccessible source appears irreplaceable. An old study may be the only evidence for a population. A paper may contain the only available estimate needed for a meta-analysis. A methodological article may be the original source of a procedure repeatedly described secondhand.

In such cases, another retrieval attempt can have substantial informational value. Library assistance, archival searching, author contact, or tracing related reports may be justified even when you would have abandoned a peripheral source much earlier.

Do not use citation count or journal prestige as your stopping rule

A highly cited paper is not automatically important to your particular research question, and an obscure paper may contain exactly the evidence you need. Likewise, journal prestige does not determine whether retrieving an individual source is worth your time.

Evaluate relevance, uniqueness, and consequence. Retrieval effort should follow your research need, not the social status of the citation.

Set stopping rules before frustration sets them for you

A simple stopping rule can make retrieval decisions more consistent. For example, you might decide that a potentially important source receives the standard retrieval routes first, followed by one or more escalation routes appropriate to your project. If those fail, you document the source as unobtainable unless new information emerges.

The exact rule should fit the research design. A doctoral thesis, rapid review, scoping review, historical investigation, and casual literature search operate under different resource constraints. What matters is that the rule reflects methodological purpose rather than your patience on a particular afternoon.

Keep a retrieval log when the stakes are high

For formal reviews and other reproducible research, record the routes you attempted and their outcomes. A brief log might include the DOI or citation, publisher access, library search, repository search, interlibrary-loan request, author contact, dates, and final retrieval status.

This serves two purposes. It prevents repeated searches that have already failed, and it allows another researcher to understand why the source was not used. More importantly, it separates “we did not bother to obtain this paper” from “we made reasonable attempts but could not retrieve it.” Those are methodologically different situations.

Two Inaccessible Papers Do Not Necessarily Deserve the Same Effort

The difference is not that Paper A is inherently “better.” Its absence creates greater uncertainty for the particular research task. Retrieval effort should respond to that uncertainty.

Common Mistakes When Deciding How Long to Keep Searching

Use a Proportionate Retrieval Strategy

For an informal literature search, your stopping rule may be quite modest. For a systematic review, evidence synthesis, replication, or historical analysis, the standard should generally be more explicit because missing evidence can alter the resulting evidence base.

Most importantly, do not turn “I could not obtain it” into “it does not matter.” Those are separate judgments. Sometimes the correct outcome is to stop searching while still acknowledging that the missing paper leaves uncertainty.

Before You Spend Another Hour Looking for One Paper

Questions About How Long to Keep Searching for a Paper

Spend Retrieval Effort Where Missing Evidence Matters Most

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.