How Do You Find Negative or Null Studies That Were Never Published?

How do you search for a study precisely because it may never have become a paper?

A database search can retrieve a published article with a positive result and a published article with a null result. The harder problem is the study that was conducted, produced an unremarkable, negative, or null finding, and never entered the searchable journal literature at all.

This matters because empirical evidence has repeatedly shown that publication can be associated with the direction or statistical significance of results. A systematic review of publication-bias studies found that trials with positive findings were more likely to be published than trials with negative or null findings.

You cannot solve that problem simply by adding “null results” to a PubMed or Scopus query. If the study was never published there, the database has nothing to retrieve. You need to search for evidence that the study existed independently of whether its results became a journal article.

Search for studies before you search for their findings

Finding missing null results requires a different search strategy

Negative, null, and unpublished are not synonyms

Researchers sometimes use these terms loosely, but they describe different things. A null result generally indicates that the study did not detect the hypothesized difference or association under the analysis performed. A negative result may refer to a result that does not support the hypothesis, favours a comparator, or is otherwise regarded as unfavourable. Definitions vary by context.

Unpublished describes dissemination status. An unpublished study could have positive, negative, null, mixed, or inconclusive results.

Keeping those concepts separate protects you from an important inferential error: finding a missing study does not tell you why it is missing.

Why might null or negative findings be harder to find?

Publication bias occurs when the likelihood of publication is influenced by the direction or strength of results. WHO uses this definition in its clinical-trial terminology and identifies publication bias and selective reporting as reasons trial registration is important.

Empirical research has found that positive clinical-trial findings have historically been published more frequently and more quickly than negative or null findings. One systematic review estimated that trials with positive findings were substantially more likely to reach publication than those with negative or null findings.

The result can be a published evidence base that differs systematically from the underlying research that was conducted. This is the practical problem behind the file drawer problem.

Do not search only for the words “negative” and “null”

Searching bibliographic or web databases for “negative results” can retrieve some useful material, but it cannot recover studies whose results were never publicly described using those terms.

Instead, search for studies based on characteristics that existed before their results were known: intervention, population, research question, project title, investigator, sponsor, institution, grant, protocol, trial registration, conference presentation, or study identifier.

This is the core methodological shift. You are trying to reconstruct the population of studies, then determine which results are available.

Search study registries when the research type is registered

Clinical trial registries are particularly valuable because registration can create a public record before results exist. WHO notes that a registry record may be the only publicly available document about a trial until its results are published and explicitly recognizes the relevance of registry records for ongoing or unpublished trials.

Look for completed studies without identifiable journal reports, then investigate whether summary results have been posted. WHO's current guidance emphasizes public reporting of both positive and negative trial results and notes that incomplete reporting can leave a misleading evidence base.

For health-intervention reviews, this makes clinical trial registries an important route to completed but unpublished research.

Search grant databases for projects that disappeared from the publication trail

Funding records can establish that research was planned and funded independently of whether it later generated a paper. Search project titles, principal investigators, institutions, dates, abstracts, and grant numbers, then trace the project forward.

A completed grant with no obvious publication is not proof of a null result. The study may have changed, failed to proceed, produced outputs you have not found, or still be awaiting publication. Nevertheless, grant databases can reveal research projects worth investigating.

Conference material can reveal studies that subsequently vanished

Conference abstracts, posters, programmes, and proceedings may record research before full publication. If an apparently eligible conference study never becomes a journal article, trace the authors, study title, institution, registration number, and project identifiers.

A conference abstract may provide too little information for reliable extraction or appraisal, but it can establish a useful lead. When the full report cannot be found, the problem becomes what to do with a conference study that never appears as a full paper.

Search theses, dissertations, reports, and institutional repositories

Not every null finding is completely undisclosed. Some results may exist outside journals in dissertations, technical reports, institutional repositories, working papers, funder reports, government publications, or organizational research.

This is why searching theses, dissertations, reports, and working papers can recover evidence that conventional publication searches miss.

Use authors, supervisors, institutions, project titles, distinctive study terms, and dates to connect grey-literature records with known research programmes.

Regulatory documents can expose missing clinical studies and outcomes

For medicines and other regulated interventions, regulatory submissions and assessment documents may contain information about studies that are absent from journal databases or provide results not fully reported in publications.

Searching regulatory documents for evidence missing from journal articles can therefore be important when the intervention has undergone a regulatory process.

Follow researchers, not just papers

If you identify an investigator associated with a potentially missing study, search the investigator's publication history, institutional profile, ORCID record where available, project pages, co-authors, and subsequent papers.

A study title may change substantially between protocol, conference presentation, dissertation, report, and journal publication. Investigator names, sample characteristics, study dates, interventions, and institutional affiliations can help establish whether apparently different records describe the same study.

Contact researchers when the missing study could matter

Direct contact can sometimes clarify whether a study was completed, whether results exist, whether a manuscript was published under another title, and whether data can be shared.

Researcher contact is not guaranteed to succeed, and selective responses can introduce their own problems. It should therefore be systematic where possible rather than reserved for studies you happen to find interesting. Whether direct contact is justified depends on the importance of the missing evidence and the resources available.

Search for missing studies symmetrically

There is a subtle methodological trap here. If you search aggressively for unpublished studies only after seeing a surprisingly positive meta-analysis, your additional retrieval process can become outcome-driven.

A better strategy is to specify in advance which unpublished sources you will search and apply those procedures consistently. Research on publication bias has cautioned that searches for unpublished studies can themselves introduce selection problems when retrieval is uneven or dependent on researcher networks.

Finding the study does not mean you can use its result

A registry entry, grant record, or conference title may establish that research existed without providing enough information for inclusion in a meta-analysis.

Separate three questions: Does the study exist? Does it meet your eligibility criteria? Are usable results available?

A study can be important to your assessment of missing evidence even when the answer to the third question is no. WHO guidance recognizes that unpublished and selectively reported results can leave decision-makers with an incomplete view of intervention benefits and harms.

How to trace a study that disappeared after registration

Common mistakes when looking for unpublished negative or null results

Build the search around study existence, not expected results

If negative or null research may be missing, your search strategy should identify studies independently of what they found. That means moving upstream from publications toward registrations, grants, protocols, conferences, organizations, and researchers.

Searching can expand indefinitely, so comprehensiveness has practical limits. Decide prospectively which sources matter most and how far you will pursue unresolved studies. The appropriate stopping point depends on whether searching for unpublished research is worth the effort for your particular synthesis.

Before concluding that negative or null studies are missing

Questions about finding unpublished negative and null research

Look for the missing study before guessing what it found

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.