# Research Misconduct vs. Questionable Research Practices Research misconduct means fabrication, falsification, or plagiarism committed deliberately, while questionable research practices (QRPs) are lower-level behaviors that fall short of misconduct, such as selective analysis, sloppy methods, or careless reporting. The rule of thumb is simple: if the behavior involves inventing, altering, or stealing data or ideas on purpose, it belongs in the misconduct category. If it involves cutting corners, selective reporting, or bad judgment without outright fabrication, it belongs in the QRP category. Getting this distinction right matters because the two categories carry different consequences, different investigation procedures, and different responsibilities for everyone working in research. This guide defines each term, compares them side by side, walks through realistic scenarios, and explains how the responsible conduct of research (a framework of principles and standards for honest, accountable science) treats both categories. ## What Is Research Misconduct? Research misconduct is the most serious category of integrity violation. In the United States, the federal definition used by agencies that fund health research covers three behaviors, often abbreviated as FFP. **Fabrication** means making up data or results and recording or publishing them. A researcher who fills in missing patient measurements with invented numbers has fabricated data. **Falsification** means manipulating research materials, equipment, or processes, or changing or omitting data so that the record does not accurately represent the research. Deleting inconvenient data points to make a result look stronger is falsification. **Plagiarism** means appropriating another person's ideas, processes, results, or words without giving appropriate credit, including through unattributed reuse of your own previously published text in some institutional policies. The formal definition also includes two limiting conditions. The behavior must be committed intentionally, knowingly, or recklessly, and it must represent a significant departure from accepted practices in the relevant field. Honest errors and differences of opinion do not count as misconduct. Authorship problems occupy an adjacent space. Gift authors, ghost authors, and editors who manipulate authorship for career gain are treated as professional misconduct in the editorial literature, because authorship functions as academic currency that shapes salaries, grants, and standing [1]. A survey at a public university in Ethiopia found authorship misconduct was the most commonly reported form of misconduct at 47.5 percent of respondents, ahead of fabrication and falsification at 40.6 percent [2]. ## What Are Questionable Research Practices? Questionable research practices are behaviors that deviate from best practice without meeting the threshold of fabrication, falsification, or plagiarism. They damage the credibility and efficiency of science, but they usually sit in a gray area where intent is unclear, rules are ambiguous, or the practice has become normalized within a field. Common examples include: - **P-hacking**, which means analyzing data in many ways until a statistically significant result appears, then reporting only that analysis - **Selective reporting**, meaning publishing only the outcomes or analyses that worked - **Selective citing**, meaning citing papers to favor reviewers, friends, or your own work rather than the most relevant literature - **HARKing**, meaning hypothesizing after results are known and presenting them as if predicted in advance - **Careless supervision or training** that lets bad habits spread through a lab - **Weak authorship practices**, such as adding a lab head to every paper regardless of contribution A major classification study organized QRP definitions from dozens of surveys and found a consistent pattern. Researchers report QRPs most often when they affect interpretation, analysis, or citations, and when the practice involves omitting information rather than adding or changing it. The most commonly reported types were selective analysis, selective citation, and selective inclusion of covariates. The least commonly reported were denying authorship to deserving contributors and the outright FFP behaviors, which are misconduct rather than QRPs [4]. QRPs are diverse enough that researchers struggle to define their boundaries. A recent scoping review of 229 empirical studies published between 2016 and 2025 found that severe violations dominate how researchers measure integrity problems, while routine, relational, and structurally induced practices remain weakly captured. Aggregate harm is rarely defined explicitly and is instead inferred from perceived prevalence and severity [5]. ## Side-by-Side Comparison | Criterion | Research Misconduct | Questionable Research Practices | |--|--|--| | Core behaviors | Fabrication, falsification, plagiarism (FFP) | Selective analysis, selective reporting, selective citation, poor supervision, HARKing | | Intent | Intentional, knowing, or reckless | Often unintentional, normalized, or motivated by pressure | | Formal status | Formally defined, investigateable, sanctionable | Rarely formally defined, usually handled informally | | Who investigates | Institutional research integrity officers, government agencies | Mentors, editors, peer reviewers, the researcher personally | | Typical consequences | Retraction, termination, funding bans, debarment | Corrections, criticism, damaged reputation, wasted resources | | Detection | Whistleblowers, audits, plagiarism software | Rarely detected, often invisible in the published record | | Effect on the record | Findings are false or stolen | Findings may be unreliable, biased, or hard to replicate | | How common | Comparatively rare | Very common, admitted by most researchers in surveys | | Appropriate label | Use "research misconduct" when FFP with intent can be alleged | Use "QRP" when behavior falls short of FFP | ## Why the Line Between the Two Matters ### Consequences differ in severity A misconduct finding can end a career. Institutions can terminate appointments, funders can bar researchers from applying for grants, and journals retract the affected papers. Retraction data show how serious the consequences are in practice. A review of 333 retracted dental research articles found that more than half, 57.4 percent, were retracted because of misconduct such as plagiarism, data duplication, and data fabrication [6]. In rheumatology, scientific misconduct accounted for 75.3 percent of 381 retracted articles, with data fabrication, fake peer review, duplication, and authorship issues among the most common causes [7]. QRPs rarely trigger formal sanctions. The usual responses are criticism during peer review, corrections, replication failures, and editorial letters. Yet the aggregate damage is substantial, because QRPs distort the published record at a scale misconduct never reaches. ### Investigation procedures differ ![A branching flowchart from a witnessed concern through evidence review to a finding of misconduct or no misconduct, with federal oversight.](/article-images/illustrations/research-misconduct-vs-questionable-research-practices-misconduct-investigation.webp) Allegations of misconduct follow formal procedures. In the United States, research institutions appoint a Research Integrity Officer who handles allegations, secures evidence, and coordinates with federal oversight agencies. Preparing whistleblowers for what the process involves reduces stress and helps them make informed decisions about reporting, and integrity officers who cover more topics in the first contact are better at building that confidence [8]. Anonymity is possible but uncommon. Of 986 allegations received by the Office of Research Integrity and its predecessor from 1989 through 1997, only 8 percent of complainants remained anonymous, and few anonymous complaints were substantive enough to pursue [9]. QRPs have no equivalent procedure. Nobody opens a formal case because a co-author dropped an inconvenient condition from an analysis. The remedies are cultural: better training, better mentoring, preregistration of studies, and journals demanding fuller reporting. ### Whistleblowing dynamics differ ![A three-question decision path that sorts a suspected integrity problem into a quiet word, a QRP discussion, or a formal misconduct report.](/article-images/illustrations/research-misconduct-vs-questionable-research-practices-report-or-discuss.webp) People who witness QRPs face a genuinely hard decision, and willingness to report varies widely. A study of Malaysian higher education institutions found that less than half of respondents were willing to blow the whistle, citing weak protection of identity, tedious investigation processes, and cultural norms of deference to senior colleagues [10]. Because QRPs are normalized, witnesses often do not even recognize them as reportable problems. Qualitative work with dental students and mentors in Karachi found that students described QRPs as normalized research shortcuts and relied on mentorship as ethical scaffolding, meaning mentors either reinforced or corrected those habits [11]. ## How Common Is Each Category? The evidence base paints a consistent picture. Outright misconduct is a minority behavior. QRPs are close to universal. In the largest cross-national QRP survey to date, 9 out of 10 researchers admitted using at least one QRP in their recent publications, with a median of three out of nine QRP items. Perceived prevalence estimates ran roughly three times lower than self-reported use, suggesting researchers view their own behavior as less widespread than their colleagues'. Perceived pressure was the strongest systemic predictor of prevalence [12]. A survey of 1,573 researchers across six fields in Portugal found QRPs were widespread across all fields and seniority levels, with younger and more prolific researchers admitting more QRPs [13]. Misconduct shows lower but still concerning rates where measured. The Ethiopian survey cited above found 37.7 percent of researchers reported engaging in at least one form of misconduct, with publication pressure significantly associated with misconduct (an adjusted odds ratio of 3.18, meaning researchers under publication pressure had roughly three times the odds of reporting misconduct) [2]. There is also evidence that QRPs are linked to how researchers think about success. Using an implicit association test with 11,747 scientists in Austria, Germany, and Switzerland, researchers found that only a very low fraction of respondents associated serious misconduct with success, but roughly one in five associated QRPs with success. That association declined with career status, and was most notable among PhD students [14]. This suggests QRPs can function as a perceived career strategy, especially early in training. ## The Gray Zone: Honest but Unacceptable Practices The misconduct-versus-QRP distinction is itself being refined. Clinician researchers have argued for a third category, honest yet unacceptable research practices. These are practices that involve no dishonesty but still cause harm through sheer prevalence and consequences, particularly in health research where flawed studies can affect patient care [15]. A poorly designed trial with an honest analysis can still mislead treatment decisions just as much as a deliberately manipulated one. Psychology has produced a parallel critique at a deeper level. Some researchers argue the field's replicability problems stem not only from QRPs but from questionable research fundamentals, meaning poorly developed theories, concepts, and measurement approaches that no amount of honest practice can fix [16]. The lesson for students is that integrity is necessary but not sufficient. A study can be honest and still be bad science. ## Worked Scenarios ### Scenario 1: The invented data point A graduate student running a cell viability assay loses three plates to contamination the night before a deadline. Rather than repeating the experiment, the student copies values from an earlier plate and enters them as new data. This is fabrication, which is research misconduct. The data were invented, the intent was to deceive, and the behavior departs from accepted practice. The correct response is to report the lost plates, repeat the work, and document what happened. If you observe this as a lab mate, it belongs in a formal report to your research integrity officer or equivalent, not a quiet word. ### Scenario 2: The flexible analysis A researcher tests twelve possible statistical models on a dataset, finds one that produces a significant result, and reports only that model without disclosing the others. No data were invented or altered. Every number in the paper is real. This is a QRP, specifically selective analysis, one of the most commonly reported QRPs in survey data [4]. It can also be described as p-hacking, the practice of exploiting analyst flexibility until significance appears [3]. It is not misconduct under the formal definition, because no data were falsified and the behavior may not be reckless. But it produces a result that will likely fail replication, and it belongs in the category of practices that mentorship, preregistration, and reporting standards are designed to prevent. ### Scenario 3: The courtesy author A department head is added to a paper despite contributing nothing. The actual experimental work was done by a junior researcher whose contribution is then minimized in the author order. This sits at the boundary. If the addition was a deliberate exchange for favors, or if a deserving contributor was removed knowing the contribution requirements, it rises toward authorship misconduct, which the editorial literature treats as professional misconduct [1]. If it reflects unexamined departmental habit, it functions as a QRP that normalizes unfair credit allocation. Either way, authorship problems were the most commonly reported misconduct category in the Ethiopian survey, and students should learn authorship criteria before their first submission [2]. ## Common Mistakes and Limitations - **Calling every QRP misconduct.** Overusing the term "research misconduct" for ordinary sloppiness damages careers and makes formal channels less credible. Reserve the term for fabrication, falsification, or plagiarism with intent, knowingness, or recklessness. - **Treating QRPs as harmless.** The distinction is about formality, not harm. With 9 out of 10 researchers admitting at least one QRP [12], the cumulative distortion of the literature from QRPs likely exceeds that from misconduct. - **Confusing use with prevalence.** Survey researchers distinguish self-reported use, meaning whether someone has ever used a practice, from prevalence, meaning how often it occurs in the literature. Conflating the two inflates estimates. - **Assuming intent is obvious.** The same behavior, such as excluding data points, can be a defensible methodological decision or falsification depending on justification and documentation. Allegations require evidence, not suspicion alone. - **Believing codes of conduct solve the problem alone.** A study of researchers in Europe, Canada, Australia, and the USA found that awareness of institutional integrity policies, belief in their effectiveness, and confidence in them all related to lower QRP engagement, but the relationship depended heavily on researchers' own integrity self-confidence and attitudes [17]. Policies work only when researchers accept them. - **Assuming the problem is confined to one country or field.** Response patterns in QRP surveys are remarkably similar across countries, with a statistical agreement of 0.85 between Danish and international respondents [12]. QRPs appear across fields, seniority levels, and career stages [13]. - **Ignoring early training.** QRPs become normalized during undergraduate and graduate training, which is when intervention has the best chance of changing habits [11]. Peruvian university students showed high knowledge of integrity concepts but only regular practice levels, showing that knowing the rules does not guarantee following them [18]. ## Responsible Conduct of Research: How the System Addresses Both The responsible conduct of research framework treats integrity as a set of positive standards, not just a list of prohibitions. Codes of conduct operationalize this. The Netherlands Code of Conduct for Research Integrity, for example, sets out five core principles (honesty, diligence, transparency, independence, and social responsibility) with 61 standards. When 332 PhD candidates rated those standards, 83 percent of clarity ratings and 77 percent of relevance ratings scored in the top two categories, but perceived frequency of non-adherence varied across standards from 5 to 34 percent, and candidates saw some standards as unclear or hard to apply in practice [19]. This shows that even well-designed codes need interpretation and discussion, not just distribution. Practical steps for students and early-career researchers: 1. **Preregister your studies.** Specifying hypotheses, sample sizes, and analyses before data collection removes the flexibility that enables selective analysis. 2. **Report all outcomes.** Journals increasingly require disclosure of all measured variables and all analyses conducted. 3. **Document your data trail.** Raw data, analysis scripts, and lab notebooks protect you from unfounded allegations and make honest error correctable. 4. **Discuss authorship early.** Agree on authorship criteria at the start of a project, not at submission. 5. **Know your reporting channels.** Find out who your institution's research integrity contact is before you need them. Whistleblowers who understand the process in advance report less stress and make better decisions [8]. 6. **Treat mentorship as part of the method.** Students absorb norms from how their mentors handle ambiguous cases, so choose environments where integrity is discussed openly [11]. Institutional policy matters, but so does the research climate. The empirical literature increasingly takes a systemic view, examining supervision, organizational conditions, and climate rather than only individual behavior, and it highlights the need to integrate individual perceptions with institutional conditions when assessing integrity risks [5]. ## Frequently Asked Questions ### What is the main difference between research misconduct and questionable research practices? Research misconduct is the deliberate fabrication, falsification, or plagiarism of data or ideas, and it triggers formal investigations and sanctions. Questionable research practices are lower-severity behaviors like selective reporting or p-hacking that fall short of FFP but still undermine the reliability of science. ### Is p-hacking research misconduct? No. P-hacking, meaning analyzing data repeatedly until a significant result emerges, is a QRP because no data are invented or altered. It is still harmful, since it inflates false positive findings, and some journals now require disclosure of all analyses to prevent it [3]. ### When should I use the term research misconduct? Use "research misconduct" only when you can point to fabrication, falsification, or plagiarism committed intentionally, knowingly, or recklessly, and when formal reporting channels are appropriate. For ambiguous behavior such as sloppy methods or selective citation, use "questionable research practice" and address it through discussion, correction, or improved training. ### Can a questionable research practice get a paper retracted? Retraction for QRPs alone is uncommon, but QRPs often accompany or lead into misconduct that does end in retraction. In retracted rheumatology articles, misconduct accounted for 75.3 percent of retractions, and the median time from publication to retraction was 18 months [7]. ### How common are questionable research practices? Very common. In the largest cross-national survey, 9 out of 10 researchers admitted using at least one QRP in their recent publications, with a median of three out of nine practices, although prevalence in the literature was estimated roughly three times lower than self-reported use [12]. ### What should I do if I suspect a colleague of misconduct? Document what you observed factually, avoid confronting the person directly about a serious allegation, and contact your institution's research integrity officer or equivalent confidential channel. Formal allegations follow defined procedures, and integrity officers are expected to prepare whistleblowers for the process, which reduces stress and protects everyone involved [8]. ## Sources 1. [Authorship misconduct: professional misconduct in editorial handling of authorship.](https://pubmed.ncbi.nlm.nih.gov/39551550/) 2. [Magnitude and Factors Associated with Research Misconduct at a Public University in Ethiopia: A Cross-Sectional Survey.](https://pubmed.ncbi.nlm.nih.gov/40026541/) 3. [Questionable Research Practices: p-Hacking.](https://pubmed.ncbi.nlm.nih.gov/40048601/) 4. [Questionable Research Practices: A Principled Classification and Ranking Based on Survey Data.](https://pubmed.ncbi.nlm.nih.gov/41963575/) 5. [Scientific misconduct, questionable research practices, and research climate: empirical approaches to integrity problems.](https://pubmed.ncbi.nlm.nih.gov/42324565/) 6. [Evaluation of retracted publications related to oral health: a scoping review.](https://pubmed.ncbi.nlm.nih.gov/40118981/) 7. [Retractions in Rheumatology: Trends, Causes, and Implications for Research Integrity.](https://pubmed.ncbi.nlm.nih.gov/41555662/) 8. [Preparing whistleblowers for reporting research misconduct.](https://pubmed.ncbi.nlm.nih.gov/23009270/) 9. [Anonymity and pseudonymity in whistleblowing to the U.S. Office of Research Integrity.](https://pubmed.ncbi.nlm.nih.gov/9609854/) 10. [Whistle blowing and research integrity: Potential remedy for research misconduct in Malaysian institutions of higher education.](https://pubmed.ncbi.nlm.nih.gov/30489163/) 11. [Shaping research integrity: Perceptions of mentorship and questionable research practices among undergraduate dental students and mentors in Karachi, Pakistan.](https://pubmed.ncbi.nlm.nih.gov/42343295/) 12. [Is something rotten in the state of Denmark? Cross-national evidence for widespread involvement but not systematic use of questionable research practices across all fields of research.](https://pubmed.ncbi.nlm.nih.gov/39133711/) 13. [The 'shades of grey' in research integrity-Researchers admit to questionable research practices that they do not perceive to be serious.](https://pubmed.ncbi.nlm.nih.gov/41525269/) 14. [Are questionable research practices considered a successful career strategy? A novel implementation of the implicit association test.](https://pubmed.ncbi.nlm.nih.gov/40861029/) 15. [Honest yet unacceptable research practices: when research becomes a health risk.](https://pubmed.ncbi.nlm.nih.gov/40541433/) 16. [Psychology's Questionable Research Fundamentals (QRFs): Key problems in quantitative psychology and psychological measurement beyond Questionable Research Practices (QRPs).](https://pubmed.ncbi.nlm.nih.gov/40927341/) 17. [Shared responsibility to address questionable research practices? - A study of perceived efficacy of organizational research integrity policies.](https://pubmed.ncbi.nlm.nih.gov/42240276/) 18. [Knowledge and practices about research integrity among Peruvian university students.](https://pubmed.ncbi.nlm.nih.gov/41388540/) 19. [How do PhD candidates perceive good research practices in the Netherlands Code of Conduct for Research Integrity?](https://pubmed.ncbi.nlm.nih.gov/42271526/)