Research Ethics and Integrity

Exam questions about research ethics rarely ask you to recite a definition in isolation. Instead, they present a short scenario, a researcher who reused a classmate's survey wording, a lab report with a suspiciously tidy data set, a bibliography missing a key source, and ask you to identify what went wrong and why it matters. Succeeding on this kind of question requires more than familiarity with terms like plagiarism or data fabrication; it requires a disciplined method for reading a scenario, isolating the relevant facts, and matching them against a specific ethical standard.

This lesson builds that method. It assumes you already understand the broad principles covered in responsible research conduct, honesty, transparency, and protection of participants, and focuses on the applied skill of evaluating concrete cases, recognizing plagiarism in its subtler forms, and using precise citation and data practices to support a defensible judgment.

A well-constructed scenario question embeds its ethical issue inside ordinary-sounding details. A researcher "streamlines" an inconvenient data point before submitting results; a student "borrows the structure" of a published argument without citing it; a team member receives co-authorship "because he has seniority." Each phrase signals a specific violation, selective reporting, structural plagiarism, and unearned authorship credit, but only if you slow down and ask what standard the action fails to meet.

Before deciding whether a scenario describes a violation, ask three questions: What action did the researcher actually take? Which stakeholder, a participant, a reader, a co-author, is affected by that action? And which principle, accuracy, attribution, consent, or fair credit, does the action either satisfy or breach? Scenario questions on exams are almost always answerable by working through these three questions in order rather than by searching for a single buzzword.

Outright copying is the easiest form of plagiarism to detect, and the rarest to appear on a well-written exam. More commonly, a scenario will describe patchwriting, where a passage is rearranged word by word while keeping the original's sentence structure and argument order, or mosaic plagiarism, where phrases from several sources are stitched together without any one being copied wholesale. Both count as plagiarism even though no single sentence matches a source exactly, because the underlying idea and its expression were taken without attribution.

A reliable test: if you removed the citation entirely, would a reader assume the idea and its phrasing originated with the writer? If so, and the idea did not originate with the writer, the passage is plagiarized regardless of how many individual words were changed. This test applies equally to a paragraph of prose, a data visualization copied without credit, or a research question lifted from an unpublished source.

Correct citation does more than satisfy a formatting requirement; it allows a reader to verify a claim and traces responsibility for an idea back to its source. A scenario that describes a citation pointing to the wrong source, omitting a co-author, or citing a secondary source as if it were primary, is testing whether you understand citation as a mechanism of accountability rather than a cosmetic rule.

Data integrity questions work the same way. A scenario might describe a researcher excluding outlier results without disclosing the exclusion, rounding figures to produce a cleaner trend, or reporting only the trials that supported a hypothesis. None of these actions alters the raw data directly, which is precisely what makes them easy to overlook; each one nonetheless distorts what a reader understands the evidence to show, and each therefore violates the same underlying standard of accurate, undistorted reporting.

The diagram below summarizes the sequence worth following on any exam question that presents a research scenario rather than asking for a definition directly.

Applying this to a sample scenario: "A student submits a lab report in which three trials that contradicted her hypothesis were left out of the final data table, though she recorded them in her notebook." Working through the sequence, the action is selective omission of recorded data; the affected stakeholder is any reader relying on the report to judge the hypothesis fairly; the standard at stake is accurate reporting, not fabrication, since the data was genuinely collected, but its exclusion without disclosure still misrepresents the evidence. A complete exam answer names this precisely as a data integrity violation rather than mislabeling it as plagiarism or fabrication, because each term describes a distinct failure even when the scenarios resemble one another on the surface.

Patchwriting: Rewording a source's sentences while preserving its structure and sequence of ideas without proper attribution, a subtler form of plagiarism.

Mosaic Plagiarism: Combining phrases or ideas from multiple sources into a new passage without crediting any of them.

Selective Reporting: Presenting only the data that supports a desired conclusion while omitting contradictory results without disclosure.

Data Integrity: The requirement that collected data be reported completely and accurately, without distortion through omission, rounding, or rearrangement.

Authorship Credit: The assignment of recognition for a research contribution based on actual intellectual or practical input rather than status or seniority.

Citation Accountability: The function of a citation in allowing a reader to trace a claim back to its original source and verify its accuracy.

Related lessons

• Responsible Research Conduct