Research integrity, and working with the people your study is about

Six dilemmas, the norms behind them, and your own statement

Research studio R9 of Energy Systems, Step by Step: sort everyday practices by the norm they serve (consent, credit, conflicts of interest, data licences, honest reporting), work through six fictional dilemmas with the norm and its source named, map who a study affects, and write an integrity statement checked against the course rubric.
  • Researcher · L4
  • Research studio
  • 30 min guided · about 30 min practice
  • After R1

The task

  1. Produce: an integrity statement for your own study, planned or done: funding and conflicts, contributions, data and licences, ethics and consent, stakeholders, corrections. “Pending” lines become your next actions.
  2. From: six dilemmas, the norms behind them and the people your study affects.
  3. Done when: the rubric's bar for integrity: licences checked, credit and conflicts stated, stakeholders treated fairly.

About 30 minutes guided here, then about 30 minutes on your own: drafting your statement and listing its next actions (permissions and consultation take longer, and come later).

Five norms

  • Consent (people agree, knowing what you will do): Tell survey participants how their answers will be used, and that they can say no. Get the ethics board's decision before the first interview.
  • Credit (who did what, and whose work you used): List who did what, using contributor roles. Cite the dataset, not only the paper that describes it.
  • Conflicts of interest (funding and ties, declared): State who funded the study and what they saw before publication. Declare that you advise the utility whose programme you evaluate.
  • Data licences (the terms that come with data): Read the data agreement before you post a co-author's data. Keep the attribution line and the licence when you share an edited map layer.
  • Honest reporting (all the results, and corrections): Say how scenarios were chosen, and report the results that weaken your story. Ask the journal for a correction when you find an error.

Dilemma 1 · A funder wants a headline · Conflicts of interest

A regional utility association funds your study of a battery for Prairie Town. Across four scenarios, bills fall by 2 to 9% (illustrative). Its draft press release says “Batteries cut bills by 9%”.

  • Agree: 9% is one of your results. Hard to defend. It reports the best case as the result. Letting a funder shape how results are reported is an unacceptable practice in the European code (§3.1).
  • Ask for “2 to 9%, depending on the scenario”, and name the funder in the paper and the release. Defensible. Accurate, with its range, and the support declared (§2.7). The funder still gets a usable sentence.
  • Decline to comment on the release. Partly. Your paper stays accurate, but the misleading number still goes out near your name. Better to correct it.

Source: ALLEA (2023), §2.7 and §3.1.

Dilemma 2 · A convenient scenario · Honest reporting

You ran eight scenarios. One shows a large benefit; seven show small ones. After seeing the results, a co-author suggests presenting the large one as “the central case”, with no other reason for choosing it.

  • Present all eight, say how each was chosen, and give the range. Defensible. Honest reporting keeps the runs that weaken the story: negative results can be as relevant as positive ones (§2.7).
  • Put the large one in the main text and the other seven in a supplement. Partly. Everything is there, but the emphasis misleads: a central case needs its own justification, not its result. Say in the main text how each case was chosen.
  • Present only the large one. Hard to defend. Omitting results without justification counts as falsification in the European code (§3.1).

Source: ALLEA (2023), §2.7 and §3.1.

Dilemma 3 · A co-author's data · Data licences

A co-author shared a rural co-op's hourly meter data with you under a data agreement. You want to post the data with your code so that others can rerun the study.

  • Post it with the co-op's name removed. Hard to defend. A missing name does not override the agreement, and hourly meter data can still point to households.
  • Post it: the paper is out, so the data are public now. Hard to defend. Publication does not change the agreement. Posting may break it and expose the co-op's customers.
  • Read the agreement; post the code, a synthetic sample and how to request the data; cite the data and credit your co-author. Defensible. As open as possible, as closed as necessary: a synthetic sample lets others run the code, and the route to permission is stated (§2.5).

Source: ALLEA (2023), §2.5 and §2.7.

Dilemma 4 · A community meeting · Working with stakeholders

Your model ranks three sites for a solar farm near Prairie Town (illustrative). At a public meeting, residents ask why their concern about losing farmland is not in the model.

  • Say what the model includes and leaves out, record the concern, and report back how you treated it. Defensible. Fair process and recognition: the people affected are heard and can see what changed. Treating farmland could mean a new criterion or a sensitivity run.
  • Explain that the model is objective, so the ranking stands. Hard to defend. A ranking encodes choices about what counts. Calling it objective dismisses the people it affects.
  • Promise that the site they oppose will be dropped. Hard to defend. Researchers do not decide where farms go. A promise you cannot keep costs trust.

Source: Jenkins et al. (2016); ALLEA (2023), §2.4.

Dilemma 5 · A survey without consent · Consent

A student on a DOE-supported project plans to survey renters door to door next week about their energy bills, recording names and addresses. No one has asked the ethics board.

  • Go ahead: a short survey is harmless. Hard to defend. Research with people needs an ethics determination first. For a DOE-supported study, even a survey that may be exempt needs one, from a review board, its office or their designee (§4.a).
  • Go ahead, but do not record names. Partly. Collecting less is good practice, but the ethics determination is still needed, and addresses identify people.
  • Pause; send the plan to the ethics board; add a consent statement; collect only what the question needs. Defensible. Prior review and informed consent, as DOE Order 443.1D requires for this DOE-supported study (§4.a); participants told how their data will be used (§2.5).

Source: DOE Order 443.1D (2026), §4.a; ALLEA (2023), §2.4 and §2.5.

Dilemma 6 · An error after publication · Honest reporting

Months after publication you find a unit slip: one plant's capacity was entered in kW instead of MW (1 MW = 1,000 kW). Fixing it changes your main result from 12% to 10% (illustrative).

  • Fix it quietly in the code repository. Partly. Helpful, but the published record stays wrong. Do both.
  • Tell your co-authors and the journal promptly, and ask for a correction that says what changed. Defensible. Authors issue corrections promptly, and get credit for doing so (§2.7). The journal decides which kind of notice fits.
  • Say nothing: the conclusion still holds. Hard to defend. The published numbers are wrong, and others may reuse them.

Source: ALLEA (2023), §2.7.

With JavaScript on, you sort the practices, choose an action in each dilemma before reading the norm, and write your own statement; your draft stays in this browser.

Cases: fictional, set around Prairie Town; their numbers are illustrative, and no real person or organisation is described.
Norms: from the primary sources listed below, checked on 3 November 2026.
Your study: rules on research with people differ by country, funder and institution; your own ethics board or office has the last word.

Two answers to the same task

The fact sheet both answers use (fictional)

Study
Would a battery at Prairie Town's co-op lower members' bills? A model on public grid data and the co-op's meter data, plus a household survey.
Funding
A regional utility association. Its contract lets it see the results one week before publication; it cannot change them.
Team
Author A shaped the question, ran the analysis and wrote the draft; author B cleaned and documented the data and edited the text.
Ties
Author B advises the co-op on tariffs.
Data
EIA-930 hourly grid data (public domain); the co-op's hourly meter data, shared under an agreement that bars passing them on.
Survey
The university's ethics board approved it before it began; each household gave written informed consent.
People
Results were shown to residents and the co-op board before publication. Residents asked how renters' bills would change.
Response
A renter case was added to the results, and a one-page summary of the change went to everyone at the meeting.

Weak answer · Developing

The authors declare no competing interests [1]. Data available on request [2]. All participants agreed to take part [3]. All authors contributed equally [4].

  1. Falls short: Funding is not mentioned at all, and author B's advisory role on the fact sheet is a competing interest to declare.
  2. Falls short: Which data, under which terms? Public data should be cited; restricted data need a stated route.
  3. Falls short: Agreement is not informed consent, and the ethics review on the fact sheet is not named.
  4. Falls short: It hides who did what; contributor roles make credit checkable.

Strong answer · Proficient

Funding: a regional utility association funded the study; it saw the results one week before publication and could not change them [1]. Competing interests: author B advises the co-op on tariffs [2]. Contributions: A, conceptualization, formal analysis, writing (original draft); B, data curation, writing (review and editing) [3]. Data: EIA-930, public domain, cited with its access date; the co-op's meter data are under an agreement that bars sharing, so the code, a synthetic sample and the request route are public [4]. Ethics: the university's ethics board approved the survey before it began; each household gave written informed consent [5]. Stakeholders: results were shown to residents and the co-op board before publication; residents asked about renters' bills, so a renter case was added and a summary sent to the meeting [6]. Corrections: readers can report errors to the corresponding author; any error that changes a result goes to the journal as a correction [7].

  1. Works: The funder is named, with its role and its limits.
  2. Works: An advisory tie is declared, not hidden.
  3. Works: Contributor roles (CRediT) make credit specific.
  4. Works: Each source with its terms: as open as possible, as closed as necessary.
  5. Works: Review before collection, and consent, both stated.
  6. Works: One concern, the change it led to, and how people were told: they can see how their view was treated.
  7. Works: A route for errors, promised before anyone finds one.
Weak answer Strong answer
Funding Not mentioned Named, with what the funder could and could not do
Data “On request” Each source with its terms and a route
People “Agreed” Ethics review, consent, one concern and the change it led to

Who your study owes what

Group Ring What it is owed In the dilemmas
You and your students The study team Training, a safe place to raise a problem, and credit for the work each person did. A survey without consent
Co-authors The study team Authorship agreed on stated criteria, contributor roles, and a say before anything is submitted. A convenient scenario; A co-author's data; An error after publication
Survey participants and households People the study is about or uses Informed consent, privacy, and only the data the question needs. A co-author's data; A survey without consent
Residents near the site People the study is about or uses To be heard, to see how their concerns were treated, and a share in who decides. A community meeting
The co-op that shared data People the study is about or uses The terms of the agreement kept, and the data cited. A co-author's data
The funder People who use or pay for it Disclosure of its support, and findings it cannot change. A funder wants a headline
The utility and the regulator People who use or pay for it Results with their limits, all scenarios, and no overclaiming. A convenient scenario; A community meeting
The journal and its readers People who use or pay for it All the results, prompt corrections, and the data or code where allowed. A funder wants a headline; A convenient scenario; An error after publication
Ratepayers People who use or pay for it A clear account of who pays and who benefits. A funder wants a headline
Future residents People who use or pay for it Effects that last, such as land use, counted even though they cannot speak at the meeting. A community meeting
Always, then three questions
Declare funding and ties in every study. Then: people or their data? Someone else's work or data? Could your reporting mislead readers? Each yes adds a duty.
Who is owed what
The team is owed credit; the people a study is about are owed consent and a hearing; those who use or pay for it are owed all the results and the funding declared.
Write it down
An integrity statement turns the norms into lines a reader can check.
Go deeperRules differ by place and funder
  • In the United States, research with people funded by federal agencies follows a common federal rule; the Department of Energy applies it through its order on human research subjects, built on the principles of the Belmont Report (1979): respect for persons, beneficence and justice.
  • In Europe, personal data in research fall under the General Data Protection Regulation, which the European code names when it asks researchers to tell participants how their data will be used (§2.5).
  • Journals add their own policies on authorship, competing interests and data availability: read them before you submit.
Go deeperWhat this studio leaves out
  • How to run interviews, focus groups or participatory workshops well: a qualitative-methods course, and a mentor, are the next step.
  • Misconduct investigations: the European code's §3.2 sets out how institutions handle them fairly.
  • Peer review: writing a review, and answering one point by point, is practised in the Researcher checkpoint (CK-R).

Check your own work against the rubric

Competence Developing Proficient (the bar) Strong
Integrity
Look for: Your statement covers funding and conflicts, contributor roles, the terms of every data source, ethics review and consent, and how the people affected were heard.		 Licences or credit unchecked. Licences checked, credit and conflicts stated, stakeholders treated fairly. Data and code released with documentation where allowed.
Go deeper · ResearcherFrom this statement to your capstone

The capstone research project asks for this statement with its proposal, and its mixed track (a model with a stakeholder or survey part) needs the ethics step before any data are collected. The data licences here are the ones you checked in R4; the Researcher checkpoint (CK-R) asks you to defend them under review.

Sources and what’s next

Norms named on this page

Norm Where it is written Link
Consent DOE Order 443.1D, §4.a, for DOE-supported research: review before research with people starts; a study that may be exempt still needs a determination by a review board, a human-subjects office or their designee. European code §2.4–2.5. energy.gov · approved 5 August 2026
Credit European code §2.7 (authorship criteria, contribution statements); CRediT, 14 contributor roles, proposed by Brand et al. (2015) and a standard (ANSI/NISO Z39.104-2022) since 2022. NISO 2022 · Brand et al. 2015 · checked 3 November 2026
Conflicts of interest European code §2.7 (declare conflicts and support) and §3.1 (funders must not shape reporting). ALLEA 2023
Data licences European code §2.5 (as open as possible, as closed as necessary; how to get permission); each source's own terms. EIA example · checked 3 November 2026
Honest reporting European code §2.7 (negative results, prompt corrections) and §3.1 (omitting data or results without justification is falsification). ALLEA 2023
Working with stakeholders Energy justice: who pays and who benefits, who decides, whose concerns are recognised; European code §2.4. Jenkins et al. 2016

Next

CR · Your own study: from a question to a report

Sources and further reading

Spotted an error, or stuck? Write to hlli@ksu.edu. Corrections are made in place and dated here.

Text licensed CC BY 4.0, code MIT: teachers and students may reuse and adapt it, with credit to Honglin Li.