On Being a Scientist: A Guide To Responsible Conduct In Research

On Being a Scientist: A Guide To Responsible Conduct In Research#

Book Summary Research

by the National Academy of Science, National Academy of Engineering and Institute of Medicine of the National Academies [1]

The National Academies Advisors to the Nation on Science, Engineering, and Medicine#

The National Academy of Science (NAS) 1https://www.nasonline.org/ (created in 1863) is a private, non-profit, self-perpetuating society of distinguished scholars that are required to advise the federal government on scientific and technical matters.

The National Academy of Engineers (NAE) 2https://www.nae.edu/ (created in 1964) is a parallel organization to NAS that also advises the federal government.

The Institute of Medicine 3Now links to the National Academy of Medicine (created in 1970) is a subsidiary of NAS aimed at advising the government about public health policy matters.

The National Research Council 4https://www.nationalacademies.org/home was created in 1916 to broadly advise on research matters. It is now the primary concern of both NAS and NAE.

Preface#

Science is built upon trust. When trust is misplaced and professional standards are violated, researchers feel attacked that the base of their profession has been undermined.

New researchers are not learning how to conduct themselves as scientists (a broad definition that includes anyone who is involved with the pursuit of knowledge) 5Very broad. Within software engineering there are efforts to distinguish Research Software Engineers — who support collaborators by developing software to support scientific pursuit — and scientists [16].. This used to be the responsiblility of more senior researchers, as they are role models, teachers, and mentors to junior researchers. But as the speed of scientific developments accelerates, this task is often ignored. Furthermore, exciting opportunities keep appearing faster than science can keep up, further contributing to the problem of not educating new researchers on best practices.

A Note on Using On Being a Scientist#

This book explores options for handling situations, not defining what exactly to do in a particular situation. It should be read by a group of researchers that discuss the topics within and figure out how to apply the lessons to their practice 6You brought friends, right?.

The existence of professional standards implies that there is a proper way to go about a task or field. However, this book doesn’t aim to cover specifics. To address specific concerns, a committee consisting of a diverse group of individuals from varying experience levels should be formed to identify and discuss what is the professional standards. Furthermore, if your professional has an established code of conduct or professional standards, those should be referred to for specifics.

The material within this book is not exhaustive, and many texts and websites now exist to help supplement this material.

Further reading awaits


For software engineering, consider the ACM Software engineering code of ethics [2], or the Agile Manifesto [3].

Introduction to the Responsible Conduct of Research#

Researchers pride themselves on observing or explaining phenomenon which haven’t been seen or explained previously. However, researchers are often under a great deal of personal and professional pressure; failure to properly collect, report, or handle data, experiments, or papers can result in lost time, money, and respect 7In my experience, this pressure is felt by all researchers..

Researchers have created professional standards over many centuries for many different fields. It is therefore expected that researchers will adhere to the standards of their respected fields. A scientific standard is the application of the ethical values of honesty, fairness, objectivity, openness, trustworthiness, and respect for others 8The scientific standard proliferates throughout the entirety of science, but scientific practices (i.e., the application of standards) differs from field to field..

Researchers have several obligations. Some of them include honoring their colleagues trust, conducting themselves in a responsible manner, and to the properly report their results to the public as their work has the potential to impact policies at an international level. A researchers irresponsible actions — or scientific misconduct — can potentially undermine an entire field and cause permanent damage to their reputation. Scientific misconduct is defined as, “fabrication, falsification or plagiarism in proposing, performing, or reviewing researcher, or in reporting research results,” by the US government 9Those that operate with scientific misconduct are said to work with questionable research practices. [4].

Advising and Mentoring#

It is often best to build a diverse community of mentors as not one person has the same experience, education, or background as another.

Being a mentor is not a one way street where the mentee only benefits. As a mentor, you experience the benefits of being exposed to new ideas, building a strong research program and collaboration network, and gain new friends plus respect from beginning researchers. However, mentors have influence over the start of a junior researcher’s career, and must therefore be careful not to abuse it. The main role of the mentor is to help the mentee move along a productive and successful career trajectory.

Mentees also have responsibilities towards their mentors. They need to develop clear expectations regarding meeting times and availability with their mentors, and need to seek out their mentors, rather than expect a mentor to be provided to them.

Further reading awaits


For more information on this subject, please read Adviser, Teacher, Role Model, Friend: On Being a Mentor to Students in Science and Engineering [5].

Treatment of Data#

Researchers who maliciously manipulate data to deceive others violate the basic values and accepted professional standards of science. In doing so they mislead colleagues, potentially impede progress in their field, undermine their own authority and trustworthiness, and they introduce misinformation into both the scientific record and the public sphere 10Retraction Watch is a watchdog that attempts to identify research misconduct..

To ensure the integrity of science, researchers have developed and continually improve upon methods such as statistical tests of significance, double-blind trials, and proper survey phrasing ensured by the academic’s relevant Internal Review Board (IRB). Because it is paramount that the integrity of the work is valid, papers must include a description of the methodology used to generate the data. This is because if the methodology to generate the data is incorrect or not present, then the integrity of the data and the work can be called into question. Furthermore, this helps ensure the reproducibility of science 11Well
 It’s getting there. A lot of science is irreproducible, as explained in this video. And thanks to computers, AI, and digital analysis, if the codes necessary to reproduce the findings in the paper are not made availible, it can be difficult to reproduce the study [17]..

New researchers are not trained on how to generate, share, store, and publish data. Additionally, new technologies to store, transmit, generate, and analyze data are constantly being created. Thus researchers face a dilemma of how to store data permanently so that it can be used in future works (either by themselves or with others). However, when academic works are released the data must be made available. If a researcher refuses to share the data, then they fail to maintain the standards of science 12Platforms to share code and data include GitHub, Zenodo, and FigShare..

Mistakes and Negligence#

All science is susceptible to error. However, researchers have an obligation to be as accurate and as careful as possible both for their profession and the public.

To make progress, researchers must believe the works of other researchers. It is well known that by, “looking for an answer”, or believing that there exists an answer prior to running experiments, that researchers bias themselves to believe that something exists when it doesn’t 13See the Litany of Tarski [18].. However, these inaccuracies and mistakes must be remedied in the scientific record as soon as they are identified so as to allow the continue proliferation of science 14Shoutout Retraction Watch!.

Research Misconduct#

Scientific misconduct, as defined by the U.S. Office of Science and Technology Policy, defines it as, “fabrication, falsification, or plagiarism in proposing, performing, or reviewing research, or in reporting research results” [4].

The specific definitions for fabrication, falsification, and plagiarism are:

  • Fabrication is “making up data or results.”

  • Falsification is “manipulating research materials, equipment, or processes, or changing or omitting data or results such that the research is not accurately represented in the research record.”

  • Plagiarism is “the appropriation of another person’s ideas, processes, results, or words without giving appropriate credit.”

Scientific misconduct significantly departs from accepted practices. Identification requires that misconduct be proven to be

    15However, identifying the intent to deceive is difficult to do, as honest mistakes can also result in incorrect data.

  • intendend to decieve the scientific process 15However, identifying the intent to deceive is difficult to do, as honest mistakes can also result in incorrect data.,

  • knowingly or recklessly conducted against norms, and

    • 16Thankfully, differences of opinion are not scientific misconduct.

  • a preponderance of evidence exists to prove the misconduct 16Thankfully, differences of opinion are not scientific misconduct..

To reduce scientific misconduct, co-authors should validate each others work, as well as have third parties review their results prior to publication. Misconduct can set back individuals, institutions, and research fields. It can also draw the eye of the public and policy makers to make decisions that are negative for the progression of science.

Responding to Suspected Violations of Professional Standards#

Science (typically) self-regulates itself 17A case study of when an outside body came in to regulate science is the 2023 U.S. Congressional hearings on antisemitism.. Research communities set standards and practices that researchers should follow. When confronting others about results, retaliation can occur. Thankfully, this is prohibited by U.S. laws [6].

All research institutions that receive federal funding must have methods to investigate and report misconduct. Anyone who is aware of misconduct must follow these procedures.

All scientists and research institutions should discourage questionable research practices (QRPs) 18This blog post goes into more detail about QRPs.. Additionally, they need to take responsibility in determining what QRPs warren reprimands from the institution 19Not all fields define QRPs the same (e.g., salami-publishing where researchers deliberately divide research results across multiple papers is viewed differently from field to field)..

To structure your thinking prior to confronting a researcher, consider talking to a friend or advisor who can keep the situation confidential. To raise suspicions about misconduct, ask neutral questions rather than be accusatory. If your institution receives federal funding, to talk to your research integrity officer. However, keep in mind your own motivations and biases prior to raising questions so as to be courteous and kind.

Human Participants and Animal Subjects in Research#

Any research done on humans or animals is subject to federal, state, local, and institutional regulations.

The U.S. federal regulations on experiments that involve humans is known as the Common Rule [1]. The Common Rule specifies exactly what falls under the domain of human participation in experiments. Furthermore, any research institution receiving federal grants must have an IRB. The job of the IRB is to ensure that all research involving human participants follows the Common Rule. Many private institutions also have IRBs.

In some cases where human participation is required, formal training in bioethics is needed to identify if the experiment is ethical towards the participants. These issues are often evaluated by the committee’s such as the President’s Council on Bioethics in the United States [8].

Animals also have federal protections with regard to animal participation in experiments under the U.S. Federal Animal Welfare Act [2]. The U.S. Public Health Service’s Policy on the Human Care and Use of Laboratory Animals [3] applies to all research supported by the National Institute of Health (NIH). Additionally, the Guide for the Care and Use of Laboratory Animals [4] is required to be followed by all NIH funded research institutions. This ensures that all animal testing follows the three R’s:

  • Reduction in the number of animals used,

  • Refinement of techniques and procedures to reduce pain and distress,

  • Replacement of conscious living higher animals with insentient material

Overseeing the ethical experimentation on animals is an institution’s Institutional Animal Care and Use Committees (IACUCs) 20Here is my alma mater’s IACUC page..

Laboratory Safety in Research#

Researchers should review safety information and procedures at least once a year.

A shortlist of topics to review includes:

  • Appropriate usage of protective equipment and clothing,

  • Safe handling of materials in laboratories,

  • Safe operation of equipment,

  • Safe disposal of materials,

  • Safety management and accountability,

  • Hazard assessment processes,

  • Safe transportation of materials between laboratories,

  • Safe design of facilities,

  • Emergency responses,

  • Safety education of all personnel before entering the laboratory, and

  • Applicable government regulations.

Sharing of Research Results#

In the 17th century, researchers kept their results secret in order to protect their results from others claiming them. To solve this, Henry Oldenburg [12] started to guarantee that not only researchers could have their results rapidly published, but that the Royal Society of London [13] would defend authors integrity. He furthermore invented the modern review practice of sending out manuscripts to experts prior to publication for review.

To publish in a peer-reviewed journal is often the most important way to share information among researchers. This is because the first to publish often gets the credit. Once results are published, others can cite the author in their work. Unpublished manuscripts under review are to be treated as confidential.

Citations need to be scrutinized during the peer-review process in order to assess their validity and usefulness 21OpenAlex is a great service to enable this.. Researchers have a duty to thoroughly search the available literature on a subject and attribute all relevant authors.

Information can be disseminated through seminars, conferences, abstracts, and posters as well as peer-reviewed journal papers. Often considered to be preliminary in nature, spreading data, results, or manuscripts that haven’t been peer-reviewed online is possible as well 22See arXiv.. However, by bypassing traditional quality assurance measures, researchers take on a risk that their work won’t serve science well 23For example, AI generated slop that inhibits scientific progress..

Publishing or announcing results prior to peer-review is dangerous. Without the peer-review checks, it is possible for misinformation to be misinterpreted by the public, policy makers, and other researchers as fact. Additionally, by posting information prior to a journal submission online, it can be considered to be a “prior publication”, thus making it unusable for publishing within the journal.

Publication abuses are possible. For example, researchers may publish the “least publishable unit” of information. This is when the smallest number of results are collected or written about to be published, rather than undergoing an intensive research process. This wastes the time and energy of reviewers and can give the researcher a reputation of producing shoddy or incomplete work.

Authorship and the Allocation of Credit#

When a paper is published, all of those that contributed to the paper are considered authors and should be listed as such. Authorship conventions and practices differ from field to field, and even lab to lab [14]. Open and honest conversations about how authorship is handled is paramount to ensure that everyone knows the rules for the author list order.

Honorary authors (i.e., those that contributed zero to little actual content to the research) should not be added in the author list. They can be acknowledged in an “Acknowledgment” section however. Additionally, ghost-writers should not be used to write a paper.

All authors on a paper must take responsibility for the content of the paper and must be able to answer questions about the research.

Intellectual Property#

Researchers should be aware of the potential monetary value that their intellectual property has. Intellectual property is a legal right to control the application of an idea within a specific context (through a patent) or control the expression of an idea (through a copyright).

  • Patent owners can protect their intellectual property rights by excluding others from making, using, or selling an invention so long as the patent owner provides a full description of how the invention is made, is used, and functions.

  • Copyrights protect the expression or presentation of ideas, but they do not protect the ideas themselves.

Most research institutions have policies that detail how intellectual policy rights should be handled. Additionally, patent and copyright laws differ from country to country. Publication of work should not be delayed because of a potential patent though.

Competing Interests, Commitments, and Values#

Conflict of interest refers to situations where researchers have personal, intellectual, financial, and professional interests that conflict with the ongoing research [15]. Conflicts of financial interests undergo much scrutiny. Should research be done with a known conflict of financial interest, it severely damages the researcher’s credibility and integrity. Timely and full disclosure of conflict of interests must be made to governing bodies of science when requested.

Personal relationships can also conflict with research. Often journals and funding agencies ask for personal relationships in order to identify potential conflict of personal interests. Romantic relationships often count as a personal conflict of interest.

Conflicts of commitment are different than conflict of interests. This can cause significant strain on a researcher’s life and should be identified early.

The values that a researcher has can cause a conflict of interest as well. Religious, philosophical, cultural, or political beliefs can all cause conflicts of interests. Values cannot and should not be separated from science, but should be acknowledged and understood while conducting science.

The Researcher in Society#

Researchers have the duty to ensure that not only does their work further science, but also improves the public’s welfare. They are put into an interesting situation within the public sphere and may have to wear many hats. However, it is paramount that they remain honest, fair, collegial, and open in society as these are all core scientific tenants that researchers should follow.

Bibliography#

[1]

Committee on Science, Engineering, and Public Policy. On Being a Scientist: A Guide to Responsible Conduct in Research: Third Edition. National Academies Press, Washington, D.C., third edition edition, March 2009. ISBN 978-0-309-11970-2. URL: http://www.nap.edu/catalog/12192 (visited on 2025-11-11), doi:10.17226/12192.

[2]

Don Gotterbarn, Keith Miller, and Simon Rogerson. Software engineering code of ethics. Commun. ACM, 40(11):110–118, November 1997. URL: https://dl.acm.org/doi/10.1145/265684.265699 (visited on 2025-11-11), doi:10.1145/265684.265699.

[3]

Kent Beck, Mike Beedle, Arie van Bennekum, Alistair Cockburn, Ward Cunningham, Martin Fowler, James Grenning , Jim Highsmith, Andrew Hunt, Ron Jeffries, Jon Kern, Brian Marick, Robert C. Martin , Steve Mellor, Ken Schwaber, Jeff Sutherland, and Dave Thomas. Manifesto for Agile Software Development. 2001. URL: https://agilemanifesto.org/ (visited on 2025-11-11).

[4] (1,2)

The Office Of Research Integrity. Definition of Research Misconduct. URL: https://ori.hhs.gov/definition-research-misconduct (visited on 2025-11-11).

[5]

Committee on Science, Engineering, and Public Policy. Adviser, Teacher, Role Model, Friend: On Being a Mentor to Students in Science and Engineering. National Academies Press, Washington, D.C., July 1997. ISBN 978-0-309-06363-0. URL: http://www.nap.edu/catalog/5789 (visited on 2025-11-11), doi:10.17226/5789.

[6]

The Office of Research Integrity. ORI Statutory Authority - 42 U.S.C. § 289b. November 2025. URL: https://ori.hhs.gov/ori-statutory-authority-42-usc-%C2%A7-289b (visited on 2025-11-11).

[7]

U.S. Department of health And Human Services. Federal Policy for the Protection of Human Subjects ('Common Rule). June 2009. URL: https://www.hhs.gov/ohrp/regulations-and-policy/regulations/common-rule/index.html (visited on 2025-11-11).

[8]

George W. Bush. President's Council on Bioethics. November 2001. URL: https://www.federalregister.gov/documents/2001/11/30/01-29948/creation-of-the-presidents-council-on-bioethics (visited on 2025-11-11).

[9]

U.S. Department Of Agriculture. Animal Welfare Act. 2025. URL: https://www.nal.usda.gov/animal-health-and-welfare/animal-welfare-act (visited on 2025-11-11).

[10]

U.S. Department of Health and Human Services. PHS Policy on Humane Care and Use of Laboratory Animals. 2015. URL: https://olaw.nih.gov/policies-laws/phs-policy.htm (visited on 2025-11-11).

[11]

National Research Council. Guide for the Care and Use of Laboratory Animals: Eighth Edition. National Academies Press, Washington, D.C., 8 edition, December 2011. ISBN 978-0-309-15400-0. URL: http://www.nap.edu/catalog/12910 (visited on 2025-11-11), doi:10.17226/12910.

[12]

Henry Oldenburg. July 2025. Page Version ID: 1298364691. URL: https://en.wikipedia.org/w/index.php?title=Henry_Oldenburg&oldid=1298364691 (visited on 2025-11-11).

[13]

The Royal Society. The Royal Society. 2025. URL: https://royalsociety.org/ (visited on 2025-11-11).

[14]

Xie He, Arash Habibi Lashkari, Nikhill Vombatkere, and Dilli Prasad Sharma. Authorship Attribution Methods, Challenges, and Future Research Directions: A Comprehensive Survey. Information, 15(3):131, March 2024. Publisher: Multidisciplinary Digital Publishing Institute. URL: https://www.mdpi.com/2078-2489/15/3/131 (visited on 2025-11-11), doi:10.3390/info15030131.

[15]

Gordon Long. Fundamental Research Security. JASON JSR-19-2I, The MITRE Corporation, December 2019. URL: https://nsf-gov-resources.nsf.gov/files/JSR-19-2IFundamentalResearchSecurity-12062019FINAL.pdf.

[16]

Florian Goth, Jan Philipp Thiele, and The teachingRSE Project. Foundational Competencies and Specializations of a Research Software Engineer. Computing in Science & Engineering, 27(2):27–34, April 2025. URL: https://ieeexplore.ieee.org/document/10930785/ (visited on 2025-10-21), doi:10.1109/MCSE.2025.3552156.

[17]

IEEE Author Center. Research Reproducibility. 2025. URL: https://journals.ieeeauthorcenter.ieee.org/create-your-ieee-journal-article/research-reproducibility/ (visited on 2025-11-11).

[18]

Eliezer Yudkowsky. The Meditation on Curiosity. October 2007. URL: https://www.lesswrong.com/posts/3nZMgRTfFEfHp34Gb/the-meditation-on-curiosity (visited on 2025-11-11).

Contents