On Being a Scientist

A Guide to Responsible Conduct in Research (3rd Edition)

Committee on Science, Engineering, and Public Policy

Introduction: The Foundations of Scientific Trust

• "The scientific enterprise is built on a foundation of trust."
• Society trusts that research results are an honest, accurate reflection of a researcher's work.
• Researchers trust that their colleagues:
  • Gathered data carefully and kept accurate records.
  • Used appropriate analytic and statistical techniques.
  • Reported results accurately and treated others' work with respect.
• When this trust is broken, the profession's base is undermined, damaging the relationship between science and society.

Responsible conduct in science is driven by three distinct dimensions of obligation:

1. Obligation to Colleagues:

• Science is a cumulative enterprise; new work builds on previous results.
• Inaccurate or deceptive results waste others' time, resources, and grant funds.

2. Obligation to Oneself:

• Irresponsible conduct can ruin a career, make it impossible to earn a degree, lose funding, or prevent achieving tenure.

3. Obligation to the Public:

• Research directly impacts public health, safety, and policy (e.g., climate change, stem cell research, drug toxicity).
• Much scientific work is funded directly by taxpayer dollars.

Terminology: Values, Standards, and Practices

• Ethical Values:
  • The everyday moral rules that apply in life: honesty, fairness, objectivity, openness, trustworthiness, and respect.
• Scientific Standards:
  • The application of these values to research contexts (e.g., openness in sharing research materials, fairness in peer review, honesty in data reporting).
• Scientific Practices:
  • Specific procedures that vary by discipline or laboratory (e.g., data recording methods, standard software, authorship order).

Misconduct vs. Questionable Practices

• Research Misconduct (FFP):

  • Officially defined by the U.S. government as Fabrication, Falsification, or Plagiarism in proposing, performing, reviewing, or reporting research.
  • Requires a significant departure from accepted practices, committed intentionally, knowingly, or recklessly, proven by a preponderance of evidence.
  • Does not include differences of opinion or honest mistakes.

• Questionable Research Practices (QRPs):

  • Actions that violate standards but do not meet the FFP threshold (e.g., "salami publishing"—choosing the "least publishable unit" to inflate paper counts, or failing to share unique materials).
  • Discouraged through informal and institutional penalties.

Advising and Mentoring: Defining the Relationship

• The Academic Adviser:
  • Focuses primarily on the conduct of research, offering guidance and technical advice connected to the project.
• The Mentor:
  • Takes a deeper, personal, and professional interest in the long-term career development of a researcher.
  • Mentors suggest directions, offer encouragement, introduce trainees to networks, help them secure funding, and model ethical behavior.
• The Trainee's Role:
  • Trainees have an active obligation to establish clear expectations regarding availability, communication, and project timelines.

Case Study: "A Change of Plans"

• The Situation:
  • Deniz, a graduate student, returns from summer vacation convinced he can finish his Ph.D. in one more semester. He has already started drawing up a list of companies to apply to.
  • His adviser immediately objects, stating his current measurements are insufficient for his dissertation committee. She insists he stay at least two more semesters to perform more measurements.
  • Deniz wonders if her advice is self-serving, as her own federally funded research project would benefit significantly from the additional data points.

Student Task 1: The Thesis Stalemate

Group Discussion & Roleplay

Break into groups of three. Assign roles: Deniz (Student), The Adviser (Faculty PI), and The Thesis Committee Chair (Neutral Mediator). Discuss and resolve:

1. How should Deniz approach his adviser to compare what his current data shows versus what the new measurements would add?
2. Is it appropriate for Deniz to bypass his adviser and consult other members of his thesis committee directly?
3. Preventative Strategy: What concrete steps should Deniz and his adviser have taken in the preceding years to avoid this sudden misalignment of expectations?

The Treatment of Data: Manipulation and Integrity

• The Visual Threat:
  • In 2002, the Journal of Cell Biology began screening images in all accepted manuscripts.
  • Result: Approximately 25% of accepted papers showed evidence of inappropriate image manipulation.
  • Fraud: In 1% of papers, editors uncovered fraudulent manipulation that affected the scientific conclusions, leading to paper rejection.
• Data Selection Pitfalls:
  • Omitting outliers or choosing data that matches a preferred theoretical curve is a violation of basic scientific standards.
  • You must keep a permanent, unalterable, and complete record of all primary data in bound or secure digital systems.

Case Study: "The Selection of Data"

• The Scenario:
  • Deborah (graduate student) and Kamala (postdoc) measure a new semiconductor material using an expensive national laboratory neutron test.
  • During tests, unexpected electrical power fluctuations occurred. They suspect these fluctuations affected their detector, but don't know which data points were compromised.
  • A new theoretical curve predicts their results. Two anomalous data points deviate widely from this curve.
  • Kamala suggests dropping those two points, arguing they are obviously errors caused by the power fluctuations, and that keeping them will make the paper harder to publish.
  • Deborah is concerned that dropping them without evidence of a specific fault is inappropriate.

Mistakes, Negligence, and "Believing is Seeing"

All scientific research is susceptible to human error. However, science handles mistakes differently based on intent and care:

• Honest Error:
  • Working at the absolute limits of technology or detection.
  • Example: Adriaan van Maanen (early 20th century) made painstaking measurements of spiral nebulae and reported "unwinding motions" which proved they were within the Milky Way. Edwin Hubble later showed they were distant galaxies. Van Maanen's expectations had unconsciously biased his measurements. ("If I hadn't believed it, I never would have seen it.")
• Negligence:
  • Mistakes caused by haste, carelessness, inattention, or failure to follow disciplinary protocols. Negligence compromises public trust.

Case Study: "Discovering an Error"

• The Dilemma:
  • Esin (epidemiologist) and Kerem (statistician) have published two well-received papers modeling the spread of infections.
  • Kerem discovers a coding error in his simulation software.
  • The Correction: Fixing the error does not alter the average time it takes for an infection to spread (their main conclusion). However, the corrected model shows much higher uncertainty, making their predictions less definite.
  • Marie opposes sending a correction to the journals: "Both papers will be seen as suspect if we do that, and it doesn't affect the main average anyway." She wants to simply use the corrected model in future papers.

Student Task 2: The Errant Code

Think-Pair-Share

Turn to your neighbor and analyze the Esin and Kerem case:

1. What specific obligations do Esin and Kerem owe to the scientific community and public health officials who might be using their modeling papers?
2. If their code is open-source and posted online, does that change their obligation to publish a formal erratum in the journal?
3. The Compromise: Draft a 2-sentence response explaining how Esin and Kerem can frame their corrigendum/erratum to preserve their professional reputation while correcting the public record.

Research Misconduct: Case Study of Bell Labs (Jan Hendrik Schön)

• The Breakthrough (1998-2001):
  • Jan Hendrik Schön, a young physicist at Bell Labs, published a series of highly celebrated papers in Science and Nature describing molecular-level switching and organic superconductivity.
• The Suspicion (2001-2002):
  • Outside labs failed to replicate his results.
  • Scientists noticed identical noise patterns and identical curves in graphs representing entirely different experimental systems.

• The Investigation:
  • A committee discovered Schön kept no data notebooks. He had deleted his primary files from his computer because of "lack of storage space," and his physical devices had been thrown away or destroyed.
  • The committee concluded Schön fabricated or falsified data in at least 16 papers. Schön was fired, his doctoral degree from the University of Konstanz was revoked, and his coauthors retracted 25 papers.

Responding to Suspected Violations

• Science is Self-Regulating:
  • Because governments cannot police every laboratory bench, the scientific community must monitor itself.
  • If you witness a colleague engaging in misconduct (FFP), you have an unmistakable obligation to act.
• How to Raise Concerns:
  • Suspicions are best raised in the form of questions and requests for clarification rather than aggressive allegations.
  • Consult confidentially with trusted mentors, ombudsmen, or your institution's designated Research Integrity Officer (RIO).
  • Institutional responses are divided into a preliminary inquiry (to gather facts) followed by a formal investigation if evidence warrants.

Responding to Fraud: The Hwang Woo-Suk Case

• The Rise (2004-2005):
  • Dr. Hwang Woo-Suk (Seoul National University) reported groundbreaking stem cell cloning and the cloning of a dog (Snuppy) in Science. He became a global celebrity.
• The Fall (2005-2006):
  • Co-workers and whistleblowers raised concerns about the unethical procurement of human eggs and fabricated DNA profiling data.
  • An SNU investigation proved the papers were fraudulent. Science published an immediate retraction, and Hwang was prosecuted for fraud and embezzlement.

• The Societal Cost:
  • Years of peer-reviewer and replicator time and resources were wasted trying to reproduce fabricated stem cell lines.
  • Public trust in stem cell biology was severely damaged, slowing down legitimate therapeutic legislation.

Human Subjects and Animal Welfare in Research

• Human Participants:
  • Regulated by federal guidelines known as the Common Rule.
  • Requires research protocols to be reviewed and approved by an independent Institutional Review Board (IRB).
  • Core pillars: informed consent, minimization of risks, and equitable selection of vulnerable populations (Belmont Report).
• Animal Subjects:
  • Governed by the Animal Welfare Act and reviewed by Institutional Animal Care and Use Committees (IACUCs).
  • Trainees must strictly adhere to the "Three R's" of animal testing:
    1. Reduction: Minimize the number of animals used.
    2. Refinement: Refine techniques to minimize pain and distress.
    3. Replacement: Replace conscious animal subjects with insentient material or simulations whenever possible.

Authorship and the Allocation of Credit

Authorship is the "currency" of science, establishing both credit and accountability:

• Who Qualifies?
  • A person should be listed as an author only if they made a direct and substantial intellectual contribution to:
  • The design or execution of the research.
  • The interpretation of the data.
  • The drafting or critical revision of the manuscript.
• Ethics Violations in Authorship:
  • Honorary/Guest/Gift Authorship: Including supervisors or prominent names who did not contribute directly, to appease authority or inflate paper credibility.
  • Ghost Authorship: Excluding the person who actually wrote or designed the study.

Historical Case Study: Pulsars and Jocelyn Bell

• The Discovery (1967):
  • 24-year-old graduate student Jocelyn Bell noticed "a bit of scruff" on a paper chart recorder while operating a radio telescope she helped build under the supervision of Antony Hewish.
  • She recognized the recurrence of the signal, which proved to be the first discovered pulsar.
• The Allocation of Credit:
  • The initial announcement paper was published with five authors: Antony Hewish (first), Jocelyn Bell (second), and three others.
  • The 1974 Nobel Prize in Physics for the discovery was awarded to Antony Hewish and Martin Ryle. Jocelyn Bell was excluded.
  • This case sparked intense debate. Some argued Bell's primary observation was the critical act of discovery; others (including Bell) argued she was performing her expected role in a wider project designed by Hewish.

Student Task 3: The Authorship Arbitration

Classroom Debate

Consider the following two authorship disputes. Split the room into two halves:

• Dispute A: Robert (industry scientist) has developed a novel reactor mixing model using computer simulations. His supervisor did not make any intellectual contribution but demands to be listed as last author because he runs the division and provides the computational hardware.
• Dispute B: Jocelyn Bell (graduate student) versus Antony Hewish (adviser) on pulsar discovery.

The Task:

• Group 1 (Left Side of Room): Defend the strict intellectual contribution rule. Why must Robert deny his supervisor, and why should Jocelyn Bell have shared the Nobel Prize?
• Group 2 (Right Side of Room): Defend the institutional/supervisory view. Why is providing lab space, funding, and theoretical architecture enough to justify co-authorship?

Competing Interests, Commitments, and Values

• Conflict of Interest:
  • Situations where a researcher's financial, professional, or personal interests could bias their professional scientific judgment.
  • Example: Acceptable funding sources vs. biased sponsors. The tobacco industry funded the Center for Indoor Air Research to publish biased studies to counter the EPA's finding that secondhand smoke is a Class A carcinogen.
• Conflict of Commitment:
  • Strains arising from how researchers divide their time and energy (e.g., academic dissertation work vs. consulting for a private corporate sponsor).
• The Resolution:
  • Full, transparent, and timely disclosure to institutions and journals is the primary mechanism to manage and mitigate bias.

The Scientist in Society: Arthur Galston and Agent Orange

• The Discovery (1943):
  • As a graduate student at the University of Illinois, Arthur Galston discovered that a synthetic chemical could hasten the flowering of soybeans, enabling growth in cold climates.
  • He noted that at high concentrations, the chemical was highly toxic, acting as a defoliant.
• The Weaponization:
  • Military researchers read Galston's thesis and developed Agent Orange, spraying 50,000 tons of herbicides over Vietnam forests.

• The Scientific Response:
  • Horrified by the toxic impact on humans and agricultural land, Galston actively organized scientific opposition, circulating petitions and directly presenting toxicity data to President Nixon's science adviser in 1969.
  • The spraying was phased out in 1970.
  • Galston's Lesson: "I used to think one could avoid involvement in the anti-social consequences of science... I have learned things are not that simple. The only recourse is for a scientist to remain involved with it to the end."

Core Takeaways: On Being a Scientist

• Scientific Progress is Social: It relies completely on the shared integrity of thousands of individuals working across boundaries.
• Intervene and Engage: Science is not passive; it requires active curation of data, meticulous record-keeping, and the courage to correct mistakes.
• Maintain the Moral Authority: By modeling and upholding ethical standards in mentoring, authorship, data selection, and funding disclosures, you protect the reputation of the scientific enterprise.
• Stay Involved to the End: Scientists cannot decouple their physical discoveries from their ultimate ethical consequences in society.