INDIAN MARITIME UNIVERSITY (IMU)
RESEARCH & PUBLICATION ETHICS (RPE)
Course Work:-Subject: RPE02- Scientific Conduct
Faculty: Anil Kumar Mishra, IMU-MPC
Module 5: Selective Reporting and Data
Misrepresentation
1. What is Selective Reporting?
1.1 Definition and impact
Definition of Selective Reporting
Selective reporting occurs when researchers intentionally or unintentionally
publish only a subset of their data that supports a particular hypothesis while
omitting conflicting or negative results. This practice distorts the scientific
record and can mislead readers, policymakers, and other researchers.
Impact of Selective Reporting
1. Misrepresentation of Research Findings
➢ Hides contradictory or null results, leading to biased conclusions.
➢ Can exaggerate the effectiveness of a drug, treatment, or intervention.
2. Hindering Scientific Progress
1
➢ Other researchers may unknowingly repeat studies with suppressed negative
outcomes, wasting resources.
➢ Skews meta-analyses and systematic reviews, which depend on complete
datasets.
3. Ethical and Legal Consequences
➢ Journals and funding agencies increasingly demand data transparency to
counter selective reporting.
➢ Regulatory bodies can penalize researchers for withholding critical
information.
1.2 Case studies on selective reporting
Case Studies on Selective Reporting
1. Tamiflu Clinical Trials (2009-2014)
➢ Pharmaceutical company Roche selectively reported positive results of
Tamiflu’s effectiveness against flu while withholding unfavorable data.
➢ A later independent review revealed that the drug had limited benefits,
raising concerns about transparency in medical research.
2. Antidepressant Trials in the 2000s
➢ Studies on selective serotonin reuptake inhibitors (SSRIs) showed that
only positive trials were published, while negative trials were left
unpublished.
➢ When all data was considered, some SSRIs were found to be no more effective
than placebos.
3. Food Industry-Funded Sugar Studies
➢ In the 1960s, the sugar industry funded research that selectively
downplayed the role of sugar in heart disease while blaming fats instead.
➢ This influenced dietary guidelines for decades, impacting public health.
1.2.1 Preventing Selective Reporting
➢ Mandatory Data Sharing: Journals and funding agencies require full datasets.
➢ Preregistration of Studies: Researchers must define their methodology before
data collection to prevent post hoc data manipulation.
2
➢ Stronger Peer Review Standards: Reviewers should assess whether omitted data
affects conclusions.
[Link] of Data in Research
2.1 How data can be manipulated?
Data manipulation in research refers to unethical practices where data is
altered, misrepresented, or selectively reported to support a desired
conclusion. This can happen in various ways:
1. Fabrication – Creating fake data that was never collected.
➢ Example: A researcher invents survey responses to fit a hypothesis.
2. Falsification – Modifying or omitting data to mislead results.
➢ Example: Adjusting temperature readings in climate research to
exaggerate trends.
3. Selective Reporting (Cherry-Picking) – Publishing only favorable results
while ignoring contradictory findings.
➢ Example: A pharmaceutical company reports only successful trials of
a drug while hiding negative outcomes.
4. Misuse of Statistical Methods – Using incorrect statistical techniques
to exaggerate significance.
➢ Example: Repeating experiments until a statistically significant
result appears, then only reporting that outcome (p-hacking).
5. Image Manipulation – Editing visual data such as microscopy images or
graphs to mislead readers.
➢ Example: Altering an image in biological research to show
nonexistent results.
2.1 Consequences of data misrepresentation
3
Misrepresentation of data has severe consequences for researchers, institutions,
and society at large:
1. Scientific Repercussions
➢ Retraction of papers by journals.
➢ Loss of credibility in the research community.
2. Legal and Institutional Consequences
➢ Disciplinary actions by universities or employers.
➢ Legal penalties in cases involving fraud.
3. Impact on Public Trust and Policy
➢ Misleading research can influence policy decisions, public health,
and funding priorities.
Example: The fraudulent Wakefield study on vaccines led to a decline
in vaccination rates.
4. Ethical and Moral Ramifications
➢ Damage to research integrity and the ethical foundation of science.
➢ Hindered progress due to misleading findings.
By promoting transparency, accountability, and ethical reporting practices,
researchers can uphold the integrity of scientific inquiry.
[Link] Transparency in Research
3.1 Best practices for accurate reporting
Accurate reporting is essential to maintaining the credibility of scientific
research. Researchers must ensure that their data, methodology, and findings
are presented transparently and without bias. The following best practices help
uphold research integrity:
1. Full and Transparent Data Presentation
4
➢ Avoid selective reporting; include all relevant findings, whether
positive or negative.
➢ Use clear methodologies so results can be replicated by other
researchers.
2. Proper Citation and Acknowledgment
➢ Credit previous research to avoid plagiarism.
➢ Disclose conflicts of interest that may affect research outcomes.
3. Use of Ethical Statistical Methods
➢ Avoid data manipulation techniques like p-hacking or
misrepresenting statistical significance.
➢ Report exact values and confidence intervals instead of misleading
visualizations.
4. Adherence to Publication Guidelines
➢ Follow journal-specific requirements for research reporting,
including ethics statements and data availability.
➢ Register clinical trials and research protocols in public databases
before data collection begins.
3.2 Role of journals and peer review in preventing misrepresentation
Scientific journals and the peer-review process serve as essential safeguards
against research misrepresentation. They ensure that only valid, credible, and
ethically sound research gets published.
1. Editorial Oversight
➢ Journals check for plagiarism, duplicate submissions, and ethical
compliance before accepting a paper.
➢ Many journals now require raw data submission to verify research
findings.
2. Peer Review Process
5
➢ Experts evaluate research validity, methodology, and conclusions
before publication.
➢ Peer reviewers can request clarifications, revisions, or even reject
flawed studies.
3. Retraction and Correction Mechanisms
➢ Journals can retract or correct articles found to contain falsified
data or misrepresentation.
Examples: Retractions of fraudulent studies in medicine and
psychology (e.g., the Andrew Wakefield case on vaccines).
4. Encouraging Open Science
➢ Some journals promote open peer review and data-sharing policies to
enhance transparency.
➢ Preprint servers allow for community scrutiny before final
publication.
By following these best practices and ensuring strong peer-review mechanisms,
the scientific community can prevent the spread of misinformation and maintain
the credibility of research.