Table of Contents
Academic integrity means doing academic work with honesty and responsibility, so that research and education stay trustworthy. Misconduct happens when someone copies without credit, makes up data, changes data dishonestly, or uses the same study/paper in unfair ways.
In Real Life: One copied paragraph or one fake data entry can break trust in the whole dissertation, even if 95% is genuine work.
Exam Point of View: UGC NET commonly asks you to identify the misconduct type from a short case and choose the most ethical action.
1. Academic Integrity and Misconduct
1.1 Academic integrity meaning
Academic integrity is “intellectual honesty” while proposing, performing, and reporting academic/research work.
It is the opposite of shortcut culture. It means your work must show originality, correct credit, and honest reporting.
1.2 Core values and principles you should remember
Many research integrity guides highlight common principles. A famous one is the Singapore Statement which gives four key principles.
- Honesty: Tell the truth in methods, data, and reporting.
- Accountability: Take responsibility for your work and its outcomes.
- Professional courtesy and fairness: Treat co-authors, participants, and peers respectfully.
- Stewardship: Careful use of resources and data (stewardship means “taking care like a responsible guardian”).
Another useful idea is Merton’s norms (scientific ethos). These are ideal rules for good science.
- Universalism: Judge ideas by evidence, not by person/status.
- Communality: Share knowledge for collective progress.
- Disinterestedness: Do research for knowledge, not personal gain.
- Organized skepticism: Critically check claims before accepting them.
1.3 Research misconduct meaning
A widely accepted definition says research misconduct includes fabrication, falsification, and plagiarism.
- Fabrication = making up data.
- Falsification = changing/omitting data to mislead.
- Plagiarism = using others’ work without credit.
1.4 Misconduct vs questionable practices
Some actions may not look “criminal,” but they still damage research quality.
- Misconduct (FFP): Fabrication, falsification, plagiarism.
- Questionable Research Practices:
- Selective reporting (show only good results)
- P-hacking (trying many tests until p-value becomes “significant”)
- HARKing (hypothesizing after results are known, but writing as if planned)
- Gift authorship (adding a name without real contribution)
- Ghost authorship (hiding the real writer/contributor)
- Citation manipulation (forcing unnecessary citations)
- Using predatory journals (fake/low-quality journals that accept fast for money)
Exam Point of View: When a question says “changed/removed data to fit hypothesis,” it is usually falsification, not “data cleaning.”
1.5 Consequences of misconduct
Misconduct is not only a personal problem. It harms the entire system.
- Wrong findings can mislead policy, education, medicine, and society
- Degrees lose value when dissertations are not original
- Careers can be blocked through penalties, bans, and loss of credibility
- Institutions lose reputation and funding opportunities
2. Plagiarism
2.1 Plagiarism meaning and scope
Plagiarism means taking someone else’s work or idea and passing it as your own.
It is not limited to copy-paste text. It can happen in:
- Ideas and theories
- Data and results
- Tables, charts, diagrams, images
- Tools like questionnaires and scales
- Computer code, scripts, algorithms
- Translations (copying by translating)
2.2 Types of plagiarism
| Type | What it means | Simple example |
|---|---|---|
| Direct plagiarism | Word-to-word copying | Copying a paragraph exactly |
| Mosaic plagiarism | Patchwork of copied phrases | Mixing copied lines with small edits |
| Paraphrase plagiarism | Same idea rewritten without credit | Synonyms used but no citation |
| Self-plagiarism | Reusing your own published text without disclosure | Same “literature review” used again |
| Idea plagiarism | Taking a concept/model without citation | Presenting a known theory as yours |
| Source plagiarism | Citing a source you did not actually use | “As per X” but copied from Y |
| Figure/table plagiarism | Using visuals without attribution | A borrowed chart without credit |
| Code plagiarism | Using code without license/credit | Copying GitHub code without mention |
Situational Example: A student rewrites a book paragraph using synonyms and submits it without citation. This is still plagiarism because the idea is borrowed.
2.3 Quotation vs paraphrase vs summary
- Quotation: Exact words + quotation marks + citation.
- Paraphrase: Same meaning in your own words + citation.
- Summary: Shorter main points + citation.
A simple rule works well.
Rule: If the idea is not yours, citation is needed in all three.
2.4 UGC plagiarism rules and regulations
UGC (2018) explains application, definitions, prevention duties, similarity exclusions, levels, panels, and penalties.
2.4.1 Who these regulations apply to
These rules apply to students, faculty, researchers, and staff in Higher Educational Institutions.
2.4.2 What HEIs must do (prevention duties)
UGC expects institutions to build a real anti-plagiarism system.
- Make plagiarism policy and publish it on the institution website
- Provide plagiarism detection tools and access to all researchers
- Train people to use plagiarism detection tools and reference management tools
- Ask every student to submit an undertaking that work is original and checked
- Ask supervisors to certify the work is plagiarism-free
- Submit soft copies of dissertations/theses to ShodhGanga and maintain an institutional repository
2.4.3 Similarity exclusions (what is NOT counted)
Similarity checks exclude these parts.
- Quoted work with proper permission and/or attribution
- References, bibliography, table of contents, preface, acknowledgements
- Generic terms, laws, standard symbols, standard equations
Note: UGC also mentions common knowledge/coincidental terms up to fourteen consecutive words are excluded, and core original sections like abstract, summary, hypothesis, results, conclusions should not carry similarity.
2.4.4 Similarity levels (UGC)
UGC quantifies plagiarism into levels.
- Level 0: Similarity up to 10%
- Level 1: Above 10% to 40%
- Level 2: Above 40% to 60%
- Level 3: Above 60%
2.4.5 Panels for handling complaints (DAIP and IAIP)
UGC describes a two-stage system.
- DAIP: Departmental Academic Integrity Panel (department-level inquiry and recommendation)
- IAIP: Institutional Academic Integrity Panel (final review and report to head of institution)
Both are expected to follow principles of natural justice (fair hearing).
2.4.6 Penalties for thesis/dissertation submission
| Level | Similarity | Penalty for students (thesis/dissertation) |
|---|---|---|
| Level 0 | Up to 10% | No penalty |
| Level 1 | >10% to 40% | Revise and resubmit within 6 months |
| Level 2 | >40% to 60% | Debar from resubmission for 1 year |
| Level 3 | >60% | Registration cancelled |
Note 1: Penalty for repeated plagiarism becomes one level higher than the previous level.
Note 2: If plagiarism is proved after degree/credit is already obtained, the degree/credit may be put in abeyance for a recommended period.
2.4.7 Penalties for academic and research publications
| Level | Similarity | Penalty for publications |
|---|---|---|
| Level 0 | Up to 10% | No penalty |
| Level 1 | >10% to 40% | Withdraw manuscript |
| Level 2 | >40% to 60% | Withdraw + deny 1 increment + no new supervision for 2 years |
| Level 3 | >60% | Withdraw + deny 2 increments + no new supervision for 3 years |
Note: Repeated Level 3 offence may lead to disciplinary action as per service rules (including suspension/termination).
2.5 Plagiarism in the digital era (images, code, AI)
Plagiarism today often hides inside “digital work.”
- Using a diagram from Google Images without attribution
- Reusing dataset columns from an online repository without citation
- Copying code snippets without license/credit
- Using AI tools to generate text but submitting as “my original writing” without verification and transparency
A safe practice is simple.
Practice: Treat software, datasets, online reports, and AI outputs like sources that may need credit and careful verification.
2.6 How to avoid plagiarism (complete practical method)
A strong prevention method looks like a routine, not a last-minute check.
- Write notes in two parts: your understanding + exact quote with source
- Put citations while writing the first draft
- Use quotation marks immediately for exact lines
- Paraphrase after closing the source (forces your own wording)
- Keep a running reference list from day one
- Use reference manager tools to format consistently
- Run a similarity check before final submission and revise properly
3. Fabrication of Data
3.1 Fabrication meaning
Fabrication means making up data or results and recording or reporting them.
This is not a “small shortcut.” It directly creates false knowledge.
3.2 Common forms of fabrication
- Creating fake survey responses to complete the sample size
- Inventing interview quotes that never happened
- Writing lab readings that were never measured
- Adding imaginary participants to increase “N” (sample size)
3.3 How to prevent fabrication (research habit list)
- Keep raw data safely with timestamps
- Use data collection logs and maintain proper records
- Store consent forms and participant codes separately
- Use version history so changes are traceable
- Ask for peer/supervisor verification at checkpoints
4. Falsification of Data
4.1 Falsification meaning
Falsification means manipulating processes or changing/omitting data so the research record is not accurately represented.
It often looks “smart” because some data is real, but the final story becomes dishonest.
4.2 Examples of falsification
- Removing outliers only because they reduce significance, without reporting
- Changing values to match the expected trend
- Editing images/graphs to make results look stronger
- Hiding negative results to show only success
4.3 Fabrication vs falsification vs honest data cleaning
Many students confuse falsification with “cleaning the dataset.” Data cleaning is allowed when it is transparent and rule-based.
| Action | Honest data cleaning | Falsification |
|---|---|---|
| Outlier handling | Pre-decided rule + reported in method | Deleted secretly to get better result |
| Missing values | Reported method (mean/median/imputation) | Filled with “convenient” values |
| Data correction | Fix entry errors with proof | Change values to support hypothesis |
| Image edits | Adjust clarity without changing meaning | Alter image to mislead |
Situational Example: If a researcher removes 10 “low scores” only because they weaken the hypothesis and does not mention it, it becomes falsification.
5. Duplicate Publication
5.1 Duplicate publication meaning
Duplicate publication means publishing substantially overlapping work again without clear disclosure and proper linkage to the earlier publication.
5.2 Why it is unethical
Major editorial guidance warns against duplicate submission/publication because it wastes review resources and can mislead the evidence base.
It can also create a false impression that “many studies” support a claim, when it is actually the same study repeated.
5.3 Common forms of duplicate publication
- Submitting the same manuscript to two journals at the same time
- Republishing the same study with very minor changes
- Publishing the same data with same conclusions in another journal
- Translating a paper and publishing again without disclosure
5.4 When overlap can be acceptable
Overlap can be acceptable when transparency is maintained, permissions are followed, and editors are informed (for example, justified secondary publications or properly disclosed versions).
6. Salami Publication (Fragmentation)
6.1 Salami publication meaning
Salami publication means slicing one meaningful study into multiple thin papers mainly to increase publication count, without strong academic justification.
6.2 Clear signs of salami publication
- Same sample, same method, same dataset repeated
- Each paper reports only a small piece with weak independent value
- Multiple papers look like “one study split into parts”
- Readers are misled into thinking these are separate independent studies
6.3 Duplicate vs salami vs ethical multiple outputs
| Point | Duplicate publication | Salami publication | Ethical multiple outputs |
|---|---|---|---|
| Core issue | Same paper repeated | One study split | Different strong questions from a larger project |
| Disclosure | Usually hidden | Often hidden | Proper cross-references and editor disclosure |
| Reader impact | Double-counting risk | Inflated research output illusion | Clear, useful, non-misleading contributions |
Exam Point of View: If the question says “same dataset, several small papers,” it most commonly points to salami publication.
7. Ethical Use of ICT in Research
7.1 Data security and privacy in digital research
Digital research uses online forms, cloud storage, emails, apps, and analytics tools. So ethics includes protecting participant data.
India’s data protection framework is strengthened by the DPDP Act, 2023 and DPDP Rules, 2025, which emphasize lawful use of personal data and clear responsibilities.
DPDP Rules, 2025 describe seven core principles.
- Consent and transparency
- Purpose limitation
- Data minimisation
- Accuracy
- Storage limitation
- Security safeguards
- Accountability
Practical actions for researchers:
- Use strong passwords and two-factor authentication
- Encrypt sensitive files (encryption means converting data into unreadable form without a key)
- Use access control (only required people can open files)
- Separate identifiers (name/phone) from research responses
- Maintain secure backups with version history
- Share only anonymized data when possible (anonymized means identity cannot be traced back)
7.2 Responsible use of software and digital resources
Ethical ICT use also means using digital tools honestly and legally.
- Use licensed software and respect copyright
- Cite software, datasets, and online tools you used
- Keep raw data and processed data in separate folders
- Maintain an audit trail (audit trail means a step-by-step record of what changed and when)
- Avoid predatory tools/services that promise “guaranteed publication”
7.3 Ethical use of AI tools in research work
AI can help in grammar, summarizing, and formatting, but it can also create false citations or incorrect facts.
Safe rules:
- Never use AI to generate or “fill” missing data
- Verify every citation and every statistic manually
- Do not upload confidential participant data into public AI tools
- Follow your institution/journal disclosure rules when AI assistance is used
Situational Example: If a student pastes participant interview transcripts into an online AI tool without consent and privacy safeguards, it becomes an ICT ethics violation even if there is no plagiarism.
Key Points – Takeaways
- Academic integrity is intellectual honesty in academic/research work.
- Research misconduct commonly includes fabrication, falsification, and plagiarism.
- Plagiarism is not only text copying; it includes ideas, data, figures, and code.
- Paraphrase and summary also need citation when ideas are borrowed.
Exam Point of View: If the idea is borrowed and there is no citation, the safest answer is usually “plagiarism,” even when words are changed.
- UGC excludes quotations with attribution, references, bibliography, and standard terms from similarity checks.
- UGC levels are based on similarity bands: 0, 1, 2, 3.
- UGC has strong penalties for thesis and publications, including withdrawal and supervision bans.
- Repeated plagiarism can lead to one level higher punishment.
Exam Point of View: Numbers-based questions are common, like “45% similarity = which level?” or “Level 2 thesis penalty = what?”
- Fabrication means making up data or results.
- Falsification means changing/omitting data so record becomes inaccurate.
- Duplicate publication repeats substantially overlapping work without proper disclosure.
- Salami publication splits one study into many thin papers and inflates output.
Exam Point of View: In assertion–reason questions, “double-counting evidence” strongly supports “duplicate publication is unethical.”
- DPDP Rules, 2025 list seven principles like purpose limitation and security safeguards.
- Ethical ICT means secure storage, controlled sharing, and responsible tool usage.
- Transparency (about methods, cleaning, and tools) is the easiest way to stay safe.
Research Integrity Workflow (Prevent → Detect → Correct → Report)
A workflow is a step-by-step routine that reduces mistakes and also protects you if someone questions your work.
1) Prevent
- Build citation habits from day one
- Maintain data logs and secure storage
- Use consent forms and proper approvals
- Use reference managers and plagiarism policies
2) Detect
- Similarity check before final submission
- Data audit (spot impossible values, missing blocks)
- Supervisor/peer review checkpoints
3) Correct
- Rewrite properly with citations
- Report cleaning rules transparently
- Maintain version history of changes
4) Report
- If misconduct is suspected, follow institutional mechanism using DAIP and IAIP style panels.
| Step | Goal | One strong practice |
|---|---|---|
| Prevent | Avoid mistakes early | Cite while writing |
| Detect | Catch issues before submission | Similarity + data audit |
| Correct | Fix transparently | Versioned corrections |
| Report | Handle fairly | Panel-based inquiry |
Examples
Example 1: A student copies definitions from a coaching PDF into an assignment and submits without citation. This is direct plagiarism because the words are borrowed and no credit is given.
Example 2: During survey research, a student collects only 60 responses but creates 40 additional fake responses to reach the required 100 sample size. This is fabrication because the added data never existed.
Example 3: A researcher removes low-scoring participants from analysis only because the result becomes statistically significant, and the method section does not mention this removal. This is falsification because the record is changed to mislead.
Example 4: Ravi was close to his dissertation deadline and felt pressure to “show perfect results.” He edited the dataset so the trend line looks strong and removed the negative results quietly. In viva, the examiner asked for raw data and a clear log of changes. Ravi could not explain the edits, and the entire research lost credibility because honesty was missing.
Quick One-shot Revision Notes
- Academic integrity = intellectual honesty in research work.
- Misconduct (FFP) = fabrication, falsification, plagiarism.
- Plagiarism includes words, ideas, figures, data, tools, and code.
- Quotation needs quotes + citation.
- Paraphrase needs citation.
- Summary needs citation.
- UGC excludes references/bibliography and properly attributed quotations from similarity count.
- UGC similarity levels = 0 (≤10), 1 (10–40), 2 (40–60), 3 (>60).
- Thesis Level 2 = 1-year debar from resubmission.
- Publication Level 3 = withdraw + 2 increments denied + supervision ban (3 years).
- Repeated plagiarism can lead to higher level punishment.
- Fabrication = made-up data.
- Falsification = changed/omitted data.
- Duplicate publication repeats overlapping work without proper disclosure.
- Salami publication = one study split into thin papers.
- DPDP Rules, 2025 core principles include purpose limitation and security safeguards.
Mini Practice
Q1) A student rewrites a paragraph from a book using synonyms and submits it without any citation because “I changed the words.” What is this?
A) Quotation
B) Proper paraphrase
C) Plagiarism
D) Fabrication
Answer: C
Explanation: Changing words without citing the source still steals the idea.
Q2) Which option correctly matches the concept?
A) Fabrication = changing data; Falsification = making up data
B) Fabrication = making up data; Falsification = changing/omitting data
C) Fabrication = duplicate publication; Falsification = salami publication
D) Fabrication = citation style; Falsification = bibliography
Answer: B
Explanation: Fabrication creates false data; falsification distorts real data.
Q3) Choose the correct statement(s).
- A summary never needs citation.
- Copying a table without attribution can be plagiarism.
- Paraphrasing needs citation if the idea is borrowed.
A) Only 1
B) Only 2 and 3
C) Only 1 and 2
D) 1, 2 and 3
Answer: B
Explanation: Summary also needs citation if ideas are borrowed, and tables/paraphrases need proper credit.
Q4) As per UGC levels, a similarity index of 45% falls under which level?
A) Level 0
B) Level 1
C) Level 2
D) Level 3
Answer: C
Explanation: Level 2 is above 40% to 60%.
Q5) Assertion (A): Duplicate publication is unethical because it can mislead the evidence base.
Reason (R): The same data may be counted twice and look like extra support.
A) Both A and R are true, and R explains A
B) Both A and R are true, but R does not explain A
C) A is true, R is false
D) A is false, R is true
Answer: A
Explanation: Editorial guidance warns against overlapping publications because they create misleading duplication.
FAQs
What is academic integrity in one line?
Academic integrity is intellectual honesty while creating and reporting academic or research work.
Is paraphrasing without citation still plagiarism?
Yes, because the idea is borrowed even if the words are changed.
What is the difference between fabrication and falsification?
Fabrication makes up data; falsification changes or omits real data.
What does UGC exclude from similarity checks?
Quoted work with attribution, references/bibliography, and standard terms are excluded.
What are DPDP Rules, 2025 useful for in research?
They highlight principles like purpose limitation and security safeguards for handling personal data.
Why is duplicate publication treated seriously?
It wastes editorial resources and can mislead readers through overlapping publications.