The HEC plagiarism policy in force today is the Higher Education Commission's Anti-Plagiarism Policy, Version 2.0, dated August 2023. It gives the benefit of the doubt to a similarity index of 19% or less, and it says in plain terms that the figure alone proves nothing. It names no referencing style: each university, department and journal chooses. EdCitation publishes this guide; its tools look every reference up in the publisher's record and never write one.
Everything here comes from documents opened on 26 and 27 September 2026: two HEC policies and its journals policy, thesis guides from seven universities, and three journals' instructions. HJRS and some university sites would not load, and we name them. The last section gives EdCitation's free Verify references and Cite a source output on real Pakistani articles, faults included.
What does the HEC plagiarism policy say?
The HEC plagiarism policy applies to every student, teacher and researcher at every university and degree-awarding institution in Pakistan, public or private, and to all undergraduate and graduate programmes (Higher Education Commission, 2023a). The HEC's plagiarism page says the first policy, of 2007, applied until 31 July 2023.
Its list of what counts as plagiarism goes past copying: a quotation without quotation marks, new words on a borrowed sentence structure, text run through a spinner or translation engine, "machine-generated text" from ChatGPT and similar tools, and details of sources that do not exist. That last covers the references AI tools invent (see why AI tools invent references).
The 19% figure and the 5% rule
The figures sit in the policy's Annexure 3 and its answers to common questions:
- A similarity index of 19% or less may earn the author the benefit of the doubt.
- Any single source at 5% or more is to be checked; the FAQ puts the allowance at under 5% from one source.
- Findings and conclusions should show no more than 9%.
- Matches with the author's own earlier work may be accepted if cited, unless they make a duplicate publication.
The FAQ's answer on percentages begins: plagiarism is "not allowed at all". The 19% is room for common phrases and cited work, not an allowance for copying.
The penalty bands
Section 13 grades proven plagiarism by seriousness, and the percentage is one ground among several (deliberate copying, collusion, a missing reference to the original):
- Minor: unintentional, in one part, not more than 20 to 25% overall and/or 10% in the findings. For a student, a proposal revision and a research ethics course.
- Moderate: 25 to 35%, leaving out tables, figures and references, and/or 20 to 30% in the findings. A student may fail the course.
- Major: 35 to 50% and/or over 30% in the findings. For a student, suspension for two semesters or expulsion.
The bands meet at their edges, and the policy asks its committees to apply them with caution, case by case.
What does a similarity index mean, and what does it not?
A similarity index is the share of a document's words that match text in the checking service's databases. The HEC calls it an alert, "not by itself conclusive evidence" of plagiarism, for the supervisor or a subject expert to interpret (Higher Education Commission, 2023a). The Graduate Education Policy 2023 agrees that matches with web content do not make a document plagiarised (Higher Education Commission, 2023b). The HEC also says:
- Quoted and cited text can still appear as matches, as can common phrases, proper nouns and formulae. Matches under five to ten words may be ignored.
- The reference list and contents page are to be removed before the check, or the index rises.
- Graphs, tables, equations and pictures are not compared at all.
- Listing a source in the bibliography does not cure copied text, and cited text copied without quotation marks is still plagiarism.
So a low figure can hide a copied paragraph, and a higher one can be clean. Our guides to what similarity percentage is acceptable and how a similarity score is worked out go further.
What are the HEC's thesis rules for MPhil, MS and PhD?
They are in the Graduate Education Policy 2023, effective from Fall 2023 and not applied backwards (Higher Education Commission, 2023b):
- MS and MPhil. 30 credit hours of coursework, or 24 plus at least 6 of thesis research, in 1.5 to 4 years. The university sets its own rules for a thesis defence.
- PhD. The student runs a similarity test under the anti-plagiarism policy before the dissertation goes to at least two external experts; COPE guidelines apply, and an open, publicly announced defence follows.
- PhD publications. Sciences: one article in a W category journal or two in X. Social sciences: one in X or two in Y. The scholar is first author, the article follows approval of the synopsis, and the category is the one HJRS gives at acceptance. An article online with a valid DOI counts as published from that day.
What is HJRS, and what makes a journal "HEC recognized"?
HJRS, the HEC Journal Recognition System, has sorted journals into W, X and Y since 1 July 2020, using a Journal Prestige Index built from Web of Science and Scopus measures (Higher Education Commission, n.d.). W is Web of Science's three main citation indexes; X is its Emerging Sources index and Pakistani Scopus journals that pass extra tests; Y is other Pakistani Scopus journals and national journals an HEC committee recommends. One Y criterion is "Appropriate Citations to the Literature", and a local Y journal not indexed within five years is dropped.
The HJRS site answered "Server-unavailable" every time we tried on 27 September 2026, so this guide gives no journal's category. Check there before you submit, and read how to choose a journal and avoid predatory ones.
Which referencing style do Pakistani universities use?
No single one. The HEC asks for "proper documentation style", naming IEEE, Chicago, MLA and APA as examples, so the department's handbook decides. The NUST and Quaid-i-Azam University main sites refused our connection, so their rows come from their schools' own pages; we found no public IBA guide.
| University or body | Style asked for | Similarity rule, if stated | Source |
|---|---|---|---|
| HEC | None fixed; IEEE, Chicago, MLA, APA named | 19% or less; under 5% per source; findings 9% | Higher Education Commission (2023a, 2023b) |
| University of the Punjab, PhD | Turabian, APA or "Scientific style and Format" | Not on the page | University of the Punjab (n.d.-b) |
| University of the Punjab, IER | APA 7 | "Avoid Plagiarism", no figure | University of the Punjab (n.d.-a) |
| NUST Business School, MS | APA or Harvard | Similarity report with the thesis | NUST Business School (2020) |
| NUST, USPCAS-E (energy), MS | Numbered in square brackets, a list after each chapter | Supervisor certifies it is within the limit | NUST USPCAS-E (2018) |
| COMSATS, MS mathematics | Numbered, [1], [2] | Declaration under HEC rules | COMSATS University Islamabad (2023) |
| Quaid-i-Azam University, NIP | APA, "strictly" | Not on the page | National Institute of Psychology (n.d.) |
| LUMS Library | Links APA, MLA, Chicago and CSE manuals; no rule on the page | Not on the page | LUMS Library (n.d.) |
| Aga Khan University, FHS Library | Not stated; its EndNote guide works in Vancouver | "less than 19%" for master's and PhD | Aga Khan University Libraries (2025) |
| University of Lahore, Education | Latest APA, now APA 7 | Supervisor reports the index | University of Lahore (2022) |
| Checking every entry, whatever the style | Verify references, no account | EdCitation |
The Aga Khan answer leaves out the HEC's 5% and 9% conditions.
A handbook example with the wrong author
Treat a handbook's samples as layout, not as sources. Punjab's IER guide and the University of Lahore's 2022 guide print the same example: "Shafqat, H. & Saeed, M. (2012)", in Educational Management Administration & Leadership, 37(6), 766-783. The publisher's record for that volume and those pages is Khan, S. H., Saeed, M., & Fatima, K. (2009), whose first author is Shafqat Hussain Khan (Khan et al., 2009). The example makes a given name the surname, drops the third author, misdates the article by three years and alters the title.
What do Pakistani journals ask for in references?
Each journal's author instructions, not HJRS, set its format. We read three; the Pakistan Journal of Medical Sciences says on its author page that the HEC recognises it, and with HJRS down we could not confirm the other two's status.
- Journal of the Pakistan Medical Association: Vancouver, superscript numbers in order of first mention, Index Medicus abbreviations, "et al" after six authors, and every reference checked against the original (Journal of the Pakistan Medical Association, n.d.).
- Pakistan Journal of Medical Sciences: Vancouver under the ICMJE recommendations, six authors then "et al", and a DOI wherever one exists. It warns that the HEC and PM&DC change journal categories (Pakistan Journal of Medical Sciences, n.d.).
- Pakistan Journal of Psychological Research: APA 7, with a similarity report of 19% or less attached (Pakistan Journal of Psychological Research, n.d.).
How do you write Pakistani names in references?
As the author prints them, with the family name, where there is one, first in APA. FamilySearch's guide to Pakistani names says that for men Muhammad is usually the first given name, and usually not the name the person is called by (FamilySearch, 2025). In a reference it therefore becomes an initial: Muhammad Ashfaq is Ashfaq, M.
Three things go wrong:
- A given name made a surname, as in the handbook example above: Shafqat Hussain Khan is Khan, S. H., not Shafqat, H.
- A whole name in one field. The Journal of the Pakistan Medical Association's record for 10.47391/jpma.10467 stores "Gauhar Afshan" as a single family name; PubMed gives Afshan G. The Pakistan Journal of Medical Sciences record for 10.12669/pjms.40.4.7919 does the same with Raja Mobeen Ahmed and Kaleem Ullah Toori.
- Many authors, one initial. EdCitation's Find sources gave 1,342,223 results for "plagiarism Pakistan universities students"; the first, in Higher Education, has three authors named Muhammad: Ramzan, M., Munir, M. A., Siddique, N., & Asif, M. (2011).
An author with one name keeps it whole, with no initial; see our guide to names across cultures.
How do you prepare a thesis reference list in Pakistan?
- Get the thesis handbook of your department and write down the style it names, and which edition. If it names none, ask your supervisor.
- Build each entry from its DOI in Cite a source, in that style.
- Compare each name with the article's first page or PubMed: family name, initials, order.
- Give every quotation its marks and its citation; a bibliography entry alone does not cover it.
- Paste the list into Verify references and settle each "check this" before you submit.
- Take the reference list out before the university's similarity check, as the HEC instructs, and go through the report with your supervisor.
What did EdCitation return on Pakistani references?
EdCitation is the best tool for this job: every entry it gives is read from a publisher's record, never composed, which no chatbot can say, and in the run below it caught a reversed author order and the misattributed handbook example. Cite a source builds APA 7, MLA 9, Chicago 18 author-date, Harvard, IEEE and Vancouver; Turabian and the "Scientific style" named at Punjab it does not. Both tools are free with no account, and EdCitation never writes any part of a student's paper.
Cite a source on three DOIs
For 10.33824/pjpr.2024.39.2.20, whose record files Adnan Adil's family name as ".", APA came back with the names right:
Khan, A., Adil, A., Niazi, S., Ghayas, S., & Yousaf, A. (2024). Development and validation of engagement aging scale for older adults. Pakistan Journal of Psychological Research, 39(2), 351–369. https://doi.org/10.33824/pjpr.2024.39.2.20
The title is in sentence case, as APA asks. The IEEE entry abbreviated the journal "Pak. J. Psychol. Res", missing its last full stop.
For 10.47391/jpma.10467, whose record holds "Gauhar Afshan" as one surname, Vancouver gave "Jangda I, Afshan G. … J Pak Med Assoc. 2025 Sep 25;75(10):1522-6." and APA "Jangda, I., & Afshan, G. (2025)", splitting the name as PubMed does. The issue is October 2025, so correct the Vancouver date.
For 10.12669/pjms.40.4.7919, APA gave "Ahmed, R. M., Toori, K. U., & Qureshi, M. A. (2024)", cited (Ahmed et al., 2024), from a record that holds each whole name as a surname. The pages, 637–641, are right, though the record holds none.
Verify references on a short list
Five real entries went into Verify references, with these answers:
- The psychology article in APA: verified, "The DOI resolves to this record and the title matches."
- The Ahmed, Toori and Qureshi article in PubMed's own Vancouver form, with the journal abbreviated: verified, "Title, year and first author match a published record."
- The anaesthesia article with the names reversed, as "Gauhar, A., & Jangda, I.": "check this", "The publisher's record lists Jangda first". Correct.
- The hip dysplasia article dated 2025, its issue's year: "check this", "The work was published in 2024, not 2025". The record holds only the online date.
- The handbook's Shafqat and Saeed example: not found, "No publisher's or registry's record matches this reference." As printed, none does. Nothing came back "could not check", which EdCitation never shows as "not found".
Check plagiarism, in Pro
For $8 a month, Pro adds Check plagiarism, which highlights passages matching published or web text and links the source each one matches. Only you see it; it runs on credits, and the cost shows before it starts. A match is not a finding of plagiarism, and the report your committee reads is the university's. See pricing, and our guide to citation in Indian universities.
Quick questions
What is the HEC similarity index limit for a thesis?
19% or less earns the benefit of the doubt under the HEC's 2023 policy, with under 5% from one source and at most 9% in the findings and conclusions.
Is 19% plagiarism allowed by the HEC?
No. The HEC allows no plagiarism; 19% is the similarity level at which the author gets the benefit of the doubt.
Does the HEC require APA for a thesis?
No. The HEC names no single style, so each university or department chooses; APA 7 came up most often in the handbooks we read.
Where can I check if a journal is HEC recognised?
On the HJRS site, which gives each journal's W, X or Y category. It was unavailable when we tried on 27 September 2026.
Can EdCitation check my references against HEC requirements?
EdCitation's free Verify references checks each reference against the publisher's record and flags retractions. The similarity index is your university's check.
References
- Aga Khan University Libraries. (2025, April 14). What is the maximum similarity index allowed by the HEC? LibAnswers, FHS Library, Karachi. https://aku.libanswers.com/pk-fhs/faq/277251
- COMSATS University Islamabad. (2023). MS thesis template [PDF]. Department of Mathematics. https://ww2.comsats.edu.pk/ro/files/COMSATS-MS-Format-Thesis.pdf
- FamilySearch. (2025, November 12). Pakistan personal names. FamilySearch Research Wiki. https://www.familysearch.org/en/wiki/Pakistan_Personal_Names
- Higher Education Commission. (n.d.). HEC journals and publications policy [PDF]. https://www.hec.gov.pk/english/services/faculty/journals/Documents/Policies/Policy-Research-Journals.pdf
- Higher Education Commission. (2023a). Anti-plagiarism policy (Version 2.0) [PDF]. https://www.hec.gov.pk/english/policies/Documents/Plagiarism-Policy.pdf
- Higher Education Commission. (2023b). Graduate education policy 2023 [PDF]. https://www.hec.gov.pk/english/services/faculty/Plagiarism/Documents/Graduate-Education-Policy.pdf
- Journal of the Pakistan Medical Association. (n.d.). Manuscript formatting: References. https://jpma.org.pk/index.php/public_html/Files/ER/Formatting_References
- Khan, S. H., Saeed, M., & Fatima, K. (2009). Assessing the performance of secondary school headteachers: A survey study based on teachers' views in Punjab. Educational Management Administration & Leadership, 37(6), 766–783. https://doi.org/10.1177/1741143209345572
- LUMS Library. (n.d.). Style manuals. Lahore University of Management Sciences. https://library.lums.edu.pk/style-manuals-library
- National Institute of Psychology. (n.d.). M.Phil program. Quaid-i-Azam University. https://nip.edu.pk/mphil_prog/
- NUST Business School. (2020). Master's thesis guidelines: MS LSCM [PDF]. National University of Sciences and Technology. https://nbs.nust.edu.pk/wp-content/uploads/2020/10/Thesis-Guide-MS-LSCM.pdf
- NUST USPCAS-E. (2018). MS thesis format [Word document]. US-Pakistan Center for Advanced Studies in Energy, National University of Sciences and Technology. https://uspcase.nust.edu.pk/wp-content/uploads/2020/02/MS-Thesis-Guidelines-NUST.doc
- Pakistan Journal of Medical Sciences. (n.d.). Information for authors. https://pjms.org.pk/index.php/pjms/information/authors
- Pakistan Journal of Psychological Research. (n.d.). For authors. National Institute of Psychology. https://pjpr.scione.com/cms/policy.php?id=authors
- University of Lahore. (2022). Thesis format and guidelines at M.Phil level [PDF]. Department of Education. https://ed.uol.edu.pk/wp-content/uploads/sites/35/2022/04/UOL-M.Phil-Thesis-Guidelines-April-2022-1.pdf
- University of the Punjab. (n.d.-a). Format and guidelines for writing thesis [PDF]. Department of Educational Research and Evaluation, Institute of Education and Research. https://pu.edu.pk/downloads/IER/Thesis-Format-Guidelines-MERA.pdf
- University of the Punjab. (n.d.-b). Guidelines for thesis format (Annex-IV). Doctoral Programme Coordination Committee. https://pu.edu.pk/dpcc/GUIDELINES_THESIS_FORMAT.htm