What Is Plagiarism? Definition, Types, and Global Data

TL;DR

Plagiarism is using someone else's words or ideas without attribution. The AMI's D4 dimension measures it. Pakistan (72) and Russia (72) score highest; Norway (32) and Sweden (38) lowest. Detection uses tools like Turnitin (90%+ of Q1 institutions); culture and policy matter as much as technology.

Definition

Plagiarism is the use of someone else's words, ideas, data, or other original work without proper attribution and presented as one's own. The misconduct lies in the misrepresentation — the reader is led to believe the student produced the work themselves when the work originated elsewhere.

The standard is intent-neutral in most institutional policies. Accidental plagiarism (forgetting to cite) is typically treated as plagiarism, though usually with lower sanctions than deliberate copying.

Types of plagiarism

Direct plagiarism

Verbatim copying without quotation marks or citation. The clearest and most easily detected form.

Mosaic plagiarism (patchwriting)

Combining phrases from multiple sources without attribution. Often produced by lightly editing copied text — replacing some words while preserving structure and content. Harder to detect than direct copying.

Paraphrasing plagiarism

Restating ideas in different words without citing the source. Sometimes called "idea plagiarism." The misconduct is in the failure to credit the original thinker, even though no words are directly copied.

Self-plagiarism

Reusing one's own previously submitted work without disclosure. Common in graduate work where students may want to develop a single research idea across multiple assignments. Most institutional policies require disclosure of any reuse.

Translation plagiarism

Translating from a foreign-language source and presenting the result as original work without citation. Difficult to detect with conventional tools, which match against same-language sources.

Image, data, and figure plagiarism

Using charts, figures, datasets, or images without permission or citation. Particularly relevant in scientific and technical disciplines.

Contract cheating as plagiarism

Submitting work written by someone else is typically classified as contract cheating (the AMI's D1 dimension) rather than plagiarism, but it shares the core misconduct of misrepresented authorship.

What the AMI data shows

D4 scores on a 0–100 scale across the 39-country set:

High D4 scores cluster in countries with limited mandatory detection deployment, large higher education sectors, and historically lower institutional emphasis on academic writing conventions. Russian and Pakistani D4 scores reflect documented dissertation-level plagiarism (Dissernet, Pakistani HEC).

Detection methods

Plagiarism detection tools

Turnitin, iThenticate, Copyleaks, Compilatio, Urkund, Antiplagiat (Russia), CopyKiller (Korea), JSA (Poland), and others compare submitted text against:

• Web content
• Published academic literature
• Previously submitted student work (institutional repositories)
• Subscription database content

Coverage and effectiveness vary by tool and by language. English-language detection is strongest; less-resourced languages have weaker coverage.

Cross-reference checking

Human review of suspicious passages against likely sources. Slow but effective for high-stakes work like doctoral dissertations.

Patchwriting detection

Advanced detection tools attempt to identify mosaic plagiarism through phrase-level analysis. Reliability is lower than direct match detection.

Why D4 scores vary across countries

Several factors drive cross-country variance:

Language market for detection

English-language detection tools (Turnitin, iThenticate) have larger corpora than tools for less-resourced languages. Countries with English-medium higher education benefit from stronger detection capability.

Mandatory deployment

Countries with mandatory plagiarism checking for theses (Poland's JSA, India's UGC requirements, Pakistan's HEC) capture more cases. The detection-prevalence confound means stronger detection produces higher reported rates even with the same actual prevalence.

Citation culture

Academic writing conventions vary across educational systems. Anglophone systems heavily emphasise citation and quotation marks; some other systems traditionally emphasised demonstration of mastery over source attribution, producing different baseline behaviour.

Survey availability

Country-specific D4 scoring depends on the availability of self-report survey data. The McCabe / ICAI surveys cover some countries directly; others rely on regional extrapolation, which compresses cross-country variance.

Sources

• McCabe, D. L. (ICAI / Rutgers). Multi-decade self-report survey programme
• Turnitin global deployment data
• AMI v1.5 methodology document
• Country-specific integrity literature

Full methodology | Download dataset

Related

Read the full methodology