{"resourceId":"plos-one-nguyen-duong-ai-cheating-factors-vietnam-20260928","versions":[{"version":"external-d1a567ad2fb026d4aba5f6a58d65255512f70f42cfbe8ad2d6c8d39f320c266c","resource":{"id":"plos-one-nguyen-duong-ai-cheating-factors-vietnam-20260928","title":"PLOS One: unclear AI-use boundaries and institutional gaps associated with self-reported AI-assisted cheating among 863 Vietnamese undergraduates","organization":"Nguyen and Duong (PLOS One)","sector":"Higher education student success / academic integrity and AI-use policy","geography":"Vietnam (single university of science and technology)","publishedAt":"PLOS One, published 28 September 2026","publicationDate":"2026-09-28","eventDate":null,"sourceName":"PLOS One","sourceLabel":"Peer-reviewed cross-sectional survey with factor analysis and SEM","sourceUrl":"https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0359549","evidenceClass":"academic-research","outcomeClass":"cautionary","topics":["governance-procurement","knowledge-work"],"finding":"A cross-sectional self-report survey of 863 undergraduates at one Vietnamese university of science and technology found a factor combining ethical ambiguity, technological affordances and institutional gaps was positively associated with self-reported AI-assisted cheating (beta 0.228, p<0.001), while academic pressure and peer comparison showed a small negative association (beta -0.148, p=0.022). Male and non-STEM students reported higher scores; seniority was not significant. The model explained about 5% of variance. The authors recommend clear boundaries on acceptable AI use, explicit course-level expectations and consistent institutional guidance, and caution against reading the pressure result as protective.","sledRelevance":"Published 28 September and not yet archived, it is included now because it is independent scrutiny of the course-level rule variation surfaced this week at Notre Dame and Maastricht: unclear boundaries, not pressure alone, track with misuse.","evidence":"N=863, convenience sample, one institution; EFA, CFA and covariance-based SEM; R-squared about 5%; self-reported behavior.","architectureImplications":"Limited relevance; detection difficulty is one perceived affordance, but no tooling is evaluated.","governanceImplications":"Supports explicit permitted, disclosure-required and prohibited AI-use categories at course and institution level.","securityPrivacyImplications":"Limited relevance; self-report surveys on misconduct require anonymity protections.","caveats":"Cross-sectional and self-reported; convenience sample uneven by gender, field and seniority; single Vietnamese institution; modest explanatory power (about 5%); associations, not causal effects.","streamIds":["student-success"],"roles":{"sales":"The framing point is that unclear rules, not just student pressure, track with AI misuse. A PLOS One survey of 863 undergraduates linked ethical ambiguity and institutional guidance gaps to self-reported AI-assisted cheating. Use it to support policy-clarity and faculty-guidance work rather than detection purchases. Keep claims bounded: one Vietnamese institution, self-report, cross-sectional, and the model explained only about 5% of variance.","engineering":"For engineering, the framing point is making AI-use rules explicit and consistent in systems students touch. Assignment templates in the LMS can carry permitted, disclosure-required and prohibited AI-use labels. Proof of value should check whether students can correctly identify the rule for a given assignment before and after labeling, because this study measures perceptions and associations, not the effect of any intervention.","delivery":"Delivery's framing point is a shared AI-use taxonomy across courses. Coordinate integrity office and teaching center guidance, give faculty templates for course-level expectations, and survey students anonymously on rule clarity. Interpretation: the evidence supports clarity as a reasonable priority, but given modest explanatory power and a single non-U.S. sample, measure local results rather than assuming reductions in misconduct."},"roleDetails":{"sales":{"customerProblem":"Institutions face rising AI-related integrity concerns with inconsistent course rules.","stakeholders":["Academic integrity office","Teaching and learning center","Provost","Faculty senate"],"discoveryQuestions":["Do courses label which AI uses are permitted, disclosed or prohibited?","How do students learn the rules for each assignment?","Is integrity strategy weighted toward detection or clarity?"],"valueHypothesis":"Clearer AI-use boundaries may reduce ambiguity-linked misuse; the effect size is unknown.","potentialEngagement":"A bounded AI-use taxonomy and syllabus-template project with a student clarity survey.","unsupportedClaims":["That clear policies reduce cheating by a given amount","That academic pressure prevents cheating","That detection tools solve the problem"]},"engineering":{"fit":"Institutions standardizing assignment-level AI-use signals in the LMS.","architectureIntegration":"Assignment-level AI-use labels in LMS templates and syllabus builders.","prerequisites":["Agreed AI-use taxonomy","LMS template control","Faculty adoption"],"constraints":["Single Vietnamese institution","Self-report cross-sectional data","Low explained variance"],"security":"Run student clarity and misconduct surveys anonymously.","proofOfValue":["Student rule-identification accuracy before and after labeling","Share of assignments labeled","Track integrity referrals with appropriate caution"]},"delivery":{"work":"Define a shared AI-use taxonomy, deploy templates, and survey rule clarity.","dependencies":["Integrity office","Teaching center","Faculty senate approval"],"ownership":"Academic integrity office with teaching center.","skillsAdoption":"Faculty learn to state AI expectations per assignment.","governanceCheckpoints":["Senate policy approval","Survey ethics review","Term-end review of referrals and clarity"],"acceptanceCriteria":["Most pilot assignments carry an AI-use label","Clarity survey shows improved rule identification","Taxonomy published to students"],"risks":["Overinterpreting associations as causal","Faculty non-adoption","Labels conflict with discipline norms"]}},"retrievedAt":"2026-10-11T03:16:02Z","enrichedAt":"2026-10-11T03:18:59Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Interpretation: faculty need support to write consistent, explicit course AI expectations.","procurementImplications":"Limited relevance; does not support buying detection tools as a solution.","operatingModelImplications":"Integrity offices and teaching centers should coordinate a shared taxonomy of AI uses across courses.","sourceVerification":{"openedUrl":"https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0359549","referenceExcerpt":"The model accounted for approximately 5% of the variance in AICB, indicating modest explanatory power.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}