From the Student Success edition of October 21, 2026
THE Campus: Sophia University action research finds first-year EMI students skip AI claim-checking and reflection under deadline pressure; author proposes the 5P framework with in-assignment checkpoints
Sophia University · Higher education AI literacy and assessment design · Tokyo, Japan
- Publisher
- Times Higher Education (THE Campus)
- Original publication
- Times Higher Education Campus, 19 October 2026
- Source retrieved
- 2026-10-22
What happened
Chris Harwood, associate professor of education at Sophia University, describes the 5P framework (plan, probe, pause, prove, ponder) he developed for first-year undergraduates in an English-medium instruction programme. In his classroom-based action research, reflections from high-proficiency students suggested the framework increased awareness of AI's influence on their thinking, but students reported using the prove (check claims against independent sources) and ponder (reflect) stages less consistently under pressure, such as when several assignments were due. He cites a 2026 meta-analysis of 35 studies noting that better performance with AI assistance does not necessarily mean students can regulate learning independently, and recommends two brief in-assignment checkpoints: a source link with a sentence on what the check established, and a short reflection on AI's influence.
Why it matters
Directly relevant to AI literacy programmes: knowing how to question AI is not the same as doing it when workload peaks. Student-success and teaching centres can build verification and reflection into assignment structure rather than relying on one-off training.
Evidence and measured results
Author-reported qualitative action research with high-proficiency EMI students; no sample size or measures given.
Limitations and uncertainty
Single-author practitioner piece; action research with unspecified sample of high-proficiency students in one programme in Japan; no quantitative results; not peer reviewed. Transfer to U.S. general-education courses is uncertain.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-10-22; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Students trained in critical AI use skip verification and reflection when workload peaks.
- Stakeholders
- Teaching and learning centre director
- Gateway course instructors
- AI literacy programme lead
- Academic integrity office
- Students
- Discovery questions
- Where is AI literacy taught today, and is it assessed?
- Which assignments coincide with peak workload?
- Can LMS templates carry structured checkpoints?
- How would instructors absorb extra marking?
- Value hypothesis
- Interpretation: embedding brief checkpoints in assignments could sustain verification habits under pressure, if workload stays modest.
- Potential engagement
- A bounded course-design pilot: checkpoint templates in a few gateway courses with completion and quality measures over one term.
- Unsupported claims
- Improved critical thinking
- Reduced misconduct
- Transfer to all student groups
- Quantified effect sizes
Pre-sales engineering
Role takeaway
- Fit
- Fits institutions with LMS assignment templates and an AI literacy programme.
- Architecture and integration
- LMS assignment and rubric configuration; no AI system integration required.
- Prerequisites
- LMS structured submission fields or templates
- Rubric rows for checkpoints
- Instructor guidance
- Worked examples
- Constraints
- Practitioner evidence only
- Japan EMI context
- High-proficiency students only
- No quantitative data
- Security
- Keep reflection text out of default AI-detection workflows and limit access to instructors.
- Proof of value
- Checkpoint completion rates in peak versus early weeks
- Quality rubric scores for verification notes
- Instructor time per checkpoint
- Student feedback on burden
Delivery
Role takeaway
- Work
- Pilot the two in-assignment checkpoints in selected courses and measure use under pressure.
- Dependencies
- Instructor participation
- LMS template support
- Teaching centre facilitation
- Rubric design
- Ownership
- Teaching and learning centre with participating instructors.
- Skills and adoption
- Instructor modelling of the five stages with a worked example; student practice before independent use.
- Governance checkpoints
- Course selection and instructor consent
- Mid-term completion review
- End-of-term evaluation
- Decision on scaling
- Acceptance criteria
- Proposed: checkpoint completion at least 80% in peak weeks
- Proposed: completion gap between peak and early weeks under 10 points
- Proposed: instructor marking time under five minutes per submission
- Proposed: student burden rated acceptable by most respondents
- Risks
- Checkpoints become box-ticking
- Added marking load
- Uneven effect across proficiency levels
- Weak transfer to other courses
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Limited technical relevance. Interpretation: LMS assignment templates could carry the two checkpoints as required fields.
Governance
Who approves, reviews and stays accountable for outcomes?
Suggests AI-use expectations should be embedded in assignment instructions and checkpoints, not only in policy statements.
Security and privacy
What data, permissions and controls need testing?
Not applicable to this source.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Checkpoints add modest student and instructor workload; effects on lower-proficiency or second-language students are not reported.
Procurement
What should contracts, pricing and exit terms secure?
None directly.
Operating model
Which teams own the service once it runs?
Places responsibility on course design and instructors, with whole-class feedback on recurring difficulties.
Publication history
- 2026-10-21Student Success · Issue 484 resources
Stable resource ID: the-campus-sophia-university-5p-framework-ai-checks-skipped-under-deadline-pressure-20261019