evidoria

← Back to browse

Good practice Imported

ChatGPT-3.5 as a Supplementary Skills Instructor Improves Dental Procedure Performance in a 187-Student RCT at Wuhan University

China · Wuhan · See the China profile · See the Wuhan profile

Evidence: Randomised controlled trial Top 56% 47/100 · Ask Evidence Copilot about this practice

A one-week RCT with 187 Chinese dental undergraduates found ChatGPT-3.5 access alongside standard video instruction significantly improved procedural performance and cut eye-tracking-measured cognitive load, with the biggest gains for students with lower spatial ability.

187 students
Students randomized (2024)
73.12
Practical assessment score, ChatGPT group
65.54
Practical assessment score, control group
0.137
Cognitive load (pupil diameter), ChatGPT group
0.312
Cognitive load (pupil diameter), control group
70.20
Score, low-spatial-ability students, ChatGPT group
55.41
Score, low-spatial-ability students, control group

Details

Maturity
Pilot
Promoter
Wuhan University, School & Hospital of Stomatology
Period
2024
Keywords
higher education, dental/medical training, clinical skills education

Context

Wuhan University's School & Hospital of Stomatology randomly assigned 187 third-year dental undergraduates to a one-week skills-training unit, either with standard video instruction alone (control) or with the same video instruction plus on-demand access to ChatGPT-3.5 as a supplementary instructor (intervention).

Objectives

Test whether on-demand access to a general-purpose AI chatbot alongside standard video instruction improves procedural performance and reduces cognitive load, particularly for students with lower spatial ability.

Activities

Students completed a one-week dental skills-training unit with standard video instruction, with the intervention group additionally given on-demand ChatGPT-3.5 access as a supplementary instructor. Outcomes were measured with a blinded practical assessment, eye-tracking-derived pupil-diameter measures of cognitive load, and self-reported self-efficacy and motivation.

Results

The ChatGPT group scored significantly higher on the blinded practical assessment (mean 73.12 vs 65.54, t=4.569, p<.001) and showed significantly lower cognitive load (mean pupil-diameter measure 0.137 vs 0.312, p<.001). Students with low spatial ability benefited most (mean 70.20 vs 55.41, p<.001), and self-reported self-efficacy (p=.04) and learning motivation (p=.02) both favoured the ChatGPT group.

Conclusions

The trial ran for one week at a single university using an untuned, general-purpose chatbot rather than a purpose-built or institutionally vetted AI tutor; the published study does not report data-governance measures for routing student data through a third-party, US-hosted chatbot, nor does it evaluate any teacher-facing component.

Implementation
Indicative cost
Low (< €50k) — Used an off-the-shelf, consumer ChatGPT-3.5 subscription/access alongside existing video materials; no purpose-built tool or major additional cost reported.
Time to results
Short (< 1 year) — A one-week intervention embedded within an existing skills-training unit.
Staffing & skills
Wuhan University School & Hospital of Stomatology research/teaching team

Conditions for success

  • On-demand access to a general-purpose AI chatbot (ChatGPT-3.5) alongside existing standard video instruction
  • Blinded practical assessment and eye-tracking measurement for objective outcome data

Common failure modes

  • Untuned, general-purpose chatbot rather than a purpose-built or institutionally vetted AI tutor
  • No reported data-governance measures for routing student data through a third-party, US-hosted chatbot
  • Single-institution, one-week trial with no teacher-facing evaluation

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful