Abstract
Background
The Hong Kong Social Welfare Department’s Standardised Care Need Assessment Mechanism for Elderly Services (SCNAMES) relies on comprehensive, multi-domain assessment instruments with over 300 items, spanning cognition, mood, psychosocial issues, nutrition, and physical functioning.
Amid growing service needs, Comprehensive Assessment for Psychogeriatric Care—a compulsory course in the Master of Social Science (Gerontology) programme—aims to train future assessors.
In traditional classroom learning, not all students can practise the geriatric assessments process with peers due to time limit. Only a few students could participate in role play while the majority learn through observation, leaving practical skills training under-challenging.
Description of / Practice
To bridge this gap, we developed a generative-AI simulation that role-plays an older adult based on a paper case, incorporating functional, cognitive, and psychosocial profiles. Students act as assessors—probing, clarifying, and scoring—while the chatbot responds with coherent, case-appropriate detail across scenarios of varying complexity: pain, activities-of-daily-living decline, fall risk, and caregiver strain.
The simulation enables universal participation and personalised feedback derived from AI-generated transcripts, which traditional classrooms lack. It explicitly supports self-directed learning and learner autonomy: students set goals, choose personas and difficulty, control pacing and repetition, and use structured prompts, item-specific checklists, and on-demand hints as scaffolds that fade as competence grows. Transcripts, scoring rationales, and analytics facilitate self-assessment and reflection.
Shifting from teacher-centred to student-centred learning increases autonomy and responsibility by repositioning instructors as coaches who provide targeted feedback on rapport, cultural-sensitivity, and scoring, convene concise debriefs, and set mastery thresholds rather than directing each interaction.
This iterative cycle operationalises Kolb’s experiential learning theory, preserves psychological safety by allowing mistakes without harming clients, and advances equity through multilingual, culturally-grounded cases that also support interprofessional learning among nursing, occupational therapy, and physiotherapy students.
Running on free AI platforms, the tool offers a low-cost pathway to build competence with standardised instruments before practicum or SCNAMES assessments.
Evidence of Outcomes
The tool will pilot in Fall 2025. A mixed-methods evaluation will analyse transcripts, student reflections, and relate indicators to confidence ratings and rubric scores to demonstrate learning gains and usability. Aligning gerontology education with government requirements while overcoming practical constraints, this AI simulation offers a scalable and equitable approach to achieve all-inclusive experiential learning.
Presenter(s) / Author(s)
