Table 4 Two-sample t-test results comparing AIPatient and H-SP
Metric | Question | t-statistics |
|---|---|---|
Fidelity | ||
Role/Text Adherence | The SP followed the case script without contradictions. | 0.57 |
The SP’s responses matched the intended medical condition. | 1.77* | |
Contextual Appropriateness | The SP’s responses felt natural and relevant to my questions. | 1.10 |
Emotional Realism | The SP displayed believable emotions (e.g., pain, anxiety). | 3.02 *** |
Coherence/Consistency | The SP’s dialogue was coherent (no abrupt shifts). | 1.23 |
Response Quality | The SP’s answers were directly relevant to clinical questions. | 0.17 |
Usability | ||
Ease of Use | Interacting with this SP required minimal effort. | 1.62 |
I encountered no technical difficulties (e.g., delays). | 2.68*** | |
Feasibility/Scalability | This SP could be easily integrated into our training program. | 0.47 |
Effectiveness | ||
Diagnostic Accuracy | I have reached a preliminary diagnosis at the end. | 1.59 |
Learner Satisfaction | This session improved my clinical reasoning skills. | 2.19** |
