Predictive validity of the multiple mini-interview for selecting medical trainees

Area of Science:

• Medical Education
• Assessment and Evaluation

Background:

• The multiple mini-interview (MMI) is a selection tool used in medical education.
• Its validity and the stability of candidate performance require further investigation.
• The relationship between cognitive and non-cognitive qualities in assessments needs clarification.

Purpose of the Study:

• To validate the multiple mini-interview (MMI) selection process by comparing MMI scores with national clinical skills examination results.
• To assess the stability of candidate performance across MMI administrations.
• To explore the independence of cognitive and non-cognitive qualities throughout medical training.

Main Methods:

• Predictive validity was examined by matching MMI data with licensing examination data for undergraduate and postgraduate samples.
• Reliability coefficients were generated to assess the stability of candidate performance.
• Correlations between cognitive and non-cognitive measures were analyzed at different training stages.

Main Results:

• The median reliability of the MMI was 0.73 with 12 stations and one examiner per station.
• A significant correlation (r = 0.43 and r = 0.35) was found between MMI performance and licensing examination outcomes.
• The correlation between cognitive and non-cognitive assessments increased with training duration.

Conclusions:

• The study provides further evidence for the validity of the MMI in admissions decisions.
• Findings challenge trait-based models for capturing performance.
• The distinction between cognitive and non-cognitive qualities is context-dependent on the scope of practice.