Multiple mini interviews: revealing similarities across institutions

Area of Science:

• Medical Education
• Psychometrics
• Healthcare Professional Selection

Background:

• Multiple mini interviews (MMIs) are increasingly used globally for selecting medical and health professionals.
• MMIs offer improved reliability and validity over traditional panel interviews.
• Limited research exists on whether MMIs and panel interviews measure the same underlying constructs.

Purpose of the Study:

• To determine if multiple mini interviews (MMIs) are unidimensional or multidimensional.
• To investigate if MMIs from different institutions assess similar constructs.
• To compare the constructs measured by MMIs versus panel interviews.

Main Methods:

• Analysis of interview scores from medical degree applicants at three institutions over two years (n=293).
• Two institutions utilized independently developed MMIs; one used a panel interview.
• Exploratory and confirmatory factor analyses were performed on station and overall scores.

Main Results:

• Factor analyses supported a unidimensional model for MMIs across both years.
• MMI scores were moderately to strongly correlated with each other (r=.56–.64).
• MMI scores showed weaker correlations with panel interview scores (r=.07–.48).

Conclusions:

• Both MMIs and panel interviews appear to measure a single latent construct.
• The constructs measured by MMIs and panel interviews are not the same.
• MMI methodology may be suitable for measuring emergent constructs like adaptability.