The Reliability and Validity of a Paediatric Script Concordance Test with Medical Students, Paediatric Residents and Experienced Paediatricians

Abstract

Background: The Script Concordance (SC) approach was used as an alternative test format to measure the presence of knowledge organization reflective in one’s clinical reasoning skills (i.e., diagnostic, investigation and treatment knowledge).

Methods:  The present study investigated the reliability and validity of a 40-item paediatric version of the SC test with three groups representing 53 medical students (novices), 42 paediatric residents (intermediates) and 11 paediatricians (experts).

Results:  A comparison between scoring techniques based on experts’ ratings of the items showed internal reliability coefficients from .74 for the one-best answer up to .78 for alternative scoring techniques.  An ANOVA showed an increase in test performance from medical students through to expert paediatricians (F_(2,103) = 84.05, p < .001), but did not differentiate between the postgraduate year 1 to 3 paediatric residents.  A large effect size (Cohen’s d) difference of 1.06 was found between medical students and residents total SC test scores.

Conclusions:  These results support other findings indicating the SC test format can be used to differentiate between the clinical reasoning skills of novices, intermediates and experts in paediatrics.  An alternative scoring method that includes one best answer and partial marks was also supported for grading SC test items.