Make assessments harder?

 One obvious solution would be to make the summative assessments more challenging, by redefining medicine’s culturally accepted notions of acceptable minimum competence for graduation to something more stringent. In this way, when students perceive no feedback gap between their performance and the required standard, they have at least already achieved a higher initial level of competence. However, this proposal is flawed on a number of levels. 

Firstly, it would be likely to lead to a higher initial number of students failing high-stakes assessments at the first attempt. This was one of the main problems with the assessment system in Osler’s time which led him to propose his changes; in his time failure rates of over 40% in the main high-stakes assessments were commonplace. Osler was clear on the distress this caused to students and was unsparing in his criticism of examiners who appeared to enjoy humiliating candidates. 

On a practical level and in the current economic climate, institutions would find it hard to support increased numbers of students through a remediation process. Irrespective of the logistical challenges, it is clear that medicine has much to do to improve the process of remediation. More fundamentally, making high-stakes decisions on single assessments is inherently flawed; some candidates will fail who should have passed, and vice versa. The process of defining minimal competence is fraught with difficulty, with a lack of clear agreement on hypothetical students who just about deserve to pass. As usual, Osler was here before us, acknowledging the difficulties in defining essential minimum requirements. 

Perhaps the most important flaw with this proposed solution is that it resides within the existing punitive behaviourist paradigm of existing high-stakes assessments; students are rewarded for passing and punished for failing. Raising the bar on an individual assessment would simply make the stakes even higher, so the emotional barriers to reception of feedback would be likely to be amplified. Instead, solutions which are situated within a constructivist approach to learning and assessment would seem more appropriate.

To explore other options to redesign assessments, please click below.