Toggle Main Menu Toggle Search

Open Access padlockePrints

Comparison of Cohen and Angoff methods of standard setting: is Angoff worth it?

Lookup NU author(s): Dr Laura Woodhouse, Professor Brian Lunn, Professor David Kennedy


Full text for this publication is not currently held within this repository. Alternative links are provided below where available.


Standard setting is an essential process in definingcompetency in medical education. However, currentlythere is no gold standard method of standard setting.One commonly used method is Angoff in which apanel of experts estimate the percentage of borderlinestudents predicted to correctly answer each question inan examination. This method is costly, time consumingand relies on the assumption that the panel can accuratelydefine the borderline student. Recently the Cohen methodhas been developed and subsequently modified, toovercome these disadvantages. Our aim was to comparethe standard set using our current method of Angoff toCohen and modified Cohen methods, to inform futurestandard setting practices. Cohen and modified Cohenmethods were applied to historical data for writtenexaminations across all 5 years of the MBBS programmeat Newcastle University. Data included cohort sizes of250-470 students per year, from academic year 2011/12onwards. For single best answer (SBA) examinations inyears 1 and 2, the Cohen method produced consistentlyhigher pass marks and failure rates compared to Angoff.However, the modified Cohen method produced passmarks and failure rates comparable to Angoff. For year3 SBA examinations, modified Cohen also producedcomparable pass rates and failure rates to the Angoff. Foryears 4 and 5 SBA examinations the Cohen and modifiedCohen methods produced comparable pass marks whichwere marginally higher than that determined by Angoff.This would have led to a 1-5% increase in the failure rate.With modified Cohen producing comparable standardsto Angoff it suggests that this method may be a valid andeconomical alternative to standard set SBA examinations.We are currently analysing historical data for clinicalexaminations to determine whether standard setting usingthe Cohen or modified Cohen method is feasible forpractical as well as SBA examinations.

Publication metadata

Author(s): Woodhouse L, Lunn B, Kennedy DJ

Publication type: Conference Proceedings (inc. Abstract)

Publication status: Published

Conference Name: European Board of Medical Assessors Annual Academic Conference: Crossing Boundaries: Assessment in Medical Education

Year of Conference: 2016

Pages: 75-75

Acceptance date: 02/04/2016

Publisher: University of Exeter Medical School