Jim Crossley1, Gerry Humphris, Brian Jolly. 1. Department of Paediatric Medicine, Sheffield Children's Hospital, Western Bank, Sheffield S10 2TH, UK. j.crossley@sheffield.ac.uk
Abstract
BACKGROUND: Good professional regulation depends on high quality procedures for assessing professional performance. Professional assessment can also have a powerful educational impact by providing transparent performance criteria and returning structured formative feedback. AIM: This paper sets out to define some of the fundamental principles of good assessment design. CONCLUSIONS: It is essential to clarify the purpose of the assessment in question because this drives every aspect of its design. The intended focus for the assessment should be defined as specifically as possible. The scope of situations over which the result is intended to generalize should be established. Blueprinting may help the test designer to select a representative sample of practice across all the relevant aspects of performance and may also be used to inform the selection of appropriate assessment methods. An appropriately designed pilot study enables the test designer to evaluate feasibility, acceptability, validity (with respect to the intended focus) and reliability (with respect to the intended scope of generalization).
BACKGROUND: Good professional regulation depends on high quality procedures for assessing professional performance. Professional assessment can also have a powerful educational impact by providing transparent performance criteria and returning structured formative feedback. AIM: This paper sets out to define some of the fundamental principles of good assessment design. CONCLUSIONS: It is essential to clarify the purpose of the assessment in question because this drives every aspect of its design. The intended focus for the assessment should be defined as specifically as possible. The scope of situations over which the result is intended to generalize should be established. Blueprinting may help the test designer to select a representative sample of practice across all the relevant aspects of performance and may also be used to inform the selection of appropriate assessment methods. An appropriately designed pilot study enables the test designer to evaluate feasibility, acceptability, validity (with respect to the intended focus) and reliability (with respect to the intended scope of generalization).
Authors: Thomas J Beckman; Amit K Ghosh; David A Cook; Patricia J Erwin; Jayawant N Mandrekar Journal: J Gen Intern Med Date: 2004-09 Impact factor: 5.128