Authentic tasks in higher education: Studying design principles for assessment

Abstract

Students may benefit significantly from learning through authentic tasks. But how do we assess their learning outcomes, taking into account the specific characteristics of authentic tasks? In the second presentation of this symposium on design principles for authentic tasks we present and discuss principles for assessment procedures in authentic tasks from three different domains: clinical reasoning in Veterinary Science; internships in Medicine; and essay writing in History, all at Utrecht University in The Netherlands. Findings were evaluated from the point of view of the quality criteria for assessment procedures (validity, reliability, acceptance, cost-effectiveness, and impact on learning) and related to the design principles for authentic tasks. Approaching the assessment problem from the direction of authentic tasks in general implies the claim of high validity. At the same time it raises the suspicion of low reliability. Which assessment principles, therefore, will help us design robust assessment procedures in the context of learning through authentic tasks? Analyses from the three studies indicate that in designing assessment procedures for authentic learning tasks that they will produce results that might not be as reliable and reproducible as several other testing procedures. This should be counterbalanced by the clear advantage of more authentic testing: increased validity and positive impact on learning. Procedures should be affordable: this sets limits on what is possible in practice, and we should accept these limits. Lastly, it is important to verify that all stakeholders will accept the procedures and outcomes. Therefore, transparency and communication when implementing new procedures are fundamental.