Evaluation systems: How do they frame, generate and use evidence?

Abstract

How does the design of evaluation systems affect the different ways of using the results of evaluations? This article offers a conceptual model that outlines three ‘ideal’ types of evaluation systems. It is a heuristic tool for opening up the ‘black box’ of evaluation systems and assessing their qualitative differences in terms of types of ‘owners’ of evaluations, questions asked, methods deployed, answers provided and avenues for use of evaluative knowledge. We apply the model to study the case of the Lithuanian evaluation system. In contrast to the expectations of some of the previously developed models, it does use evaluation results, and we aim to understand why the generated evidence is more often used in some areas rather than others.