The wording of the title of this chapter is significant. We have used the word 'effectiveness' in preference to 'efficiency' in identifying the purpose of evaluation. The distinction between these two terms has been elucidated by many people in many ways, some of them rather technical, yet it continues to be a source of confusion. Since our concern here is with learning, own usage draws on learning theory, specifically that of Argyris and Schön (1996), which is represented in chapter 4 of the companion reader. Their work makes the following, crucial distinction:
By single-loop learning we mean instrumental learning that changes strategies of action or assumptions underlying strategies in ways that leave the values of a theory of action unchanged. For example, quality control inspectors who identify a defective product may convey that information to production engineers, who, in turn, may change product specifications and production methods to correct the defect … By double-loop learning we mean learning that results in a change in the values of the theory-in-use, as well as in its strategies and assumptions.
(Argyris and Schön, 1996:20-1, original emphasis)
Single-loop learning improves efficiency. It results in improvements to the processes through which we seek to achieve what we know we want to achieve. This might, sometimes, fit comfortably with the economist's definition, according to which efficiency is improved if either a given output can be produced with fewer inputs, or if with constant inputs more can be produced. On other occasions it might suggest something which is (to many people's minds, anyway) rather more qualitative. For example, we might argue that a given process of institutional change is more efficient, other things been equal, the less resentment it causes among employees and/or the more opportunities it provides for personal career growth. In neither case, however, is there any fundamental intention to make changes to underlying values and purposes.
Double-loop learning improves effectiveness. It forces learners to ask whether they are sure that what they think they want to do is what they really do want to do. As