TESTING - April 1, 2010

Judging teaching or learning by set examinations is a terrible way of learning what has been transmitted in the classroom, and everything else is worse. The mantra of accountability dictates that we need to know what students actually learn, and the result of a brief test, particularly one for which the students are deliberately crammed, is a poor predictor of how much remains months or years later if the need for it arises. Yet the mechanics of preparing the learner is an ancient practice, and often is a substitute for the kind of internalization of knowledge that is the desired result of the exercise. This applies most especially where the depth of the teacher’s knowledge does not go beyond the surface material that is the most usually the content of the test. The Continental approach, in which the examined student must reach the memorized material by an intuitive and searching extraction of the subtle material to realize how the memorized process is applied to the instant question is generally criticized for demanding insight, which is generally not teachable. A very simple example of this issue is found in what is known by various names, including “word problems” or “story problems”, in which the students is required to puzzle out how the questions are related to the processes of arithmetic. One approach is to have the same insight embedded in several different circumstances, and the true test of understanding is in ferreting out the arithmetic question from the story, if that is one in a new configuration, as where the problem of filling a swimming pool is seen to be essentially the same as the digging of a ditch or the rate of accumulation of a resource. These can be basically the same in form, but if the frame is not a practiced one, most examinees are found to be at a loss. It is there that the body is really buried. But if that form is not familiar drill, the heart of the learning is usually complained of as not having been taught that way.