Designing software to maximize learning
I take it as axiomatic that those involved in developing educational software intend that it should maximize learning. The evaluation of educational software, therefore, needs to focus mainly on this central issue. There are still, however, two ways in which 'maximization' can be measured: either in terms of the increase in the amount of learning, or in terms of the increase in the quality of learning. In either case, an important contributory factor in the evaluation of educational software is the amount of take-up in terms of purchase and use. The crude indicator of BSH (number of Bums on Seats for number of Hours) is of some value, because if software is not used it cannot have any impact at all on learning. But it is an indicator which should only be used with caution because badly designed software used frequently presumably does less to maximize learning than well designed software used infrequently. I am concerned here with the design of software which maximizes the quality of learning. I will also deal briefly with the issue of sensitivity to the context of use, but the wider problem of the take-up and use of software, although important and interesting, is beyond the scope of this paper. The paper is based on a review of key texts in the literature on learning, including some which relate directly to software development. I hope it will make some contribution to both the design and the evaluation of educational software.