ABSTRACT
Decision-makers and social scientists have a common interest in understanding the quantitative dimensions of metropolitan development and its functional processes. This identity of interest is expressed by the fact that each, from his own point of view, is interested in conditional predictions regarding function and development. The scientist makes use of conditional prediction as a method of testing theories. The decision-maker or the planner uses conditional predictions in a much more practical and immediate sense. He is interested in evaluating the putative consequences of innovations and changes in policy designed to affect urban processes. These evaluative forecasts are conditional predictions in identically the same sense as those made by the social scientist for the testing of theories. The difference between the approaches in these two different contexts arises not out of the formal content of the methods, but out of the selection of variables and of measurements of consequences, that is, the selection of inputs to and outputs from the predictions, which are made by means of these experiments.
