Christian Schunn and David Klahr
Our goal is to produce a unified model of processing in all four problem spaces. This unified model will consist of separate components corresponding to processing in each of the four problem spaces. However, as indicated in Figure 1, processing within each space is dependent upon the current search state in the other spaces. For example, experiment space search depends upon the available experimental paradigms as well as the current hypothesis. Given these strong interdependencies, there is great advantage to implementing each of the components in a unified model.