Usually, the scientist's model is only a partial representation of reality and the results which the scientist predicts through his understanding of how his theoretical model works sometimes are not borne out by his observations in real life. The scientists, for example, set up a model of how the universe worked. This placed the earth in the centre and made the sun circulate round it.

The model explained certain regularities they had observed in the movement of heavenly bodies but left other regularities unexplained. When Copernicus rearranged the model and placed the sun at the centre of the planetary system, the new model explained both the regularities which had been explained by the old model and the regularities it had not explained. The new model was thus more satisfactory since its heuristic or explanatory value was greater.Social scientists are trying to do exactly the same sort of thing. They examine some aspect of reality and try to isolate what appear to them to be the relevant enduring relationships between its different components. These relationships are then expressed in a logically consistent paradigm.

This paradigm is tested against reality, and the degree to which it can be used to predict reality gives some measure of its utility as a heuristic device. The social scientist prefers to work with an imperfect paradigm the logical construction of which he understands, rather than with one which may for in fact reproduce reality more accurately but of whose composition he is uncertain. It should emerge from this that the social scientist?s concept of causality differs considerably from the commonsense view. He does not see an effect B arising from here some known cause A but rather phenomena A, B, C, D, F, and so on back to A, all related to each other in particular ways, these relationships being such that a change in say E will modify the relationships of F to F and to D, hence F to G, and D to C, G to H, and C and so on.