But often we can not do an adequate (Sufficient for the application) model. These cases arise when we want to give very accurate answers to difficult questions. While our model does not help us in the decision or gives a wrong prediction, it is a sure sign of its inadequacy. And in this case it is necessary to improve it. At this point, and helps to science, because science is a social experience of previous generations, which contains a set of models in the form of theories (the name says it not true, but only a guess). And the main thing here is not a complete return to one theory, but rather study a few. This will broaden your understanding of important aspects of the simulation. In this case you have formed a new model, which probably is a combination of previous and studied, perhaps it will exist in your mind away from the old model.
In any case, the new model is closer to reality, and its subjectivity approaching objectivity. And although we understand that will never achieve absolute objectivity, we can still improve the understanding of the world, which would entail the important knowledge, the correct prediction, and as a consequence competent management. It should also be noted that the basis of the classification model is adequate. Adequacy refers to the possibility to solve tasks with the help of this model. It is logical to assume that the more more tasks, the more difficult model, and hence for these tasks need to know, but do not necessarily know it to smaller ones.