The experiment will not give valid results unless the situational variables are controlled. If the experiment is conducted with different set of students who have been trained in a different manner of have a different value system then the results will be different in comparison to those who have been subjected to experiment earlier. Similarly the other conditions like the person who is supervising the place of supervision etc. have to be controlled. This means that all those conditions, which might affect the dependent variable, are toe controlled. Since in any experiment there are numerous conditions which are needed to be controlled it is many a time difficult to do so. Hence we take recourse to various types of experimental designs, which we have described earlier.