Some Observations on Study Design

Quick plausible "answers" in sociology and social psychology are rewarded in our culture; tedious, modest experimental design is not in demand, and hence our discipline is not cumulative. In study design the ideal model is that of a controlled experimental, even if only a fraction of it. Since full experimental desing is very expensive and not always possible, those problems should be selected whose answers are worth the cost. This requires both theory which leads to operational deductions and preliminary fumbling research, whose intrinsic wastefulness can be reduced, if the number of variables is kept down to manageable limits and if such variables as are use are unidimensional.