The Design of Behavioral Research.

The conclusions of human behavior studies can not be believed, unless the investigator takes precaution to design his study so that confusion in data interpretation and alternative explanations of the conclusions are minimized. Without such precautions, the investigator's efforts may be largely ineffective. Experimental designs which are practical are usually not perfect. Sources of difficulty in experimental design can be attributed to setting up the hypotheses to be investigated, the sample of individuals tested, the measures to be used in the experiment, and the controls used in administering these measures. Even before the researcher collects his data, his conception of the question which he wishes to answer through the experimental process can significantly determine the result. An important source of difficulty in the design of behavioral experiments is the sampling of subjects. The evaluation of behavioral data requires no more care than the evaluation of any other kind of data obtained as part of the scientific processes, but certainly no less. Consideration must be given to experimental controls.