Date of Award
Doctor of Philosophy
W. O. Jenkins
Edward E. Cureton, Samuel R. Tipton, Ernest Furchtgott
[From the Introduction]
In the usual psychological experiment certain operations are performed upon the organisms being studied and a certain portion of the total responses are measured. The operations performed upon the organism may be roughly divided into two classes, those which are systematically varied and define the various experi mental groups, and those which are held constant across groups. Certain lawful relationships are then determined between those operations which are varied and the responses measured. Those operations which are held constant are considered to be factors which may also affect the responses being measured. It is usually felt that if these are held constant, then they will affect all groups equally and will not contaminate the results. The assumption here is that what is constant for the experimenter is constant for the various organisms. For example, in a T-maze problem, if the effect of several levels of hunger are being studied, the same maze is used with all groups, and it is assumed that the maze dimensions are constant for all groups. It may well be, however, that a given maze dimension may differentially affect animals at different hunger levels. If, at each hunger level, we use two mazes of differing lengths and find that the results obtained from one maze are parallel to the results obtained from the other, then we may assume that a condition of constancy exists. In some cases it will be obvious that such a condition of constancy does exist in the experiment, but in others this condition should be tested before the inferences are generalizable.
It is to be noticed that the condition of constancy is defined by an a priori choice of the response of the organism to be measured. It may well be that if one does get results which appear to support a condition of constancy, it would not have been obtained had another response of the organism been measured. Therefore, one must use care in choosing which response of the organism he is going to measure.
In the experiments to be reported on in this paper, close attention was paid to the effect of certain of the operations which in previous experiments have been considered under the class of "constant operations." In the activity experiment, attention was focused on the effect of the apparatus used in measuring the activity, and in the T-Maze experiment, attention was focused upon a certain training procedure that has been used in many learning experiments in the past.
Strong, Paschal Neilson Jr., "Activity and T-maze Performance of the White Rat as a Function of Drive and Apparatus. " PhD diss., University of Tennessee, 1955.