Search

Alternation and repetition of both stimulus and response components were investigated in a multiple-choice situation, using a 2 × 2 factorial design (12 Ss per cell) under 0 and 100% reinforcement conditions and with 20-sec. and 20-min. intertrial intervals. The results replicate those of a previous study in that nonreinforced Ss exhibit little change in behavior over the 12 days while reinforced Ss show an increase in response repetition and a decrease in response alternation. In addition, it was found that in every case there is more repetition and less alternation of responses with distribution of trials. It appears that the stimulus component plays a minor role in this situation and also that the reinforcement variable exerts a more potent influence than the distribution variable.

METHOD Subjt-cis and apparatus.--One week after being placed on a 22-hr. food deprivation schedule ( 8 gr daily), 160 naive male albino rats were trained on a straight runway which was 24 In long, 5 in. wide, 6 in. high, covered with glass, and divided by a vertically sliding door between the starting box and reaction compartment. A 2V8-in. continuation of the runway contained the stimulus. The stimulus was a rectangle of sheet metal the size of the end of the alley and presented a flat black background upon which was painted a white circle (plane). Flush with and cut in the center of the circle was a round door 1 % in, in diameter and hinged at the top. This door could be pushed open by S's nose to secure a food pellet located in a small tray just below and behind the stimulus door. Timing was accomplished by means of two silent Hunter Klockounters activated by raising the starting box door 1% in. O n each trial two measures were obtained: ( a ) latency, time elapsing from the raising of the starting box door until S broke the light beam located 6y8 in. beyond the door; and ( b ) response time, time elapsing from the raising of the starting box door until S pushed the stimulus door % in. The measures were converted to reciprocals and referred to as starting speed and response speed, respectively. Following three days of preliminary training during which Ss learned to open the stimulus door and secure the incentive, Ss were given 1 0 reinforced trials on each of five successive days. Half of the Ss were rewarded for traversing the runway and opening the door in the center of a circle 20 sq. cm. in area, the other half for traversing the runway and opening the door in the center of a circle 79 sq. cm. in. area. After each trial S was rerurned to his home cage and given his next trial after a minimum of 2 0 min. had elapsed. RESULTS AND DISCUSSION The effects of stimulus size on reaction potential were determined by the following measures: ( a ) response speeds and starting speeds on the first five training trials, ( b ) response speeds and starting speeds on the last five training trials, and ( c ) acquisition training as a whole. For each S, the median response speed and median starting speed (l/sec.) on each block of five trials were obtained, and the mean-median group measure calculated. A comparison of the response speeds on the first five training trials for the two groups yielded a t of 1.40, while a comparison of the starting speed measures yielded a t of 0.92. A comparison of the response speeds on the last

in confounding stimulus generalization and the generalization of extinction effects. Grant and Schiller (1953) emphasized the importance of investigating the magnitude of the response during the first test trial after training in order to minimize the effeccs of discrimination training. The importance of measuring generalization of excitation in the early extinction runs in order to avoid interfering effects of inhibition was also recently discussed by Heyman (1957). The purpose of chis study was co repeat the investigation reported by Grice and Saltz, except for the utilization of distributed rather than massed acquisition trials and alteration in appararus, procedure, and measurements taken. METHOD Sabjec~s.-Ss were 160 experimentally naive male albino rats purchased from the Albino Farm at Red Bank, New Jersey. Their ages ranged from 70 to 100 days at the beginning of experimentation. Each S was assigned at random to one of the eight groups, subject to the restriction that each group contain 20 Ss. Diet during the experiment consisted of 8 gm. of Purina Dog Chow Checkers administered daily in individual cages after each experimental session. For a period of 1 wk. prior to the beginning of the experiment, Ss were adapted to a 24-hr. deprivation schedule, using 10 gm. of food.