William N. Schoenfeld

Adrenal Hypertrophy in the Rat as a Function of Probability and Frequency of Punishment

After training on a regular reinforcement procedure in which a lever-pressing response produced milk reward, each of 16 groups of water-deprived rats was exposed to a different concurrent contingency of regular reinforcement and random interval punishment for 20 experimental sessions. Under the intermittent punishment schedules used, the programmed probability and frequency of response-contingent electric shock were independently manipulated. Autopsies revealed (a) that changes in adrenal weights were about the same (independent of the amount of shock) over a wide range of punishment frequencies and (b) that the amount of adrenal hypertrophy was a decreasing function of shock probability over a range of probabilities from 0.0625 to 1.00.

DIFFERENTIATION OF RESPONSE DURATION

Milk reinforcement for 2 rats was made contingent on the duration of a panel press response with reinforcement delivered on panel release. Various response distributions and cumulative records were taken at a panel “hold” duration requirement of 8 sec. Similarities and differences between the present data and those obtained with conventional DRL schedules were noted and are discussed in terms of response chaining.

Generalization gradients from “reaction time” or latencies of the white rat to visual brightness

Presentations of a light stimulus SD (milk-reinforced), interspersed with varying SΔ light stimuli (not reinforced), yielded generalization gradients of response latency (“reaction time”) for 2 albino rats. Gradients also appeared when “failure to respond” was used as the behavioral measure. The latter type of gradient changed with continued training, first in the direction of indicating increased discriminative sensitivity, and then toward less sensitivity.

A comparison of cumulating and non-cumulating time out for escape and avoidance behavior under a temporally defined schedule of negative reinforcement

Escape and avoidance behavior of two rats under a temporally defined schedule of negative reinforcement was compared for the cases of “non-cumulating Time Out” and “cumulating Time Out.” In the former case, a response in TO (Time Out) has no effect on the duration of TO; in the latter, a response in TO extends TO, and the duration of TO is measured from the last response in it. The findings suggest that the cumulating TO contingency does not, given the parametric conditions of this experiment, add to the efficiency of “avoidance” responding.