Avoidance acquisition in cats as a function of temporal and intensity factors.

Acquisition of an active bar-pressing avoidance response was investigated in a factorial design with 3, 5 or 9 s CS-US intervals and CSs of 70- or 50-dB of white noise. As the CS-US interval increased the required time to meet the avoidance criterion decreased. The shorter the CS-US interval and the more intense the CS, the shorter the median response latencies observed at the end of the training. The latency data were analyzed in Vincentized fifths of acquisition trials. At the beginning of training, when the highest proportion of responses occurred soon after the US onset, escape latencies were shorter in groups trained with shorter CS-US intervals and the more intense CS. During the 1st and 2nd fifths of training, avoidances occurred with the same probability in each part the CS-US interval. By the middle of the training, after a performance level of 50 percent avoidance responses was reached, a rapid increase in the number of short-latency avoidance was observed, which was more pronounced in groups trained with the 70 dB CS. While the general shape of the latency distributions was preserved during the remaining two fifths of the training, a further increase in the intensity effect was noted, which occurred along with a decrease in the CS-US interval effect. The data are in favour of Two-factor Theory of Avoidance learning and suggest that at various stages of training, different variables emerge for acquisition and consolidation of the avoidance reflex. Data supported the hypothesis that two different mechanisms are involved in performance of short- and long-latency avoidance responses.