[middle of page 9 of handout]
Sidman (1953) Two temporal parameters of the maintenance of avoidance behaviour in the white rat. Journal of Comparative and Physiological Psychology, 46, 253-261.
Rats were tested in a Skinner box. If no response was made, brief shocks were given at fixed intervals e.g. every 5 seconds. This was termed the 'S-S' interval. If the rats pressed a lever, this delayed the next shock for a standard period e.g. 10 seconds. This interval is termed the 'R-S' interval.
There was no external signal for the shocks.
Individual animals were tested with a variety of settings for the two intervals.
Results: Rate of responding varied systematically with the interval settings. The maximum rate of response for a given SS interval occurred when the RS interval was just slightly longer than the SS interval, and higher rates of response were found with shorter SS intervals.
[bottom of page 9 of handout]
Herrnstein and Hineline (1966) and Herrnstein (1969)
Again rats could reduce the frequency of brief shocks in a Skinner box by pressing a lever. However shocks occurred at random, whether or not the lever was pressed. For instance, rats learned to consistently press the lever if shocks occurred on average every 6 sec without lever presses, but on average 20 sec after a lever press.
There was no external signal for the shocks. Since there could be no internal timing signal either, the authors argue that the two-factor theory of avoidance learning cannot account for the results. Instead, Herrnstein (1969) argues that a reduction in the frequency of shocks is sufficient to reinforce the learned response, and there is no need to appeal to conditioned emotional effects.