On the functional form between behavior and reinforcement on ratio schedules

In daily sessions, each of five pigeons pecked an illuminated disk in an experimental chamber in order to produce enough grain presentations according to a Fixed-or Variable-Ratio schedule to maintain its body weight at 80% of its free-feeding level. Each of two schedule types (Fixed- and Variable-Ratio) was factorially combined with seven ratio sizes (4, 8, 16, 32, 64, 128, and 256) to create 14 experimental conditions. Each condition lasted 15 days or until responding was judged stable, whichever was longer. When response rate (responses divided by session time excluding hopper duration) was plotted as a function of ratio size, response rates increased over the ratio sizes 4 to 16 and decreased thereafter. When running rate (total responses minus the first post-reinforcement response divided by session time minus the sum of the post-reinforcement pauses and hopper durations) served as the dependent variable, the resulting function decreased with increases in ratio size. This difference in functional form between response-rate and running-rate measures was attributed to the fact that the post-reinforcement pause occupies a greater proportion of session time at the smallest ratio sizes even though the post-reinforcement pauses were typically of one second's duration--an interval approximating the probable minimum transit time for the pigeon to move its head from the grain hopper to the pecking disk. Thus, the upward limb generated by ratio schedules over ratio sizes ranging from 4 to 16 may be an epiphenomenon of hopper-to-disk transit time.