Abstract
When reinforcement schedules demand that rats depress a lever for a minimum period of time, most lever presses will meet reinforcer requirements, but others will be much shorter. This results in a bimodal distribution of lever-press durations, with one peak near the reinforced duration value, and a smaller peak at less than 1 s. We conducted an experimental and descriptive analysis of short-duration presses in rats responding under a schedule that delivered edible reinforcers for 10 s of lever depression. All rats emitted biting and idiosyncratic behavior that may have both added and subtracted to the downward force necessary to maintain lever depression for extended periods. Movement of the response levers due to vigorous biting and sniffing, as well as premature hopper entries were both responsible for response durations that fell short of reinforcer requirements. Maintenance of long lever-press durations during fixed-time schedules in two out of three rats suggested that timing failure was unlikely a factor underlying bimodal distributions of response durations.