Effects of celeration rate on behavioral fluency

In precision teaching, it has been argued that learning should continue until responding is fluent (i.e., it can be performed accurately and rapidly in the presence of distracting stimuli, even after a delay since its acquisition). Furthermore, it has been argued that the speed with which learners achieve fluency (i.e., celeration) also is important. The purpose of the present experiments was to test those propositions. Subjects read binary numbers on a computer screen, and spoke the decimal equivalent of those numbers into voice-recognition equipment. In Experiment 1, responding was reinforced with money arranged on VI schedules for correct responses with inter-response times within a programmed range. This procedure allowed response rate and celeration rate to be controlled precisely while maintaining a similar reinforcement rate across conditions. Each group of subjects received one of the following terminal-rate conditions: 25, 30, 35, and 40 responses per minute. The terminal rate of 30 responses per minute was the most effective and efficient. In Experiment 2, each group of subjects received a condition in which the target rate increased by either 100% or 25% every 20 sessions until responding reached 30 responses per minute. In precision teaching, 100% increases after 20 sessions are considered optimal, and 25% increases are considered minimal. No between-groups differences were found on fluency tests. Latency of vocal responses decreased, but other topographical measures did not change. Present results suggest that terminal response is an important correlate of learning, but celeration rate is not