Abstract
Presses on left and right buttons by undergraduate students occasionally produced points exchangeable for money, according to a multiple random-ratio randominterval schedule. During interruptions in the schedule, the subjects were required to fill out sentence-completion guess sheets about how they should press the buttons to gain the points available. Three of six subjects were instructed about the rules of button pressing (the instructed group), whereas the others received no instructions but their guesses were shaped with differential points also worth money (the shaped group). The schedule was changed so that button pressing relying on the instructed or shaped rules substantially decreased the available points in the contact condition but not in the no-contact condition. A schedule change in the no-contact condition produced no performance change in either groups. In the contact condition, shaped group subjects showed a performance change, whereas instructed group subjects did not until they temporarily encountered an extinction schedule. When the no-contact condition was reintroduced, sensitive responding occurred in the shaped group but not in the instructed group suggesting that prior experience in the contact condition increased the schedule sensitivity of the shaped group. The findings clearly demonstrated that responding was more sensitive to schedule changes when the rules were shaped than when they were instructed.
