The establishment of stimulus control by instructions and by differential reinforcement.

Researchers compared three ways to teach adults a simple discrimination. One group got only differential reinforcement when they picked the correct picture. A second group got reinforcement plus brief instructions. A third group heard a clear rule that told them exactly which picture would pay off.

Everyone worked alone at a computer. The task never changed—just the teaching method did.

Reinforcement alone worked. After many trials all adults finally picked the right picture. Adding instructions cut the learning time in half. Giving a clear rule cut it even more. Rules won, instructions placed, reinforcement alone came last.

Badia et al. (1972) showed the same thing in an earlier lab: differential reinforcement by itself can build brand-new stimulus control. The 1990 study simply adds that you can speed things up if you also tell the learner what to do.

Peterson et al. (2024) pushed the idea further. They showed that the rule must be accurate; a wrong rule gives no control at all. Together the three papers make a timeline: reinforcement works, instructions help, accurate rules help most.

Raslear (1981) looks like a contradiction at first. That study got different response rates without any differential reinforcement—just louder or softer tones. The key difference is procedure. G kept reinforcement equal across tones and let intensity alone drive the response. S et al. kept stimuli equal and let the payoff schedule create the control. Both can be true; they test different levers.

When you run discrimination training start with a clear, accurate rule. If the learner cannot follow rules, add simple instructions plus immediate differential reinforcement. Save reinforcement-only trials for later checks, not for first teaching. You will save sessions and reduce client frustration.