Puzzle-Box Learning

In another experiment involving drive, Edward L. Thorndike put a cat into
a puzzle box. The cat attempted to get out via various behaviors (mewing,
scratching, and so on). By chance, it stepped on a plate that resulted in the
door opening, allowing the cat to escape. The cat was repeatedly returned to
the box, and soon it would escape right away by stepping on the plate; other,
useless behaviors were no longer manifested. The source of D in this case
was the anxiety induced by confinement in the box, which could be measured
by various physiological changes, such as heart rate and hormonal levels.
Escaping would make the anxiety disappear; D is reduced. D reduction
results in an increase in the probability that the behavior immediately preceding
it (stepping on the plate) will recur. Thorndike describes this puzzlebox
learning as trial and error, implying a blind attempt at various means of
escape until one happens to work. He states that a “satisfying effect” will create
repetition, calling this the law of effect; the essence of the satisfying effect
appears to be drive reduction. A five-stage learning cycle is then complete: It consists of need, drive, behavior, drive reduction, and behavior
repetition.