Why Reinforcers Work

186
These curves illustrate the patterns of behavior typically seen under different reinforcement schedules. The steeper the
curve, the faster the response rate. The
thin diagonal lines crossing the curves
show when reinforcement was given. In
general, the rate of responding is higher
under ratio schedules than under interval
schedules.
Learning
1,000
Fixed
ratio
Number of responses
FIGURE 5 .1 0
Results of Four Partial
Reinforcement Schedules
Chapter 5
Variable
ratio
750
Fixed
interval
500
Variable
interval
250
10
20
30
40
50
60
70
80
Time in minutes
Source: Adapted from Skinner (1961).
for example, that you are in a gambling casino, standing near a broken candy
machine and a broken slot machine. You might deposit money in the broken candy
machine once, but this behavior will probably stop (extinguish) very quickly. The
candy machine should deliver its goodies on a continuous reinforcement schedule,
so you can easily tell that it is not going to provide a reinforcer. But you know that
slot machines give rewards on an unpredictable, intermittent schedule. So you might
put in coin after coin, unsure of whether the machine is broken or is simply not
paying off at the moment.
Partial reinforcement helps to explain why superstitious behavior is so resistant
to extinction (Vyse, 2000). Suppose you had been out for a run just before hearing
that you passed an important exam. The run did nothing to cause this outcome. The
reward followed it through sheer coincidence. Still, for some people, this kind of
accidental reinforcement can strengthen the behavior that appeared to “cause” good
news. These people might decide that it is “lucky” to go running after taking an
exam. Similarly, someone who wins the lottery or a sports bet while wearing a particular shirt may begin wearing the “lucky shirt” more often (Hendrick, 2003). Of
course, if the person wears the shirt often enough, something good is bound to follow every now and then, thus further strengthening the superstitious behavior on a
sparse partial schedule.
Why Reinforcers Work
What makes a reinforcer reinforcing? For primary reinforcers, at least, the reason could
be that they satisfy basic physiological needs for survival. Yet artificial sweeteners, which
have no nutritional value, can be as powerfully reinforcing as sugar, which is nutritious.
And what about addictive drugs, which are also powerful reinforcers, even though they
threaten the health of those who use them?
Research by biological psychologists suggests that reinforcers may work by exerting
particular effects on the brain. In a classic study on this point, James Olds and Peter
Milner (1954) discovered that mild electrical stimulation of certain areas of the brain’s
hypothalamus can be a powerful reinforcer. Hungry rats will ignore food if they can press
a lever that stimulates these “pleasure centers” (Olds, 1973). It has since been discovered