Reinforcement schedules, intermittent

However, reinforcement does not necessarily happen all the time or even regularly. For example, even though many of us work every day, we do not necessarily get paid at the end of that day for what we did. And in some instances reinforcement is unpredictable, like when you receive an unexpected phone call that is rewarding to you from a close friend. When reinforcement doesn t occur in a predictable way, it is referred to as being on a variable or intermittent (random or unpredictable) schedule (or pattern). Behavioral researchers have found that a variable reinforcement schedule produces behavior patterns that are much more difficult to change than behavior patterns reinforced regularly. 

Gray T, Wise RA (1980) Effects of pimozide on lever pressing behavior maintained on an intermittent reinforcement schedule. Pharmacol Biochem Behav 72 931-935. 

Notice that the "worker" in this picture does not receive a reward for every cast. In fact, he is on an intermittent reinforcement schedule. He catches a fish once in a while. This kind... 

In the real world, consequence stimuli do not necessarily follow every occurrence of the response. In fact, typically, consequences follow the response on an intermittent basis. Paychecks, for example, are typically distributed on a weekly, biweekly, or even monthly basis, not after each instance of work-related activity that occurs. The pianist plays the entire piece of music before the audience applauds. This strategy of intermittent reinforcement of responding actually provides greater behavioral efficiency and economy as well as greater response strength and persistence than does continuous reinforcement. A response that has been reinforced after every occurrence declines much more rapidly when reinforcement is withheld (extinction) than does one that has been reinforced on an intermittent schedule. 

Performance generated by intermittent schedules of reinforcement has played an important role in behavioral pharmacology and is proving a useful tool in behavioral toxicology. On an intermittent schedule, an animal is not reinforced for every response but for a number of responses according to certain rules . Most intermittent schedules are based on reinforcing the organism as a function of the number of responses emitted, some temporal requirement for emission of responses, or a combination of these. For example, a fixed ratio (FR)... 

Intermittent schedules may also be maintained by negative reinforcement, usually by a brief mild electric shock. The most popular of these is continuous or Sidman avoidance in which each response postpones a shock by a fixed amount of time. By spacing its successive responses within this time interval, the animal may postpone shock indefinitely. This schedule is particularly useful as a comparison to behavior generated by positive reinforcement if a toxicant is suspected of producing anorexia. Simple intermittent schedules such as these have been used fairly widely in behavioral toxicology and have proved to be sensitive to the effects of a number of industrial and environmental toxicants. 

Intermittent schedules of reinforcement can be combined to form more complicated schedules such as multiple schedules of reinforcement. For example, if FR and FI schedules are presented to an animal in succession during a single test session, the resulting multiple schedule is termed a multiple FR-FI schedule. Each component of the multiple schedule is independent and occurs in the presence of a different external discrimination stimulus that signals the schedule component in effect. Schedule components are typically presented in an alternating fashion, first one schedule and then the other this allows the investigator to collect data on both types of behavior almost simultaneously. This schedule in particular has proved to be useful in detecting behavioral toxicity. 

Following FI schedule testing, each monkey was tested on another intermittent schedule, the DRL (differential reinforcement of low rate), which assessed the monkey s ability to inhibit responding. This required the monkey to wait at least 30 s before responding in order to receive a reward. Although the lead-treated monkeys were able to learn the task, they did so at a slower rate than controls and were more variable in their performance from day to day than the controls (similar to the FI results). 

A variety of operant conditioning techniques have been used to assess nervous system damage produced by exposure to lead in the monkey. Several classes of tests are proving to be sensitive in detecting behavioural impairment produced by toxicants (Laties, 1978), tests that are drawn largely from the fields of behavioural pharmacology, experimental psychology, and the neurosciences. Intermittent schedules of reinforcement offer a powerful tool... 

---