Theoretical Approaches and Questions in Operant Conditioning Psychology

Theoretical Approaches and Questions in Operant Conditioning Psychology 3306

Responses and reinforcers • How important are responses and reinforcers? • Thorndike and Skinner would say that it is essential • Tolman said ‘umm, not completely • Cognitive maps • Latent learning

So you don’t need a reinforcer? • Perhaps it strengthens the S-R bond, but is not necessary • Maybe it becomes part of a representation? • Colwill and Rescorla, 1985 • Chain - > water • Lever - > food • Food - > poison

That sound you hear is Skinner spinning in his grave • Rats won’t press the lever! • Therefore, the response and reinforcer have been connected, but not directly • It seems that associations are made between all three parts of the three term contingency

Are there 2 types of learning? • Yes, operant and classical • No, it is all associative learning • What a stupid question….

Pushing the limits • Heart rate conditioning • Biofeedback • Not always successful

So what is a reinforcer? • • • Some event that increases the likelihood of…. ENOUGH! Maybe it is need or drive reduction? Premack’s ideas Is food the reinforcer, or is it the act of eating?

Premack’s principle • Given 2 responses arranged in an operantconditioning procedure, more probable behaviors will reinforce less probable behaviors; less probable response will not reinforce more probable ones

Premack (1963) - study in Cebus monkeys Lever Pressing (L) : highest probability Door Opening (D) : medium probability Plunger Pulling (P) : lowest probability reinforces D; L reinforces P; P does NOT reinforce either D or L, etc. • “Premackian reinforcers”: activities that act as reinforcers (reading, playing) • Useful (and less costly) in token economies than object-based reinforcers • • •

Applications • Mitchell & Stoffelmayr (1973) use Premack’s principle in schizophrenics • Reinforcement items like candy, cigarettes, etc. usually not effective in schizophrenic patients • Sitting is a highly probable behavior in negative-symptom schizophrenics (catatonia, social withdrawal) • therapists made sitting contingent on doing small amount of work or activity; improved their negative symptoms • In unruly nursery-school children, high probability behaviors (running around, screaming) made contingent on low probability behaviors (sitting quietly, paying attention)

Behavioural economics • • Open economies Closed economies Elasticity of demand And behavioural ecology

Stimulus Control • Relationship between S and R • Reynolds (1961) - Attention in the pigeon • 2 pigeons reinforced for pecking to compound stimulus of white triangle on red background projected on response key • rate of pecking was then observed in each subject to a white triangle and a red background singularly

Why do you get a gradient? • Sort of the study of generalization and discrimination • How do you get data? – Probe trials – Test phases – Both done in extinction • Is it a property of the nervous system? – Pavlov – Hubel and Wiesel

Perhaps it is learning? • Lashley and Wade figured it was how the animal was trained • Borne out by Jenkins and Harrison’s work • Non differential • Presence absence • Explicit training

Duck you • Peterson (1960) and his ducklings, only saw yellow • They did not show gradients with colour! • Oops • Probably depends on the modality and the species being tested really • Or, could be relational Kohler and his chickens

• Gonzalez, Gentry, & Bitterman (1954) - chimpanzee had to pick among 9 squares of varying sizes • Squares 1, 5, & 9 presented, subject reinforced for choosing # 5 (intermediate size) • on test trial chimp reinforced no matter which square chosen • if given 4, 7, & 9 relational theory says subject will choose # 7, absolute theory says subject will choose square closest # 5 ( # 4) • subjects usually chose the intermediate size square of whichever three squares were presented

Peak Shift

OK, so that is odd • • Why does that happen? Excitatory gradient Inhibitory grandient As the Violent Femmes would say, you gotta add it up

Bird brain is not an insult • • • Concept learning Natural needed? How long lasting Delius (1982) Honig and Stewart (1988) The field of comparative cognition really grew out of much of this stuff