Lecture 25: Stimulus Control III PSYC 2330 Fall 2024 PDF

Lecture 25: Stimulus Control III Instructor: Heather Collett PSYC 2330 Nov. 11 2024 Spence’s Theory of Discrimination Learning Kenneth Spence S- S+ Assumes reinforcement...

Lecture 25: Stimulus Control III Instructor: Heather Collett PSYC 2330 Nov. 11 2024 Spence’s Theory of Discrimination Learning Kenneth Spence S- S+ Assumes reinforcement of response in presence of S+ conditions excitatory response tendencies to S+ Non-reinforcement of responding during S- conditions inhibitory properties to S- that Non-reinforcement Reinforcement suppress instrumental behavior leads to S- leads to Differential responding to S+ acquiring excitatory inhibitory response and S- reflects both properties and tendencies and conditioned excitation to S+ conditioned conditioned and conditioned inhibition to inhibition to S- excitation to S+ S- Kearns and colleagues (2005) IV Cocaine Experimental S+ S+ S- Tone Clicker After being Reinforced Reinforced Light taught to the other ½ of When present, no ½ the time the time reinforcement lever press for cocaine IV Cocaine Control S+ S+ Light Tone Clicker ½ present, reinforcement, Reinforced Reinforced other ½ time present, no Training ½ the time the other ½ of reinforcement the time Kearns and colleagues (2005): Summation Test Assessed lever pressing when tone was played and with tone + light Test was first time Tone (S+) and Light (S-) were presented together Control group pressed the lever when the Tone and the Tone+Light were present with no effect of light on lever pressing The experimental group pressed the lever with Tone alone, but Tone+Light suppressed lever pressing The suppression of responding evident in the experimental group shows that a stimulus that is a signal for nonreinforcement (S−) in a discrimination procedure acquires active inhibitory properties What does this mean for conditioned learning of drugs of abuse? Excitatory conditioning can lead to drug self- administration S- cues can exert powerful inhibitory influence on drug seeking and drug self-administration behavior Inhibitory influence counteracts excitatory effects of S+ for drug administration Suggests that drug-seeking can be reduced by inhibition even if excitatory processes remain intact Interactions Between S+ and S-: The Peak Shift Effect Intradimensional Discrimination–training procedure in which S+ and S- differ only in terms of value of one stimulus feature Jenkins and Involved in expert performance Harrison (1960, 1962) Can produce peak-shift effect A displacement of the highest rate of responding in a stimulus generalization gradient away from the S+ in a direction opposite the S−.  Hanson (1959) – Peak Shift Effect Pigeons pecking in the presence of a light that was 550 nm Group 1(Control) : No discrimination training. Group 2: S+ = 550 nm S - = 590 nm Group 3: S+ = 550 nm S - = 555 nm 550 nm 555 nm 590 nm Only 5 nm away Hanson (1959) – Peak Shift Effect All the birds were tested for their rate of pecking in the presence of test stimuli that varied in color Control group pecked most in the presence of 550 nm which was reinforced during training showed a standard excitatory generalization gradient centered at the S+. S- 590 group pecked a lot for 550 nm but pecked more for 540 S-590 nm group pecked most for 540 nm and 530 nm and not for the trained S+ 550 nm The peak of the generalization gradient is away from S+ for stimuli never shown during training Spence’s Explanation of the Peak Shift Effect Spence’s Explanation: discrimination training leads to learning of excitation to S+ and inhibition to S- Both assumed to generalize to other S+ S- S- similar cues with the excitatory stimulus generalization gradient centered around S+/inhibitory gradient centered around S- S+ and S- overlap and inhibition to S- may 550 nm 555 nm 590 nm generalize to S+/suppress responding to S+ More inhibition from S- to S+ if S- is closer to More overlap with S+ S+; results in greater peak-shift effect Less overlap with S+ Spence’s S+ Explanation of the Peak Shift Effect S+ S- 550 nm 590 nm S- Stimulus Equivalence Training Being able to generalize stimuli can be also very important in our day to day lives Equivalence training encourages generalized responding among set of stimuli or establishment of stimulus equivalence Common Outcome Training: Links each of several different stimuli to common outcome Example = noise and clicker paired with food Involves training same response to several physically different stimuli Example = “fruit” to mean apples, strawberries, etc. Honey and Hall (1989): Stimulus equivalence Stage 1: Experimental Group Both Tone and A+ N+ Clicker paired Tone Clicker with food Control Group No food with A+ clicker Tone Tone paired N- with food Clicker  Honey and Hall (1989): Stimulus equivalence Stage 2: Tried to establish Experimental fear to the clicker Group noise Clicker noise was N+ paired with a Clicker shock for both groups Control Group Would this conditioned fear generalize to clicker generalize N+ to noise? Clicker  Both A and N were associated with Honey and Hall (1989) common outcomes for the experimental group Suppression ratio when presented with stimulus A Experimental group (A+N+) showed a much great suppression ratio to the presentation of stimulus A which was never paired with shock than the control group For the experimental group both stimulus A and N had been paired with food, so learning about stimulus N, generalized to stimulus A For the control group, where stimulus A and N had initially been paired with different outcomes (food or no food), new learning about stimulus N did not generalize to stimulus A Conditioned responding was approach to the food tray Contextual Cues and Conditional Relations Discrete Stimulus–presented for brief period; has clear beginning and end; can be easily characterized Discrete discriminative stimuli occur in presence of background contextual cues (visual, auditory, olfactory) where discrete discriminative stimuli presented Contextual cues can provide important additional source of control of learned behavior! Control by Contextual Cues Contextual cues signal reinforcement same way as discrete CS Contextual cues control behavior if they serve as a signal for a US or a reinforcer Studying could be associated with a stimuli in a context, like a library, and be harder to do away from that context Conditioned Place-Preference: Akins (1998) Contextual cues can come to control behavior Looked a sexual reinforcement in male quails One side of the apparatus was paired with a female quail Male quails would spend 5 minutes in the context and then another 5 minutes in that context with the female quail Male quails in control group had access to a female quail 2-h prior to being put in the apparatus so the 2 would not be paired Side preference was measured prior to training, after 5 trials, and after 10 trials Conditioned Place-Preference: Akins (1998) Male quails that experienced the pairing of the female with one compartment of the box, showed a preference for the that side of the box over the non-paired side CS context paired with sexual reinforcement came to prefer that context. Quails in the control group, where place and the reinforcer were not paired, showed no sig. preference for one side Contextual cues can control behaviour if it is learned that the context predicts a reinforcer Drugs that show a similar effect can Trial 5 Trial 10 become drugs of abuse Conditional relation Experiment by Thomas, McKelvie, and Mah (1985) illustrates control by contextual cues that are not correlated with the availability of reinforcement Context 2 S+ S- Vertical Horizontal line line After discrimination training, S+ S- the context were changed, and the discrimination training Context 1 contingencies were reversed Conditional relation After the pigeons had learned each task, they were then tested with each task in its associated context In context 1, pigeons responded the most to vertical line and the least to the horizontal line, consistent to the environment they learned about it it In context 2, they responded more to the horizontal line and least to the vertical line Learning in context 2 did not generalize back to context 1 Instead of being associated with a particular stimulus, each context came to activate a different S+/S− contingency. Conditional relation: If Context 1, then 90+/0−; if Context 2, then 90−/0+  Control by Conditional Relations Modulator–nature of binary relation determined by third event; conditional relation Animals learn to use modulators to tell when binary relation is in effect In Classical Conditioning: Establishing conditional relation requires introducing modulator indicating when Modulator: Light/No Light presentation of auditory CS Modulators can also be called will end in food occasion setters or facilitators