Stimulus Control of Behavior Notes
Stimulus Control of Behavior
- Behaviors are controlled by environmental stimuli.
- In the wild, survival depends on adjusting behavior to environmental changes.
Stimulus Control
- Organism is under stimulus control when behavior is determined by the stimulus.
- Reynolds (1961) showed that it is not always clear what stimulus controls instrumental behavior.
- Pigeons displayed differential responding, showing stimulus discrimination.
Stimulus Generalization
- Occurs when an animal responds similarly to two or more stimuli.
- Pavlov discovered stimulus generalization; animals responded to tones of similar frequency after training with a specific tone.
Measuring Stimulus Control
- Stimulus generalization gradients measure stimulus control and reflect sensitivity to stimulus changes.
- Guttman & Kalish (1956) conducted early studies on stimulus generalization.
- The stimulus with the greatest control evokes the largest response during testing.
- Discriminative stimulus (SD) is a stimulus that has gained control of instrumental responding.
No Stimulus Control
- A flat gradient occurs when an organism responds similarly to all stimuli.
Factors in Stimulus Control
- Pavlov discovered that an animal learns about the more salient stimulus when two stimuli are presented at the same time.
- Overshadowing: A more salient stimulus prevents conditioning to a less salient one.
- Stimulus control is influenced by the type of reinforcement.
- Light easily associates with food reward
- Tone easily associates with shock avoidance.
Instrumental Repsonse
- Stimulus control depends on the type of response.
- Dobrzecka, Szwejkowska, & Konorski (1966) showed responses differentiated by location are controlled by spatial features, and those differentiated by quality are controlled by the quality of stimuli.
Stimulus Discrimination Training
- Stimulus control depends largely on an animal's learning experiences.
- Stimulus discrimination procedures effectively bring behavior under control of a stimulus.
- Campolattaro, Schnitker, and Freeman (2008) trained rats on eyeblink conditioning.
- A+ paired with shock, B- was not.
- Blinking behavior fell under control of A+.
Instrumental Conditioning
- Stimulus discrimination training establishes stimulus control.
- Rats learn to press a lever with one stimulus but not another.
- Pigeons learn to peck to one color but not another.
Discrimination Training
- Jenkins and Harrison (1960, 1962) researched pigeons pecking for food reinforcement.
- Group 1: Reinforced for pecking to a 1000 Hz (S+) tone but not for pecking to a 950 Hz tone (S-).
- Group 2: Reinforced for pecking to a 1000 Hz (S+) tone but not for pecking during absence of tone (S-).
- Group 3: Reinforced anytime it pecked and a 1000 Hz tone was always on. THIS GROUP WAS NOT DISCRIMINATING.
- Stimulus control is greatest when S+ and S- vary along the same stimulus dimension.
Interoceptive Cues
- Schall et al. (1996) showed animals discriminate between internal sensations from drug states.
- Phase 1: Pigeons injected with 3.0 mg/kg cocaine prior to session in which pecking was reinforced on a VI 2 min schedule.
- Phase 2: Pecking was reinforced on drug- injected sessions (3.0 mg/kg ) but not on non- drug sessions.
- A (3 mg/kg) Cocaine-Induced State Served As The S+.
- A Non-Drugged State Served As The S-.
Compound Stimuli
- Stimulus-element approach: Stimuli are perceived as distinct and separate.
- Configural-cue approach: Organisms treat complex stimuli as indivisible wholes.
Configural Cues
- Rats discriminate between configural cues and separate stimulus elements.
- Positive and Negative Patterning - Control of Behavior by Configural Cues
Discrimination Training Learning
- Possible behavioral strategies:
- Respond whenever S+ is present, regardless of S-.
- Do not respond when S- is present, regardless of S+.
- Respond only when S+ is present and not when S- is present.
Spence’s Theory of Discrimination Learning
- Kenneth Spence (1936) believed animals learn about both S+ and S-.
- Animals acquire excitatory tendencies to S+ and inhibitory tendencies to S-.
Excitatory and Inhibitory
- Honig, Boneau, Burstein, & Pennypacker, (1963) Group 1: Reinforced for pecking when response key was illuminated with a white light and black vertical bar (S+). Not reinforced when key illuminated with a white light and no bar (S-).
S- Inhibition
- The summation test demonstrates that S- acquires active inhibitory properties.
- Kearns et al. (2005) studied lever pressing in rats with intravenous doses of cocaine used as reinforcement.
- Experimental Group: Tone, Clicker: Reinforcement; Light: No Reinforcement
- Control Group: Tone, Clicker: Reinforcement; Light: presented half the time with clicker (and reinforcement) and half the time alone (with no reinforcement)
- The Light Inhibited Lever Pressing For the Experimental Group
- This is just differential inhibition!
S+ and S-
- Sometimes S+ learning influences responding to S-.
- Occurs in intradimensional discriminations (S+ and S- differ on one stimulus feature).
- Hanson (1959) studied discrimination in pigeons.
- All Groups were reinforcement for pecking with a key light of 550 nm (S+).
- Group 1: S- was a color of 590-nm wavelength.
- Group 2: S- was a color of 555-nm wavelength.
- Group 3: Control group – No S-. Just rewarded during S+.
- Peak Shifts Can Occur With Any Intradimensional Discriminations When S+ and S- Are Similar
Spence Peak-Shift
- Spence (1937) predicted the peak-shift phenomenon.
- Generalized inhibition from S- suppresses responding to S+.
- Peak shifts are less likely when S+ and S- are further apart.
Stimulus Equivalence Training
- Has the opposite effect of discrimination training, promoting greater generalization.
- Subjects are trained to treat similar stimuli equivalently.
- Pigeons can be trained to respond similarly to different photographs, all of which contain water, and they can also learn to respond similarly to photographs containing human beings in them.
- This represents a form of categorical learning.
Stimulus Equivalence Training
- Stimulus equivalence training can be established by linking two stimuli to a common third event, like a food stimulus.
- Honey and Hall (1989) performed the following experiment
- Phase 1: Experimental Group
- Noise– Food
- Clicker – Food
- Phase 1: Control Group
- Phase 2: All Groups
Contextual Control
- Behavior can come under the control of contextual cues.
- Akins (1998) demonstrated this in male domesticated quail.
- Exp Group: Birds were placed in CS compartment for 5 minutes. A sexually-receptive female was then placed in compartment for another 5 minutes. CS (compartment) – US (sex) paired.
Conditional Relations
- Contextual cues do not have to signal reinforcement to gain control over behavior.
- Thomas, McKelvie, & Mah (1985) trained birds in two different contexts (lighting and type of noise).
- The S+/S- contingencies were reversed between the two contexts.
Pavlovian Conditioning
- Establishing a conditional relation in classical conditioning requires introducing a modulator that indicates when a CS will be followed by reinforcement.
- Rescorla, Durlach & Grau (1985) used Noise as a modulator for pigeons pecking an illuminated orange key light.
- If Noise: Light – Food
- If No Noise: Light – No Food
Conditional Control
- Fetsko, Stebbins, Gallagher, & Colwill (2005) studied occasion setting in mice.
- Whenever a light stimulus (modulator) was present, a noise stimulus would be followed by food.
- Mice showed much more conditioned responding on trials that included the light.
- The light, by itself, did not develop excitatory conditioning properties.
Conditional Control in Pavlovian Conditioning
- Occasion Setting Light indicates US is coming
- Conditioned Inhibition Light (CS-) indicates US is not coming
- The procedure for occasion setting is the converse of the standard procedure for conditioned inhibition.