Fading, Chaining, Flashcards
What is successive Discrimination?
A provider used to train a differential respondent by alternating between S^D and S^DELTA. Relies on Generalization and Extinction
Fading
Transferring the control of one antecedent stimulus (S1) to another stimulus (S2) by gradually eliminating S1 in the presence of S2
-definition fro text: the gradual elimination of the prompt as the behavior continues to occur in the presence of S^D
Fading produces
Little to no generalization effects
-very difficult to reverse roles of S^D and S^delta
Fading procedures are best sued where
Contingencies of reinforcement are not likely to change and variability is undesirable
-multiplication tables etc.
Typically fading is more desirable than
Trial and error methods
-does not contain undesirable effects of extinction
Prompt
An antecedent stimulus/ event that increases the likelihood that a person will engage in the correct behavior at the correct time
-Prompt = controlling stimulus (i.e., S^D or S^delta)
NOTE: the definition of fading in the text is poor because fading can generate S^D or S^delta
Response Prompts
Involve the behaviour of another person
Verbal prompt
-Verbal behavior of another person results in correct behavior of the trainee
-Gestural Prompt
-a physical movement or gesture of another person that leads to the correct behavior
-Modeling Prompt
-person demonstrates the target behavior for the learner
-Physical Prompt
-person physically assists the leaner to engage in the correct behavior
Stimulus Prompts
-involve a change or addition/subtraction of an antecedent stimulus
-Within-Stimulus Prompt
-an aspect/dimension of the stimulus is changed to aid discrimination
-Extra-Stimulus Prompt
-A separate stimulus is added to aid discrimination
-Prompt/Stimulus Fading
- gradually removing the prompt/stimulus in the presence of the new S^D that is being established
- Most to least
- Least to most
-Prompt Delay
- gradually increase the delivery of the prompt in the presence of the new S^D that is being established
- initially present prompt with no delay and then gradually insert a delay between the prompt and the desired SD
1. Present the desired S^D
2. Want for response (delay)
3. If no response made after x amount of time, prompt the desired behavior
4. Observe the desired behavior
5. Reinforce the desired behavior
Behavior Chains
– a specific sequence of discrete responses each associated with specific stimulus condition
- i.e., a complex behavior consisting of 2 of more component behaviors that occur together in a sequence - most behaviors are complex combinations of simpler behaviors in a particular order
Each component of a behavior chain has
a S^D and a corresponding response to that S^D
Each response in a behavior chain
creates a new situation that functions as an S^D for the next response
Task Analysis for behavior chains
Breaks down the task into its component elements
-establishing the stimuli (S^D) responses and reinforcers
Analysis should be
- Validated
- observe practice and behaviors
- consult experts
- perform the behavior yourself
- adjust steps based on the learners’’ ability
Prior to chaining, the sub subject
Must be capable of executing each component
-revise if necessary (break down, combine or add steps)
Forward chaining steps
- Reinforce after performance of first two components
- repeat until behavior is performed reliably - reinforce after complete performance of the 1st 3 components
- reinforce after complete performance of first four components
- and so on…
Backward Chaining
- Require performance of last 2 component and reinforce
a. Repeat until smooth and prompts have been faded - Require performance with last 3 components and reinforce
- And so on…
Note: components are added from back to front
- A prompt (to be faded) may be needed to elicit new components
Chaining long-term considerations
- if learned chain is too long, responding may decrease over time due to
- weak conditional reinforcers
- weak terminal/primary reinforcers
Potential remedies of chaining
- use a stronger terminal reinforcer
- shorten the chain
- highlight natural rewards the chain provides (if any)
- intermittently reinforce smaller sections of the chain
Variable Chains.
- not all chains occur in the same sequence with every presentation
- train components out of order
- leads to flexible problem-solving
