Overview of Advance Behavior Analysis

Reinforcement: Part 1
Week 5: Increasing Behavior
Positive Reinforcement
• Delivered Immediately after a behavior
• Delivered contingent upon a behavior
• Increases probability of future behavior
Potential Rule Governed Behavior
• When no immediate consequence
• Behavior changes without reinforcement
• Large increase in behavior follows one
instance of reinforcement
• Rule exists
Some important Points
We do not reinforce people
Practice is not reinforcement
Feedback is not always reinforcement
Artificial Reinforcement does not mean “fake”
Motivating Operations
• Establishing Operations: Increases temporarily
the immediate effectiveness of a reinforcer.
– Deprivation
• Abolishing Operations: Decreases temporarily
the immediate effectiveness of a reinforcer
– Satiation
4 term Contingency
No water Drinking Drink
Fluid intake
Reinforcement Classification
• Unconditioned
– Unlearned
– Primary
• Conditioned
– Secondary
– Learned
Generalized Conditioned
• Neutral Stimulus that is paired with many
other reinforcers
– The more reinforcers are paired with a neutral
stimulus the more likely the neutral stimulus
becomes a GCR
– E.g. money
Premack Principle
• LPR Before access to HPR
• Response-Deprivation Hypothesis (Timberlake
& Allison, 1974).
– This is an important consideration when
developing treatments.
– Don’t just make HPR contingent.
– It must be contingent and be deprived relative to
free operant levels!
Identifying Potential Reinforcers
• Stimulus Preference Assessment
– Ask People
• Open ended questions
• Choice Format
• Rank-ordering
– Observe
• Contrived
• Naturalistic
– Experiment
• Single Stimulus
• Paired Stimuli
• Multiple Stimuli
Identifying Potential Reinforcers
• Reinforcer Assessment
– Concurrent Schedule of Reinforcement
• May cause premature rejection (preference does not
mean ineffective reinforcer).
– Multiple Schedule Reinforcer Assessment
– Progressive-Ratio Schedule Reinforcer Assessment
• Increase response effort (schedule) and see how
behavior changes.
Other Methods of Identifying
• Non-contingent Reinforcement
• Differential Reinforcement
Non-Contingent Reinforcement?
• What is wrong with this term?
• Presenting Reinforcement on a time schedule
Tips for Using Reinforcement
Easily Achieved Initial Criterion
Sufficient Magnitude of Reinforcer
Vary Reinforcer
Direct contingencies when possible
Use prompts with reinforcement
Reinforce each occurrence initially
Use praise/attention with other reinforcers
Gradually increase response-to-reinforcement
delay (some call this fading – it is not – Thinning)
Negative Reinforcement
• Removal of aversive stimulus
• Immediately after a behavior
• Increases probability of behavior in future
Escape Versus Avoidance
• Escape – Termination
• Avoidance – Prevention
Schedules of Reinforcement
Fixed Ratio
• Reinforcement delivered after a specific
number of target responses are emitted.
– Post Reinforcement Pause
• Larger the ratio longer the PRP
– DeLuca & Holborn (1990)
• Provided FR schedule to pedaling behavior of fat
people. Fat people started pedaling more
Variable Ratio
• Reinforcement provided contingent upon a
variable number of responses.
• Strong Consistent Performance
• Group Contingencies makes use of VR
Fixed Interval
• Reinforcement provided after first response
after specified period of time.
• PRP and Scallop Effect
• Slow rates of responding
Variable Interval
• Reinforcement provided after first response
after variable duration of time elapsed.
• Constant stable rate of responding
• Limited Hold: amount of time R+ is available
once interval elapses.
Differential Reinforcement
DR-Lower Rates
DR-Higher Rates
Compound Schedules of
• Can be:
– Successive or simultaneously
– With or without discriminative stimuli
– Reinforcement for each element or for the
combination of all.
Concurrent Schedules of
• 2 contingencies of reinforcement operating
independently and simultaneously
– E.g. you can study or go out for the night
• Behaviors are Matched to maximize rates of
Discriminative Schedules
• Multiple Schedule (Stimulus discrimination)
– Alternating two or more schedules
– Uses an S+ to signal each schedule
• (e.g. 1 behavior class but two different situations for
two different reinforcement schedules)
• Chained Schedule
– Specific order of presentation
– Behavior may be different for each schedule
– All schedules must be complete for final outcome
Nondiscriminative Schedules
• Mixed Schedules
– Same as Mult. But no discriminative stimulus
– FR15/FI1 sometimes after FR15 and sometimes
after FI1
• Tandem Schedules
– Same as chained but no discriminative stimulus
– FR 15/FI2 means first instance will be 15 response
and second will be in 2 minutes
Combined Schedules
• Alternative Schedules
– Either or Schedule (whichever comes first)
• Conjunctive Schedules
– Both Schedules being met before reinforcement is
Matching Law
• Myerson Article

similar documents