Transcript positive reinforcement

Getting a Behavior to Occur
More Often with Positive
Reinforcement
Chapter 3
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Positive Reinforcement
•
A positive reinforcer is an event that, when presented immediately
following a behavior, causes the behavior to increase in frequency (or
likelihood of occurrence).
•
The principle called positive reinforcement states that if, in a given
situation, somebody does something that is followed immediately by a
positive reinforcer, then that person is more likely to do the same
thing again when he or she next encounters a similar situation.
•
The terms positive reinforcement and reinforcement are often
used interchangeably.
•
Operant behaviors are those that operate on the environment to
generate consequences, and are in turn influenced by those
consequences; operant behaviors followed by reinforcers are
strengthened.
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Positive Reinforcement
(Example)
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Selecting the Behavior
to Be Increased
– The behaviors to be reinforced must first be
identified specifically in order to:
(a) help to ensure the reliability of detecting
instances of the behavior and changes in its
frequency, which is the yardstick by which one
judges reinforcer effectiveness;
(b) increase the likelihood that the reinforcement
program will be applied consistently.
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Copyright © 2011 Pearson Education, Inc. All rights reserved.
• Choosing Reinforcers (“Different Strokes
for Different Folks”)
– Most positive reinforcers can be classified under five somewhat
overlapping headings:
• consumable (i.e. items that one can eat or drink)
• activity (i.e. opportunities to watch TV, look at a picture book,
etc.)
• manipulative (i.e. opportunities to play with a favorite toy, color
or paint, ride a bicycle, etc.)
• possessional (i.e. enjoy some other item that one can possess)
• social (i.e. affectionate pats and hugs, praise, nods, smiles, etc.)
– The extrinsic–intrinsic distinction between reinforcers may not
be valid
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Motivating Operations
– Temporarily alter the effectiveness of a reinforcer, and alter the
frequency of behavior reinforced by that reinforcer.
– Deprivation vs. satiation
• Most reinforcers will not be effective unless the individual has
been deprived of them for some period of time prior to their use
(establishing operations vs. abolishing operations).
– Motivational variable
• A variable that affects the likelihood and direction of behavior.
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Factors Influencing the Effectiveness
of Positive Reinforcement (cont.)
• Reinforcer Size
– The size (or amount or magnitude) of a reinforcer is an
important determinant of its effectiveness.
• Instructions: Make Use of Rules
– Specific instructions will speed up the learning process for
individuals who understand them.
– Instructions may influence an individual to work for delayed
reinforcement.
– Adding instructions to reinforcement programs may help to
teach individuals to follow instructions.
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Factors Influencing the Effectiveness
of Positive Reinforcement (cont.)
• Reinforcer Immediacy
– For maximum effectiveness, a reinforcer should be given
immediately after the desired response.
– The distinction between direct- and indirect-acting effects of
reinforcement has important implications for practitioners.
– If you can’t present a reinforcer immediately following the
desired behavior, then provide instructions concerning the
delay of reinforcement.
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Example of Delayed Reinforcement
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Factors Influencing the Effectiveness
of Positive Reinforcement (cont.)
• Contingent versus Noncontingent
Reinforcement
– When a behavior must occur before a reinforcer will be
presented, we say that the reinforcer is contingent on that
behavior.
– If a reinforcer is presented at a particular time, regardless of
the preceding behavior, we say that the reinforcer is
noncontingent.
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Factors Influencing the Effectiveness
of Positive Reinforcement (cont.)
• Weaning the Student from the Program and
Changing to Natural Reinforcers
– Unprogrammed reinforcers that occur in the normal course of
everyday living are called natural reinforcers, and the settings
in which they occur are called the natural environment.
– The behavior modifier should always try to ensure that the
behavior being established in a training program will be
reinforced and maintained in the natural environment.
– If a behavior that has been strengthened by reinforcement is no
longer reinforced at least occasionally, then that behavior will
return to its original level.
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Pitfalls of Positive Reinforcement
• How the Principle Can Work Against the
Unwary
– Those who are aware of the principle of positive reinforcement
can use it to bring about desirable changes in behavior; the
principle operates equally well for those who are not aware of it.
• Other Pitfalls
– The tendency for novice behavior modifiers to assume that
presenting reinforcers noncontingently will strengthen a specific
behavior.
– To mistakenly offer positive reinforcement as an overly simplistic
explanation of a change in behavior.
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Guidelines for the Effective Application
of Positive Reinforcement
1.
Selecting the behavior to be increased.
–
The target behavior should be a specific behavior rather than a
general category. Also, if possible, select a behavior that will come
under the control of natural reinforcers after it has been increased
in frequency.
2.
Selecting the reinforcer.
–
Complete the reinforcer survey presented in Figure 3-3 and select
strong reinforcers that
•
are readily available.
•
can be presented immediately following the desired behavior.
•
can be used over and over again without causing rapid satiation.
•
do not require a great deal of time to consume (if it takes a halfhour to consume the reinforcer, this minimizes the training
time).
–
Use as many reinforcers as feasible, and, where appropriate, use a
reinforcer tray or menu.
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Guidelines for the Effective Application
of Positive Reinforcement (cont.)
3. Applying positive reinforcement.
–
–
–
–
Tell the individual about the plan before starting.
Reinforce immediately following the desired behavior.
Describe the desired behavior to the individual while the reinforcer
is being given.
Use lots of praise and physical contact when dispensing reinforcers.
4. Weaning the student from the program.
–
–
–
If a behavior has been occurring at a desirable rate, you might try to
gradually eliminate tangible reinforcers (such as treats and toys)
and maintain the behavior with social reinforcement.
Look for reinforcers in the natural environment that might
maintain the behavior once it has been increased in frequency.
To ensure that the behavior is being reinforced occasionally and
that the desired frequency is being maintained, plan periodic
assessments of the behavior after the program has terminated.
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Developing and Maintaining
Behavior with Conditioned
Reinforcement
Chapter 4
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Types of Reinforcers
• Unconditioned Reinforcers
– Unlearned
• Conditioned Reinforcers
– Learned through being paired with other reinforcers
– Tokens
• Conditioned reinforcers that can be collected and
exchanged for backup reinforcers
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Conditioned Reinforcers
• Main advantages:
– Can often be delivered more immediately
than the backup reinforcer
– Help bridge delays between behavior and
more powerful reinforcers
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Table 4-1 – Examples of Conditioned Reinforcers.
“Examples of simple and generalized conditioned
reinforcers are given in Table 4-1.” (page 57)
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Influencing Effectiveness of
Conditioned Reinforcers
• Strength of Backup Reinforcers
• Variety of Backup Reinforcers
– Simple Conditioned Reinforcer – paired with a single
backup reinforcer
– Generalized Conditioned Reinforcer – paired with many
different kinds of backup reinforcers
• Strength depends in part on number of different backup
reinforcers available for it
• Schedule of Pairing with Backup Reinforcer
– More effective if does not follow each occurrence
• Extinction of the Conditioned Reinforcer
– Must continue to pair conditioned reinforcer with backup
reinforcer, at least occasionally
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Pitfalls
• Scolding
– No backup punishment or desirable behavior
is not reinforced
• Attention may be reinforcing, thus scolding may
not be punishing
• Extinction of conditioned reinforcer
– Failure to use backup reinforcers might make
conditioned reinforcers ineffective
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Guidelines for Effective Use
of Conditioned Reinforcers
1. Conditioned reinforcer should be a stimulus that can
be managed and administered easily
2. Use the same conditioned reinforcers found in
individual’s natural environment
3. Early on, present backup reinforcer as soon as possible
after conditioned reinforcer
–
Delay can be increased gradually later
4. Use generalized conditioned reinforcers whenever
possible
5. When more than one participant is involved, avoid
destructive competition for conditioned and backup
reinforcers
6. Apply the same rules used for primary positive
reinforcers
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Decreasing a Behavior
with Extinction
Chapter 5
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Extinction
• When a response (previously reinforced)
is not followed by reinforcement, the
individual is less likely to engage in the
behavior again.
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Extinction
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Factors Influencing the
Effectiveness of Extinction
1.
Controlling reinforcers for the behavior that is to be
decreased
–
–
2.
Extinction combined with positive reinforcement for an
alternate behavior
–
3.
Reinforcers from others or from environment can undo
extinction
Make sure that the reinforcers you are withholding are
actually the ones maintaining the behavior
Extinction most effective when alternate behavior is
reinforced
Setting in which extinction is carried out
–
–
Minimize the influence of alternative reinforcers on the
undesirable behavior (from others in the old setting)
Maximize the chances of the behavior modifier persisting with
the program
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Factors Influencing the
Effectiveness of Extinction
4.
5.
6.
7.
Instructions: Make use of rules
–
May speed up behavior change
–
–
–
When behavior was constantly reinforced, quick to extinction
Intermittent reinforcement may lead to slower extinction
Behaviors slow to extinguish are said to be slow to extinction
–
–
Extinction burst – increase in responding during extinction
In situations where extinction burst may be harmful, need to
take preventative measures or extinction should not be used
Extinction may be quicker after continuous reinforcement
Behavior being extinguished might get worse before it gets
better
Extinction may produce aggression
–
Aggression lower in programs that reinforce alternative
behaviors
8. Extinguished behavior may reappear after a break
–
this is called spontaneous recovery
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Pitfalls
• Extinction program might be undone by
other, less knowledgeable people
• Desirable behavior may be extinguished
when insufficient reinforcement is given
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Pitfalls of Extinction
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Guidelines for Effective
Application of Extinction
• Select behavior to be decreased
– Be specific
– Select proper location (behavior might get worse before getting
better)
– Select behavior for which you can control the reinforcers
maintaining it
• Preliminary considerations
–
–
–
–
Keep track of how often behavior occurs before extinction
Identify current reinforcers
Identify desirable alternate behavior
Identify effective reinforcers for the desirable alternate
behavior
– Select setting in which extinction will work best
– Let all relevant individuals know what is going on (what is
being extinguished and what is being reinforced)
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Guidelines for Effective Application of
Extinction (cont.)
• Implementing the Plan
– Tell the individual about the plan before starting
– Use rules of reinforcement if rewarding alternate
behavior
– Be consistent
• Weaning from the program
– Be prepared for occasional relapses
– Three possible reasons for failure:
1. Attention withholding is not the reinforcer
2. Undesirable behavior is receiving intermittent
reinforcement from another source
3. The desired alternative behavior has not been
strengthened sufficiently
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Developing Behavioral
Persistence Through the Use of
Intermittent Reinforcement
Chapter 6
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Definitions
• Schedule of reinforcement –
– rule specifying which occurrences of a given
behavior, if any, will be reinforced
• Continuous Reinforcement (CRF):
– every correct response is reinforced; fast
learning & fast extinction
• Intermittent Reinforcement:
– only some correct responses are reinforced;
slow learning & extinction
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Definitions
• Acquisition Phase
– Behavior is being conditioned or learned
• Maintenance Phase
– Behavior has become well-learned
• Free-Operant Procedure
– Individual is “free” to respond repeatedly
– There are no constraints on successive responses
• Discrete-Trials Procedure
– Distinct stimulus is presented prior to an opportunity for
a response to occur and is followed by reinforcement
– Rate of responding is limited to the rate at which
successive stimuli at the beginning of each trial are
presented
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Intermittent Reinforcement
• Advantages
– Reinforcer remains effective longer because
satiation takes place more slowly.
– Behavior that has been reinforced intermittently
tends to take longer to extinguish.
– Individuals work more consistently on certain
intermittent schedules.
– Behavior that has been reinforced intermittently
is more likely to persist after being transferred to
reinforcement in natural environment.
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Ratio Schedules
• Based on number of responses emitted
• Fixed-ratio (FR) schedule
– Reinforcement occurs each time a set number of
responses of a particular type is emitted.
• Ratio strain – deterioration of responding from increasing
an FR schedule too rapidly
– Produce high steady rate of responding until
reinforcement, followed by a post-reinforcement
pause
• The higher the value of the FR, the longer the pause
– Initially produces high rate of responding during
extinction
– Produces high resistance to extinction
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Ratio Schedules
• Variable-ratio (VR) schedule
– The number of responses required to produce
reinforcement changes unpredictably from one
reinforcement to the next.
– Produces a high steady rate of responding.
– Produces no (or at least very small) post-reinforcement
phase
• Differences between VR and FR schedules:
– VR schedules can be increased more abruptly than FR
schedules without producing ratio strain
– Values of VR that can maintain a behavior are somewhat
higher than those of FR
– VR produces higher resistance to extinction than FR of
same value does
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Simple Interval Schedules
• Schedules based on time
• Fixed-Interval (FI) Schedule
– The first response after a fixed amount of time
following the previous reinforcement is reinforced
and a new interval begins
– Size of FI schedule: amount of time that must elapse
– No limit on how long after the end of the interval a
response can occur in order to be reinforced
– FI Schedules produce:
• A rate of responding that increases gradually near the end
of the interval until reinforcement
• A post-reinforcement pause
– Length depends on value of FI – the higher the value, the longer
the pause
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Simple Interval Schedules
• Variable-Interval (VI) Schedule
– The length of the interval changes unpredictably
from one reinforcement to the next
– Lengths of VI schedule vary around some mean
value
– Produces a moderate, steady rate of responding
and no post-reinforcement pause
– Produces high resistance to extinction
– Responding is lower during extinction after VI
than after FR or VR
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Simple Interval Schedules
• Simple interval schedules are not often
used because:
– FI produces long post-reinforcement pauses
– VI generates lower response rates than ratio
schedules
– Simple interval schedules require continuous
monitoring of behavior after each interval
until a response occurs
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Interval Schedules
with Limited Hold
There is a finite time after a reinforcer
becomes available that a response will
produce it.
–FI/LH
–VI/LH
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Limited Hold
• Short limited holds – similar results to
ratio schedules
• For small FIs, FI/LH produce results
similar to FR schedules
• Variable Interval, Limited Hold – similar
results to VR schedules
• Used when ratio-like behavior is desired,
but unable to count each instance of
behavior
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Duration Schedule
• Reinforcement occurs after the behavior
has been engaged in for a continuous
period of time
– Fixed Duration (FD) – the period is fixed
– Variable-Duration (VD) – interval changes
unpredictably
• Used only when target behavior can be
measured continuously
Copyright © 2011 Pearson Education, Inc. All rights reserved.
FIGURE 6-1 Diagrams illustrating the differences between the time-based
schedules described in the text. In each diagram, the horizontal line
represents a period of time. “In a fixed-interval (FI) schedule, a
reinforcer is presented following the first instance of a specific response after
a fixed period of time (see Figure 6-1).” Page 81
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Concurrent Schedules
of Reinforcement
• Schedules of reinforcement that are in effect at any
given time
• Herrnstein’s (1961) matching law:
– The response rate or the time devoted to an activity in a
concurrent schedule is proportional to the rate of
reinforcement of that activity relative to the rates of
reinforcement on the other concurrent activities.
• Research findings on factors influencing choice of
reinforcement:
–
–
–
–
Types of schedules that are operating
The immediacy of reinforcement
The magnitude of reinforcement
Response effort involved in different options
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Pitfalls of Intermittent
Reinforcement
• Failure to conduct extinction correctly
may turn into intermittent reinforcement
– Ex: Child tantrums – ignore first, but then
give in
• Failure to introduce intermittent schedule
gradually enough may result in loss of the
behavior
Copyright © 2011 Pearson Education, Inc. All rights reserved.
Guidelines for the Effective Use of
Intermittent Reinforcement
• Choose appropriate schedule for behavior you
wish to strengthen
• Choose schedule that is convenient to administer
• Use appropriate instruments and materials to
determine accurately and conveniently when the
behavior should be reinforced
• Frequency of reinforcement should initially be
high enough to maintain desired behavior, then
decrease gradually
• Inform individual of what schedule you are using
Copyright © 2011 Pearson Education, Inc. All rights reserved.