Unit 12: Generalization and Maintenance

Behavior traps are powerful contingencies of reinforcement with four defining features:

(a) They are "baited" with virtually irresistible reinforcers; (b) only a low-effort response already in the student's repertoire is needed to enter the trap; (c) interrelated contingencies of reinforcement inside the trap motivate the student to acquire, extend,and maintain targeted skills; and (d) they can remain effective for a long time.

multiple exemplar training

Instruction that provides the learner with practice with a variety of stimulus conditions, response variations, and response topographies to ensure the acquisition of desired stimulus controls response forms; used to promote both setting/situation generalization and response generalization. (See teaching sufficient examples.)

__________________________ can be succinctly defined as the occurrence of new responses to the original training stimulus.

Response generalization

he shift from formal intervention procedures to a normal everyday environment

can be accomplished by gradually withdrawing elements comprising the three components of the training program:(a) antecedents,prompts,or cuerelated stimuli; (b) task modifications and criteria; and (c) consequences or reinforcement variables.

Other types of generalized outcomes (e.g.,stimulus equivalence, contingency adduction, and generalization across subjects)

do not fit easily into categories of response maintenance, setting/situation generalization, and response generalization.

general case analysis

A systematic process for identifying and selecting teaching examples that represent the full range of stimulus variations and response requirements in the generalization setting(s). (See also multiple exemplar training and teaching sufficient examples.)

programming common stimuli

A tactic for promoting setting/situation generalization by making the instructional setting similar to the generalization setting; the two-step process involves (1) identifying salient stimuli that characterize the generalization setting and (2) incorporating those stimuli into the instructional setting.

Undesired response generalization occurs

when any of a learner's untrained but functionally equivalent responses produce undesirable outcomes.

Teaching a learner self-management skills with

which he can prompt and maintain targeted behavior changes in all relevant settings at all times is the most potentially effective approach to mediating generalized behavior changes.

Minimum difference negative teaching examples

which share many characteristics with positive teaching examples, help eliminate "generalization errors" due to overgeneralization and faulty stimulus control.

indiscriminable contingency

A contingency that makes it difficult for the learner to discriminate whether the next response will produce reinforcement. Practitioners use indiscriminable contingencies in the form of intermittent schedules of reinforcement and delayed rewards to promote generalized behavior change.

generalization

A generic term for a variety of behavioral processes and behavior change outcomes. (See generalization gradient, generalized behavior change, response generalization, response maintenance, setting/situation generalization, and stimulus generalization.)

lag reinforcement schedule

A schedule of reinforcement in which reinforcement is contingent on a response being different in some specified way (e.g., different topography) from the previous response (e.g., Lag 1) or a specified number of previous responses (e.g., Lag 2 or more).

teaching sufficient examples

A strategy for promoting generalized behavior change that consists of teaching the learner to respond to a subset of all of the relevant stimulus and response examples and then assessing the learner's performance on untrained examples. (See multiple exemplar training.)

behavior trap

An interrelated community of contingencies of reinforcement that can be especially powerful, producing substantial and long-lasting behavior changes. Effective behavior traps share four essential features: (a) They are "baited" with virtually irresistible reinforcers that "lure" the student to the trap; (b) only a loweffort response already in the student's repertoire is necessary to enter the trap; (c) once inside the trap, interrelated contingencies of reinforcement motivate the student to acquire, extend, and maintain targeted academic and/or social skills; and (d) they can remain effective for a long time because students shows few, if any, satiation effects.

contrived contingency

Any contingency of reinforcement (or punishment) designed and implemented by a behavior analyst or practitioner to achieve the acquisition, maintenance, and/or generalization of a targeted behavior change.

naturally existing contingency

Any contingency of reinforcement (or punishment) that operates independent of the behavior analyst's or practitioner's efforts; includes socially mediated contingencies contrived by other people and already in effect in the relevant setting.

generalization probe

Any measurement of a learner's performance of a target behavior in a setting and/or stimulus situation in which direct training has not been provided

generalization setting

Any place or stimulus situation that differs in some meaningful way from the instructional setting and in which performance of the target behavior is desired.

contrived mediating stimulus

Any stimulus made functional for the target behavior in the instructional setting that later prompts or aids the learner in performing the target behavior in a generalization setting.

Natural contingencies of reinforcement that are easy to enter, but difficult to exit are called:

Behavioral trap

generalization across subjects

Changes in the behavior of people not directly treated by an intervention as a function of treatment contingencies applied to other people

_________________ (also called ________________) is a systematic method for selecting teaching examples that represent the full range of stimulus variations and response requirements in the generalization setting.

General case analysis; general case strategy

In spite of his long history of distractibility and poor endurance, Andrew had learned to work independently for several hours at a time in the copy room at his vocational training center. Andrew has now just begun his first competitive job working as a copy machine operator in a downtown business office. His employer, however, is complaining that Andrew frequently stops working after a few minutes to seek attention from others. Andrew may soon lose his job. What intervention could you, as his behavior analyst, have done differently before he began his new job to prevent him from this failure?

Identify the characteristics of the break room in the future competitive employment setting, and set up similar settings in the training center, in order to train Andrew to wait until he is in those settings to seek social reinforcement from his co-workers.

A general finding in the Experimental Analysis of Behavior is that, for difficult sensory discriminations, i.e., where there is high likelihood of stimulus generalization from the target SD to the S-Delta, it is important that the training include exposure to both the target SD, during which responding is reinforced, and the S-Delta, during which responses go unreinforced, in order for the organism to learn to respond to the SD but not to the S-Delta. A similar issue arises in Applied Behavior Analysis and is addressed in a parallel fashion via a procedure known as _________________________________.

Instruction with negative, or "Don't do it," teaching examples

Efforts to promote generalized behavior change will be enhanced by adhering to five guiding principles:

Minimize the need for generalization as much as possible. Conduct generalization probes before,during,and after instruction. Involve significant others whenever possible. Promote generalized behavior change with the least intrusive, least costly tactics possible. Contrive intervention tactics as needed to achieve important generalized outcomes.

___________________ refers to an outcome in which the target behavior has come under the control of a stimulus class that is too broad.

Overgeneralization

Undesirable setting/situation generalization takes two common forms:

Overgeneralization,in which the behavior has come under control of a stimulus class that is too broad, and faulty stimulus control, in which the behavior comes under the control of an irrelevant antecedent stimulus.

teaching loosely

Randomly varying functionally irrelevant stimuli within and across teaching sessions; promotes setting/situation generalization by reducing the likelihood that (a) a single or small group of noncritical stimuli will acquire exclusive control over the target behavior and (2) the learner's performance of the target behavior will be impeded or "thrown off" should he encounter any of the "loose" stimuli in the generalization setting.

The extent to which a learner continues to perform the target behavior after a portion or all of the intervention responsible for the behavior's initial appearance in the learner's repertoire has been terminated is called ___________.

Response maintenance

Gabriel is a ten-year-old boy diagnosed with autism. In an effort to meet an objective on his individualized education program that targets functional language and communication skills, Gabriel's teacher taught him to say, "Hello, how are you?" as a greeting. Now, whenever Gabriel meets anyone, he invariably responds with, "Hello, how are you?" Gabriel's parents are concerned that their son's language seems rote and robotic. What could you, as his behavior analyst, recommend to the teacher as a new procedure to correct this rote performance?

Set up lag schedule of reinforcement in which Gabriel is reinforced only for saying something different from his last two greetings to a series of consecutive surrogate greeters. In addition provide the parents with training to implement the same lag schedule.

____________________ refers to the phenomenon in which a response that has been reinforced in the presence of a given stimulus occurs with an increased frequency in the presence of different but similar stimuli under extinction conditions.

Stimulus generalization

instructional setting

The environment where instruction occurs; includes all aspects of the environment, planned and unplanned, that may influence the learner's acquisition and generalization of the target behavior.

response maintenance

The extent to which a learner continues to perform the target behavior after a portion or all of the intervention responsible for the behavior's initial appearance in the learner's repertoire has been terminated. Often called maintenance, durability, behavioral persistence, and (incorrectly) resistance to extinction.

setting/situation generalization

The extent to which a learner emits the target behavior in a setting or stimulus situation that is different from the instructional setting.

response generalization

The extent to which a learner emits untrained responses that are functionally equivalent to the trained target behavior.

The general rule for teaching sufficient stimulus examples is

The more types of stimulus used during instruction, the more likely the learner will respond correctly to untrained examples or situations.

Stimulus generalization

When an antecedent stimulus has a history of evoking a response that has been reinforced in its presence, the same type of behavior tends to be evoked by stimuli that share similar physical properties with the controlling antecedent stimulus.

Benefits of developing the planning lists include

a better understanding of the scope of the teaching task and an opportunity to prioritize the most important behavior changes and settings for direct instruction.

General case analysis is

a systematic method for selecting teaching examples that represent the full range of stimulus variations and response requirements in the generalization setting.

Aubrey's stick-with-a-task-until-it's-finished behavior was poor. She frequently submitted incomplete schoolwork as "finished. Aubrey's teacher implemented a successful intervention that helped Aubrey to complete each part of multiple-part, in-school assignments before submitting them and beginning another activity. At the parent conference at the end of the quarter, she proudly reported Aubrey's new success, and terminated the special intervention. Now, however, three weeks after the quarter ended, most of Aubrey's work has deteriorated to her past performance. What strategy would you, as a behavior analyst have recommended to the teacher at the parent conference?

a. Develop a public posting procedure in the classroom, where each student's name was checked off as soon as they turned in completed schoolwork. Once all were turned in the students would be prompted to stand up and give each other applause. b. Gradually increase the delay of reinforcement to every other period, and then every other day, until Aubrey's work completion is reinforced at the end of the week, with a note sent home to her parents. & c. Train Aubrey to recruit the teacher's immediate social reinforcement when she finished her schoolwork.

Generalization across subjects refers to changes in the behavior of other people not directly treated by an intervention as a function of treatment contingencies applied to certain people. Another term for this process is ___________.

a. Spillover effect b. Ripple effect & d. Vicarious reinforcement

One strategy for promoting generalized behavior change is to teach the full range of relevant stimulus conditions and response requirements. To do this, a practitioner may:

a. Utilize general case analysis & c. Teach sufficient stimulus examples d. Teach sufficient response examples

What is called "stimulus generalization" in most studies of the spread of learned responding as environmental circumstances change should probably actually be called "failure of stimulus generalization," given that the responding typically drops off precipitously as the stimuli change. For this reason, the drop off in responding is aptly described as a "generalization gradient." Because of this, behavior analysts should _________________________.

a. not assume that generalization will occur "for free" after initial response training c. select target behaviors that will come in contact with natural contingencies of reinforcement & d. deliberately build training for generalization into their teaching programs

Not all behavior traps are desirable. The layperson's everyday term for an undesirable behavior trap is ____________________.

a. vicious circle b. addiction & c. being hooked

After Abdullahi was systematically taught a desirable, adaptive form of behavior that neither he nor other children at the center had exhibited before, the other children also began to perform the behavior. When we saw this happen we concluded that __________________ has occurred.

across subject generalization

In their seminal review paper, "An Implicit Technology of Generalization," Stokes and Baer (1977) stressed three facets of generalized behavior change. They were:

across time, settings, and behaviors

Some interventions yield significant and widespread generalized effects

across time, settings, and other behaviors; others produce circumscribed changes in behavior with limited endurance and spread.

A contrived contingency is

any contingency of reinforcement (or punishment) designed and implemented by a behavior analyst to achieve the acquisition, maintenance, and/or generalization of a targeted behavior change.

A naturally existing contingency is

any contingency of reinforcement (or punishment) that operates independent of the behavior analyst's or practitioner's efforts, including socially mediated contingencies contrived by other people and already in effect in the relevant setting.

A generalization probe is

any measurement of a learner's performance of a target behavior in a setting and/or stimulus situation in which direct training has not been provided

A generalization setting is

any place or stimulus situation that differs from the instructional setting in some meaningful way and in which performance of the target behavior is desired.

A method for programming generalization from the instructional setting to the generalization setting is called __________.

b. Programming common stimuli Incorrect c. Teaching self-management skills d. Teaching loosely

An important behavior change should not go unmade

because complete withdrawal of the intervention required to achieve it may never be possible. Some level of intervention may always be required to maintain certain behaviors, in which case attempts must be made to continue necessary programming.

A newly learned behavior may fail to contact an existing contingency of reinforcement

because it has not been taught well enough. The solution for this kind of generalization problem is to teach the learner to emit the target behavior at the rate,accuracy, topography, latency, duration, and/or magnitude required by the naturally occurring contingencies of reinforcement.

The use of intermittent schedules of reinforcement and delayed rewards

can create indiscriminable contingencies, which promote generalized responding by making it difficult for the learner to discriminate whether the next response will produce reinforcement.

The generalization map is a

conceptual framework for combining and categorizing the various types of generalized behavior change (Drabman, Hammer, & Rosenbaum, 1979).

A ________________________ is any contingency of reinforcement (or punishment) designed and implemented by a behavior analyst to achieve the acquisition, maintenance, and/or generalization of a targeted behavior change.

contrived contingency

The instructional setting is the

environment where instruction occurs and encompasses all aspects of the environment, planned or unplanned,that may influence the learner's acquisition and generalization of the target behavior.

In _________________________, the target behavior comes under the restricted control of an irrelevant antecedent stimulus.

faulty stimulus control

Behavior analysts often emphasize ________________ as an important goal of treatment programs, by which they mean developing the target response to the point that it occurs at a more-than-sufficiently high rate and/or immediate (brief) latency, in order to guarantee contact with the natural contingencies of reinforcement that are available to it.

fluency

A ____________________ is any place or stimulus situation that differs in some meaningful way from the instructional setting and in which performance of the target behavior is desired.

generalization setting

If a trained behavior occurs at other times or in other places without it having to be retrained completely at those times or in those places, or if functionally related behaviors occur that were not taught directly, then _____________ behavior change has occurred.

generalized

Negative, or "don't do it," teaching examples

help learners identify stimulus situations in which the target behavior should not be performed.

Having the learner practice a variety of response topographies

helps ensure the acquisition of desired response forms and promotes response generalization. Often called multiple exemplar training, this tactic typically incorporates numerous stimulus examples and response variations.

Planning for generalization includes

identifying all the desired behavior changes and all the environments in which the learner should emit the target behavior(s) after direct training has ceased.

Generalized behavior change has taken place

if trained behavior occurs at other times or in other places without having to be retrained completely in those time or places, or if functionally related behaviors occur that were not taught directly.

Programming common stimuli means

including in the instructional setting stimulus features typically found in the generalization setting. Practitioners can identify possible stimuli to make common by direct observation in the generalization setting and by asking people who are familiar with the generalization setting.

One way to wake up an existing but inoperative contingency of reinforcement

is to ask key people in the generalization setting to attend to and praise the learner's performance of the target behavior.

One tactic for mediating generalization

is to bring the target behavior under the control of a contrived stimulus in the instructional setting that will reliably prompt or aid the learner's performance of the target behavior in the generalization setting.

One tactic for promoting response generalization

is to reinforce response variability. On a lag reinforcement schedule, reinforcement is contingent on a response being different in some defined way from the previous response (a Lag 1 schedule) or a specified number of previo

The first step in promoting generalized behavior changes

is to select target behaviors that will meet naturally existing contingencies of reinforcement.

Another tactic for waking up a natural contingency of reinforcement

is to teach the learner how to recruit reinforcement in the generalization setting

The simplest and least expensive tactic for promoting generalized behavior change

is to tell the learner about the usefulness of generalization and then instruct him to do so.

With most successful behavior change programs

it is impossible, impractical, or undesirable to continue the intervention indefinitely.

The strategy of training to generalize is

predicated on treating "to generalize" as an operant response class that, like any other operant, is selected and maintained by contingencies of reinforcement.

Coaches hold scrimmages; music teachers hold mock auditions; and theater directors hold dress rehearsals to prepare their athletes, musicians, or actors to perform important skills in settings that include the sights, sounds, materials, people, and procedures that simulate as closely as possible those in the "real world." These practices are examples of what principle of applied behavior analysis?

programming common stimuli

Teaching loosely

randomly varying noncritical aspects of the instructional setting within and across teaching sessions— (a) reduces the likelihood that a single or small group of noncritical stimuli will acquire exclusive control over the target behavior and (b) makes it less likely that the learner's performance will be impeded or "thrown off" by the presence of a "strange" stimulus in the generalization setting.

Avery was taught to remove weeds by swinging a long weed-removal tool. Although she was never taught or asked to do so, sometimes Avery will now remove weeds by digging with the tool, or even with her bare hands. This illustrates ___________________________.

response generalization

The extent to which a learner emits untrained responses that are functionally equivalent to the trained target behavior is called: _____________.

response generalization

When programmed contingencies have been applied to some forms of behavior, and then other forms of behavior appear as a function of those same contingencies, this is called _______________.

response generalization

A desirable, adaptive form of behavior that Sofia did not exhibit until it was systematically taught in a specific training situation will, ideally, persist over time without the need for extensive additional training support. When we see this happen we say that ________________________ has occurred.

response maintenance

When a behavior change produced in the clinic has also occurred at least once in the generalization environment, but then ceases to occur, a lack of _______________________ is evident.

response maintenance

A desirable, adaptive form of behavior that Matthew did not exhibit until it was systematically taught in a specific training situation will, ideally, also occur in other appropriate circumstances without the need for extensive additional training support. When we see this happen we say that ________________________ has occurred.

setting/situation generalization

If some component(s), but not all, of the original training program are still needed to result in meaningful behavior change in the novel environment, then ________________ can be claimed, provided the components found necessary in the novel environment were insufficient in and of themselves to bring about the behavior change in the original training.

setting/situation generalization

If the complete intervention program is still required to produce behavior change in the novel environment, then no __________________ can be claimed.

setting/situation generalization

The extent to which a learner performs the target behavior in a setting or stimulus situation that is different from the instructional setting is called _________________.

setting/situation generalization

When a behavior change produced in the clinic is not also observed in the generalization environment, a lack of _______________________ is evident.

setting/situation generalization

We may correctly claim that generalization has occurred when:

special training procedures are no longer necessary for the target response to occur appropriately in situations that are distinctively different from the training situation.

The strategy of teaching sufficient examples requires

teaching a subset of all of the possible stimulus and response examples and then assessing the learner's performance on untrained examples.

The tactic for promoting setting/situation generalization called _____________ involves teaching the learner to respond correctly to more than one example of antecedent stimulus conditions and probing for generalization to untaught stimulus examples.

teaching sufficient examples

Teaching sufficient stimulus examples involves

teaching the learner to respond correctly to more than one example of an antecedent stimulus and probing for generalization to untaught stimulus examples.

One of the more seemingly ironic and surprising findings in studies of operant conditioning, in both the Experimental Analysis of Behavior and Applied Behavior Analysis, is that ____________________________.

the behavior of doing something "new," "different," or "creative" is an operant response class (i.e., learners can learn to vary their behavior if response variability is what is reinforced).

Response maintenance refers to

the extent to which a learner continues to perform a behavior after a portion or all of the intervention responsible for the behavior's initial appearance in the learner's repertoire has been terminated.

Setting/situation generalization refers to

the extent to which a learner emits the target behavior in settings or situations that are different from the instructional setting.

Response generalization refers to

the extent to which a learner emits untrained responses that are functionally equivalent to the trained response.

As a general rule

the more examples the practitioner uses during instruction,the more likely the learner will be to respond correctly to untrained examples or situations

The greater the similarity between the instructional setting and the generalization setting

the more likely the target behavior will be emitted in the generalization setting

Research in Applied Behavior Analysis supports the general finding from the Experimental Analysis of Behavior that intermittent schedules of reinforcement generate greater resistance to extinction than does continuous reinforcement. In general, it has been found that the ______________ the schedule of reinforcement in the instructional setting, the _______________ the response maintenance in the generalization setting.

thinner; greater

Researchers have developed and advanced

what Stokes and Baer (1977) called an "implicit technology of generalization"into an increasingly explicit and effective set of methods for promoting generalized behavior change.