Basic Concepts and Principles

Functional Behavioral Assessment, Diagnosis, and Treatment: A Complete System for Education and Mental Health Settings - Ennio Cipani PhD 2018

Basic Concepts and Principles


✵Students will be able to describe and provide an illustration of a direct access function

✵Students will be able to describe and provide an illustration of a direct escape function

✵Students will be able to describe and provide an illustration of a socially mediated access function

✵Students will be able to describe and provide an illustration of a socially mediated escape function

✵Students will be able to identify an establishing operation for a number of access functions as one of a state of deprivation

✵Students will be able to identify an establishing operation for a number of escape functions as one of a presenting state of aversion

✵Students will be able to discuss what a contrived contingency is and how it can “override” existing motivational variables

Chapter 1 Behavior Analysis Certification Board (BACB) Task List

4th edition

5th edition

✵FK-01 Lawfulness of behavior

✵G-04 Explain behavioral concepts using nontechnical language

✵E-01 Use interventions based on manipulation of antecedents, such as motivating operations and discriminative stimuli

✵B-4 Define and provide examples of positive and negative reinforcement contingencies.

✵B-5 Define and provide examples of schedules of reinforcement.

✵B-6 Define and provide examples of positive and negative punishment contingencies.

✵B-7 Define and provide examples of automatic and socially mediated contingencies.


Why does he do that? This is the age-old question that people ask when they see a child throw a fit in public. Why does he behave like that? To date, an often-cited explanation of such undesirable behavior involves a hypothesis about the brain’s development in the child who is “afflicted” with such behavior. The underpinning of the undesirable tantrum behavior is hypothesized to be the result of some abnormality or underdevelopment of some part or parts of the brain. As further evidence of brain involvement, in some cases, such behavior along with other behaviors form the basis of a mental disorder. Following is an excerpt from a hypothetical lecture in a child psychology class.


Dr. Trait, I have a question. Why do some children have tantrums that are clearly inappropriate for their age?

Dr. Trait:

The child throws the tantrum because he or she is immature for his or her age; the brain has not fully developed. Once the brain matures—particularly the frontal lobe, which is responsible for executive functioning—the child will not respond to social situations in that manner. Until that point, we can expect this child to continue behaving in such a fashion because of his or her inability to process events adequately. Teenagers have a similar problem with brain immaturity when they behave impulsively. Their brain is not like the adult brain; hence, they too cannot be fully responsible for their impulsivity.

Variations and extensions of this immature brain explanation exist. The following is a continuation of this conversation in a class in developmental psychology, with a slightly different explanation.


In Dr. Trait’s class, we were told that children who have severe tantrums that are clearly inappropriate for their age do so because their brain is not fully developed. Is there any experimental cause-and-effect evidence for such an assertion?

Dr. Stager:

Well, I believe there is more to it than just the brain’s development, although I would concur that neurological issues are part of the problem. Children behave in a certain manner because they have not proceeded through certain invariant developmental stages. I would say that these children have not progressed past the egocentric stage. Of course, once the brain has developed, it is more likely that these children will interpret the actions of others as reasonable and not view everything from a “me first” perspective. When this happens, these children will not react in such a manner, but will respond to conflict in a more age-appropriate way.

Suppose we believe that the child throws a tantrum because the brain is not yet fully developed. What are the ramifications of dealing with such behavior when the supposed cause is brain malfunction? Do we wait until the brain becomes more fully developed? For clients who have continued such “immature” behaviors throughout their adolescence and into adulthood, do we still continue to wait? What can be done in the interim to reduce the child’s tantrums and develop a more acceptable manner of dealing with his or her social environment?

What is wrong with these interpretations about tantrum behavior? The role of the environmental response to such behavior is trivialized. If the brain has not developed, apparently what people do in response to the child’s behavior, whatever the form, is insignificant and, therefore, irrelevant. One can only hope that the child’s brain becomes more fully developed. We believe there is a better conceptualization of why tantrum behavior occurs.




Instead of saying that the child throws a tantrum because he or she is immature, we would possibly ascribe such an incident to the purpose or function such tantrum behavior serves in that child’s environment. That conceptualization would generate an examination of observable events in the social environment. In the case of a child’s tantrum behavior, one would examine what the social environment does when the child has a fit in public. What is the antecedent context for such tantrum behavior? How does the social environment react to these tantrums in the short and long term? This examination of temporally ordered environmental events can reveal the purpose of this behavior in this context.

This approach is termed a functional behavior-analytic approach to understanding human behavior (Bailey & Pyles, 1989; Cipani, 1990; Cipani & Trotter, 1990; Iwata, Vollmer, & Zarcone, 1990; LaVigna, Willis, & Donnellan, 1989; Lennox & Miltenberger, 1989). In a functional behavior-analytic approach, all operant behavior is viewed as serving an environmental function, either to access something or to terminate/avoid something (not withstanding genetic influences for some behaviors). Although other psychological explanations invoke hypothesized traits or developmental stages to explain behavior, a functional behavior-analytic viewpoint examines the roles of both the social and physical context—it deals with events that are both observable and measurable.

A hypothetical 7-year-old child named Oskar, diagnosed with oppositional defiant disorder (see the Diagnostic and Statistical Manual of Mental Disorders [5th ed.]; DSM-5), has extreme oppositional behavior with occasional aggressive behavior. Explaining such behavior as a result of his disorder is sufficient for many mental health professionals. When asked why this child has such episodes, their response would be: “It is a symptom of his underlying disorder—oppositional defiant disorder. He behaves in an extreme fashion because he has this disorder.” As you can see, this is a trial lawyer’s dream. People do things because they have a disorder. If they have this disorder, they cannot help it.

Whenever this form of behavioral episode occurs, it is the disorder that made him do it and he will engage in such behavior from time to time. It further presumes that such a behavior will occur irrespective of context and consequences. The child engaged in the defiant and/or aggressive behavior because of his malfunctioning brain. Such brain malfunctions are not predicated on the presence of any environmental context. One never knows when the neurons misfire; when they misfire, defiant and aggressive behavior results.

In contrast, a functional behavior-analytic view would explain such behavior more from the social (and sometimes physical) context of the behavior. One would examine Oskar’s history of oppositional and aggressive behavior and how such behavior alters his existing social environment. An understanding of why the behavior occurs is accomplished through an analysis of the behavior’s ability to either produce desired events or terminate undesirable events.

For example, we may discover that Oskar often engages in such behavior when he comes home from school. Oskar’s mother wants him to stay in the house for a while and either do his homework or finish cleaning his room. Oskar, of course, wants to go outside and play with his friends. He sometimes will complain and whine. His mother will respond to such complaining with the following retort: “You need to finish your homework. How do you expect to pass third grade? Once you are done with your homework, then you can go outside.” This parental response to his behavior incurs more arguing and defiance from him, with refutations for each assertion by his mother. When Oskar sees that this verbal battle with his mother is not helping his cause (i.e., getting to go outside), he tries another tact. He states, “I’m going to leave and you can’t stop me.” When he begins to exit the house, she grabs him. At this point, he yells at her, calls her names, and hits her. After a struggle, Oskar pulls away and heads out the door. His mother, tired of fighting with her son, lets him go, complaining he is just like his father.


In 2016 many people believe that sending children with severe problem behaviors to counseling is the best method for changing these behaviors. Can anyone (through counseling) convince Oskar that aggressive behavior is not in his best interest? What is in a child’s best short-term interest when he is placed in time-out? It is getting out of time-out. As a reader of these materials, do you believe that talking to Oskar once or twice a week will convince him not to behave aggressively toward his mother when he is told he cannot go outside?

How will Oskar’s behavior change? Ultimately, it is up to the care providers and parents to change their own behavior in order to change the child’s behavior. If Oskar’s mother continues to handle this situation in the same manner, I cannot see where anyone can convince him to respect his mother’s wishes and do his homework before going outside. In a functional behavioral-analytic approach, the presenting problem is not just with the child. It is also with the way the child’s environment responds to his behavior.

You change child behavior by changing the behavior of the adults who deal with that child. Pure and simple!

With this information, what is a more plausible explanation for Oskar’s behavior during these circumstances? Does he act out because he is disordered? Or does the explanation lie in an understanding of how such a behavior impacts his environment? Does less intense arguing with his mother result in him going outside? Or does he get to go outside when he becomes assertive (walking to the door) and combative (when he hits his mother as she tries to get him to stay inside)? What is the best explanation for Oskar’s aggressive behavior in the afternoon? He does it because it “works” for him when he wants to go outside, and other behaviors such as complaining are less effective. The function of such combative and aggressive behavior provides an explanation of why it occurs. Such an explanation can be experimentally demonstrated, as you will find out in the coming pages of this book.


In a functional behavior-analytic approach, behavior is viewed as functional (i.e., purpose­ful) for certain antecedent contexts because of the contingency or contingencies involved. Simply stated, behavior always makes sense, from the standpoint of the person exhibiting such behavior. A contingency is the temporal relationship between behavior and a consequence. It is often stated as an “if, then” rule. If you get an A on your quiz, then I will take you out for ice cream. If you stick your hand in the door and it closes on your hand, then you will experience pain and yell loudly. Such social and environmental consequences influence whether the behavior that produces them will become more or less probable in the same or similar context.


Why does he do that!

For example, the manner in which Oskar’s mother responds to his mild complaining behavior makes that behavior ineffective. If Oskar wants to go out, and he complains, it seems that such behavior is not instrumental in getting him outside in the immediate future. Therefore, mild pleading and complaining behavior becomes less likely in subsequent afternoons when Oskar wants to go outside. The current arrangement between complaining and not getting to go outside makes such forms of complaining a less viable alternative in these circumstances.

However, the story for verbal defiance and physical aggression is quite the opposite. These behaviors, including defying her wish to stay inside by physically leaving, are functional in accessing the desired event. The next day, when Oskar’s request to go outside is denied, what would he likely do? You guessed it. If Oskar’s mother continues to respond to her son’s defiance and aggression in the same manner, such behaviors become functional in that context. If this relationship between aggressive behavior and going outside becomes strengthened, under certain motivating and antecedent conditions, then such a relationship defines a maintaining contingency or behavioral function. There are two major types of behavioral functions for problem (or other) 'margin-top:12.0pt;margin-right:0cm;margin-bottom: 3.6pt;margin-left:0cm;line-height:normal'>BEHAVIORAL FUNCTIONS INVOLVING POSITIVE REINFORCEMENT (ACCESS FUNCTIONS)

Positive reinforcement functions involve behaviors that produce an environmental event that subsequently increases the likelihood of that behavior under the same or similar antecedent motivating conditions. In other words, the operation of positive reinforcement involves a behavior that produces an event (activity, object) that subsequently strengthens the occurrence of that behavior in the future (under certain motivational contexts). As a result of such a behavior-contingency relationship, the behavior becomes functional under given motivating conditions (to be discussed in greater detail later in this chapter). For purposes of the function-based diagnostic system delineated in Chapter 3, positive reinforcement functions are termed access functions, that is, these behaviors access preferred items, events, or activities.

What are some hypothetical illustrations of access (positive reinforcement) functions? Milton, an inpatient client with schizophrenia, is reported to frequently pinch other clients. Is this pinching behavior the result of his schizophrenia? Is it the result of his inability to control his impulses? I believe neither explanation would serve a useful or parsimonious purpose. After some duration and frequency (to be explained in Chapter 2), when Milton pinches others we find that staff take him for a walk. Their rationale for such a response to his pinching is that they want to get him away from other clients. They report that Milton seems to be less anxious when he is on his walk and that the walk calms him down. Facility staff, thereby, interpret their use of a walk as an “anxiety-reductive procedure” and believe that this practice is clinically sound. However, what escapes them is the long-term result of this reliable contingency between what Milton does and what they do. Milton has learned how to get a walk with staff—pinch someone! We would say that pinching other people is a functional behavior when Milton desires a walk. Unfortunately, other more appropriate behaviors do not appear to be more effective (i.e., less functional) in getting a walk. The behavior of pinching others subsequently increases to a level that constitutes a major problem, which now jeopardizes Milton’s ability to remain in the current inpatient unit.

Bea, a residential adult female client, throws a tantrum (consisting of screaming and slapping herself) at certain times during the day. Bea’s tantrum behaviors are reported to consist of yelling, hitting or slapping herself, calling staff profane names, claiming she was placed in this facility by the Mafia, and making verbal threatening statements to staff and other residents. When she engages in such behavior for a period of time, one can observe staff members give her something to eat. These staff members interpret her behavior, after some duration, as a sign that she is hungry. Of course, feeding her certainly stops the threats made to them and others. Although feeding Bea may produce the desired short-term result, it creates long-term disaster. Such tantrum behaviors become more probable for Bea when she is relatively hungry (or at least when she wants certain food items). You might conclude that Bea’s tantrum behavior is maintained because it is capable of acquiring food when she is hungry. Bea may also learn to engage in the same type of tantrum behavior when she wants her CD player and is told she has to wait until after dinner for it. If such behavior reliably results in Bea getting the CD player before dinner, then tantrum behavior also becomes functional under the conditions of Bea desiring the CD player. We would expect an increase in the frequency of tantrum behavior across the next few weeks as it becomes strengthened as a functional behavior when she desires the CD player.

A 4-year-old child named Elvira, diagnosed with autism, will engage in screaming and hitting herself multiple times during the day. Many people will explain such behavior by referring to her developmental disorder. They will proclaim, “Elvira throws a tantrum because she has autism. Her autism is the cause of this behavior.” But is this really a good explanation? Can one predict that all children with autism will engage in such behaviors independent of social context? Does such behavior differentiate children with autism from children with other developmental and/or mental disorders (i.e., only children with autism hit themselves)? If such behavior is caused by autism, what options remain for the successful treatment of such behaviors, ameliorating or eliminating autism? Although eliminating or curing autism is certainly a laudable goal, is it reasonable to suspect that this will occur in time to help Elvira before she enters school? Before she becomes an adult?

A more productive and functional approach focuses on the environmental function (i.e., outcome) of problem behaviors. For Elvira, screaming might reliably access parental attention, or hugging, under conditions in which she desires such activities or events. When Elvira desires parental attention because it has been some time, tantrum behavior becomes more probable. The production of attention for some level and duration of tantrum behavior then maintains Elvira’s behavior as functional in accessing attention.


Although many people are familiar with positive reinforcement, negative reinforcement is often misunderstood (Cipani, 1995; Cipani & Spooner, 1997; Iwata, 1987). An understanding of negative reinforcement operations is critical to the design of effective treatments, particularly if you serve individuals who often engage in behavior problems during task demands, compliance situations, instructional conditions, or chores and work tasks.

In negative reinforcement, the effect of the behavior is to terminate the existence of, or postpone (for some time), the presentation of a stimulus condition or event. Such an event is commonly referred to as aversive or unpleasant (relative to the individual). Realize that the term “aversive” is relative. What is aversive to one person may not be to another; what is aversive today may be less aversive next week. The subsequent effect of a negative reinforcement contingency on behavior is one of increasing its probability under the same or similar conditions in the future. Negative reinforcement functions are called escape functions (includes avoidance functions in this terminology).

Examples of escape functions with problem behavior can be used with the previous hypothetical cases by altering the behavioral effect of the problem behavior. For example, Bea is asked to clean up her room. She will often refuse such an initial request. When staff persons at the facility warn her that she will not get to watch television that night, she screams and yells at them. After an intense episode, Bea sometimes gets put in time-out and loses her television privileges. However, with certain staff persons, if she promises not to raise a commotion, the staff person will clean up Bea’s room for her so she can watch television. Such a behavioral effect subsequently increases the probability of Bea screaming in those conditions (or similar conditions) in the future.

Can pinching people occur for a different reason than wanting a walk? Let us say that Milton also pinches people when he is asked to go to group therapy (which he finds aversive). What is the “reason” or function for pinching when confronted with this request? Suppose the following events play out when Milton pinches a staff person when it is time to go to group therapy. When he pinches a staff person, someone decides that he should be taught that this is wrong. Milton should be put in time-out to teach him that pinching is inappropriate. The staff keep him in time-out until he is quiet and is able to say he is sorry for pinching the staff person. His apology does not occur readily and the minutes go by. By the time Milton is allowed to leave time-out, his group’s therapy hour is almost over. He therefore is able to shorten his participation in group therapy by going to time-out and being adamant about his right to pinch people. Unfortunately, Milton does not learn the “lesson” that the time-out was intended to teach, and instead it intensifies his pinching behavior before group therapy. One staff person remarks that it almost seems as if Milton pinches to avoid going to therapy, but others quickly dismiss such a contention.

As a result of this imposed time-out consequence for pinching, Milton is pinching more often when it is time to go to group therapy. Consequently, he misses most therapy sessions in a given week. Can you see why he is pinching when it is time to go to therapy? Pinching avoids an activity Milton dreads—going to group therapy. As a side note, it might be interesting in this case to find out why Milton does not like to go to group therapy (i.e., what does he wish to avoid) in order to solve this behavior problem in the long term. Beyond that, one might question why he should go to group therapy. If it was to help him uncover the reasons for his pinching mode of interaction, we could now dispense with such a requirement (given its obvious lack of effectiveness).

Another example illustrating a negative reinforcement function is the self-abusive behavior of a child with autism spectrum disorder (ASD). As a general note, very often in classroom situations, problem behaviors such as self-abuse can often function to avoid or terminate instruction. Hence, under such conditions certain behaviors become very “adaptive.” This child hits his or her head with an open hand or closed fist, and such behavior often seems to occur during group instruction. Consequently, the teacher will often stop instruction or remove the child to deal with the self-abuse. Self-abuse becomes functional in lessening or avoiding such a context.

Unfortunately, self-abuse is often unintentionally exacerbated. If staff reason that the child hits him- or herself for their attention, they might conclude that ignoring this behavior will make it decrease and eventually disappear. When the function of this child’s self-abuse is misdiagnosed as attention seeking, ignoring minor forms leads to more disastrous forms of behavior. As a result of staff ignoring minor forms of self-abuse, the severity of the self-injury intensifies. The more severe head hits do result in the instruction being halted for a period of time. Consequently, more severe self-abuse becomes prevalent, as a result of its ability to effectively terminate instruction. Perhaps, at the beginning of the year, the teaching personnel reported to the individual education plan (IEP) team that this child does engage in self-abuse, but they can handle it. Of course, this was under the presumption that their ignoring strategy would work. Now, at mid-year, this child may no longer be suitable for this classroom because the self-injury has resulted in a broken nose and gashes on the forehead. The child may now require a placement where a more intensive behavioral approach is available. A minimal problem has developed into a major threat and hazard to this child’s health and educational welfare.



Assignment: A Behavior-analytic model for explaining childhood aggression (see hot link in student portal;

After reviewing the lecture entitled, “A behavior-analytic model of childhood aggression,” submit a paper addressing the following questions in order:

✵What is the behavior-analytic explanation for the possible etiology or cause of child aggression? How is that different from a traditional explanation? (See “Why Does He Do That?” section at beginning of chapter.)

✵What is the behavior-analytic explanation for the continued display of child aggressive behavior over time?

✵How do contrived punishment contingencies (adult-imposed) for aggressive behavior affect its function?

✵Does watching or observing aggression explain the latter (i.e., how young children who are aggressive engage in such behavior when they are adolescents)?


Table 1.1 provides more examples of behaviors that illustrate access (positive reinforcement) functions. Note that in all instances the effect of the reinforcement contingency is one that strengthens the behavior that produces the desired event. Table 1.2 illustrates how behaviors that previously accessed desired events now function to terminate aversive events. We conclude that the form of the client’s problem behavior does not usually give a clue as to behavioral function. Hence, a diagnostic system that focuses exclusively on symptoms to differentiate clients misses the mark.

Comparing Tables 1.1 and 1.2 provides an illustration of how the same behavior can produce different environmental effects, that is, consequences that maintain such behaviors. Note that the behavior (although the same form of response) produces two different outcomes. In the third example in Table 1.1, the behavior of Bobby hitting his brother resulted in his mother intervening and giving him the toy that his brother had. If hitting reliably results in mom’s mediation of the conflict via giving in to Bobby’s desire for some object or item, hitting becomes more probable when Bobby wants something his brother has.



Contingency Produced

Effect of Contingency on Behavior

Child cries

Gets cookie

Increases likelihood of crying when child wants cookie in the future

Man on inpatient unit stomps foot on floor, kicks wall

Gets nurse to come over and give social attention, engage him in pleasant conversation

Increases likelihood of such behaviors when man wants to socially interact with that nurse in the future

Child hits brother

Mom tells brother, “Give Bobby the toy; he is not as mature as you are” and gets toy

Increases likelihood of aggression when child wants some toy or item his brother has

Student says, “This is not fair, I never get a turn”

Teacher gives child a turn on tetherball

Increases likelihood of such demand/tantrum behaviors when child wants to get tetherball or other activity and does not want to wait for peers to give him or her access



Contingency Produced

Effect of Contingency on Behavior

Child cries

Released from “room time”

Crying is more likely when child is placed in room for discipline

Man on inpatient unit stomps foot on floor kicks wall

Gets nurse to leave him alone for a while instead of taking his medication

Such aggressive behaviors are more probable when nurses are trying to get this man to do something he does not desire

Child hits brother

Brother leaves room

Aggressive behavior becomes more probable when this child wants to be left alone

Student screams, “This is not fair, I always get more work”

Teacher talks to student, agrees to reduce assignment by half

Increases likelihood of such demand/tantrum behaviors when child wants to do less (or no) work

In contrast, in Table 1.2 note how hitting serves to remove an aversive stimulus for the same topography (form) of behavior. Bobby hitting his brother makes his brother leave the room. Hence, whenever, Bobby wants to be alone without his brother in the room, to what behavior will he now resort? Hitting! To summarize, hitting that occurs under the context of his brother playing with a toy that he wants functions to get the toy via mediation of such behavior by his mother. Hitting that occurs when Bobby wants to be alone results in the removal of the unwanted party via the brother leaving the room. This is the same behavior with two different functions.

Compliance situations can be a context where problem behaviors arise. There are two types of commands (Neef, Shafer, Egel, Cataldo, & Parrish 1983). One type involves a parent issuing a request or directive toward a child to engage in some requested behavior (called a “do” command). The contrasting type of command involves a request to desist a behavior (called a “don’t” command). Examples of compliance situations involving a do command are (a) “pick up your trash and place it in the trash can,” (b) “open the door to the laundry room,” (c) “put your sneakers on.” Examples of don’t commands are (a) “stop running through the hallway,” (b) “stop yelling,” (c) “do not throw the ball against the house again.” When oppositional behavior occurs in compliance situations (i.e., the child refuses to follow through with the request), it can be analyzed in terms of function (Cipani, 1998).

In some cases, noncompliance takes an innocuous form, such as the individual simply not attending to the person issuing the command. Such a lack of response is maintained by negative reinforcement, that is, avoidance function. A command is issued and the child or client does not respond but rather continues engaging in the ongoing activity. The form of noncompliance at this moment is simply nonresponding. If the adult making these requests often forgets about what task was requested as a result of inactivity on the part of the client, one can see that nonattending avoids engagement with task.

With some children or clients, their opposition to the request or command entails more than just ignoring the request. In many home and school compliance situations, the child is “forced” to engage in other forms of protest to successfully avoid engaging in the undesired task. Let us say that the adult fails to leave them alone when they simply opt out of complying with the request peacefully. The response of the adult to the child’s deaf ear approach (i.e., adult issues request again) does not provide escape from the compliance situation. When the adult responds with another request the child now retorts, “I’m not doing it!” Because simply ignoring the request did not work, maybe becoming insolent and disrespectful at the person making the request will force the adult to leave the child alone. The form of noncompliance can then become exacerbated, as mild forms of opposition do not have the effect desired (i.e., termination of the request). To illustrate this point, following is a hypothetical scenario with an adult client in a group home.

Staff member:

Mr. Smith, please pick up your dirty clothes from the floor and place them in the hamper.

Mr. Smith:

Leave me alone. I’m watching American Idol.

Staff member:

Mr. Smith, I need you to pick up your clothes. Someone may trip over them when they are in the middle of the day room.

Mr. Smith:

Then that would be their own stupid fault!

Staff member:

(Moves closer to Mr. Smith) Mr. Smith, would you like me to help you?

Mr. Smith:

I would like you to leave me alone! If you are so interested in my dirty clothes, you pick them up.

Staff member:

Please address me with respect.

Mr. Smith:

Quit ragging on me, you . . . .

Staff member:

Okay, Mr. Smith. I will get Raul and Robert to help me assist you.

Mr. Smith:

(Gets up and runs out of the day room with staff in tow. He hides in the bathroom and does not come out. After 25 minutes, he finally opens the door and is allowed to go back and watch television with the clothes now having been picked up by someone else.)

Note that in this scenario, simply protesting was not an effective method of being left alone. Such behavior resulted only in continued verbal requests on the part of the staff member. However, with continued requests and the threat of having several staff members help him get his clothes off the floor, Mr. Smith engages in more than just noncompliance. He runs out of the room and locks himself in the bathroom, which of course makes picking up his clothes very unlikely. Do you think Mr. Smith will conclude that the bathroom is a good place to escape from staff?


In some settings, clients engage in severe forms of self-injury, aggression to ­others, and property destruction. Many people take this as an example that these people have uncontrollable rages as a result of their disorder. “Obviously these people have no control over what they do. They are driven to destruction!

Take the case of the previously discussed child who gets sent to the time-out room and cries to get “a more lenient sentence” once there. He might be judged to be less “disordered” than another child who engages in property destruction when placed in time-out. Why does he cry and not become destructive? Although this child currently does not engage in property destruction, he may not be far from learning how to do such. Here is a scenario of how more severe tantrum behaviors can be shaped by the social environment.

Suppose his parent decides to ignore the tantrum behavior in the time-out area. He will “stick to his guns” and not allow his child to get out early. The child goes to time-out and begins his crying episode but to no avail. He cries louder, but that does not result in removal either. He then throws something at the wall. The parent then comes into the room to find out what has happened. Note the effect of the throwing behavior in the sequence of events. If crying does not work, throw something. That will bring in Dad, and possibly a reprieve!

With time, the parent ignores some of the throwing episodes until a shoe goes through the dry wall near the door. As you can see, with each new exacerbation of behavior, the parent “has to” attend to the child to stop any further escalation. Hence, months later the child is now destroying the room, as opposed to just crying when placed in time-out.


There are two ways to access positive reinforcers: directly and socially mediated. Escape behaviors can function to produce the termination of an aversive event in the same two ways: directly and socially mediated. Hence, there are four possible major functions (Table 1.3).

Direct Access (DA)1

With a direct access function (referred to in the behavior analysis literature as automatic positive reinforcement), the client’s behavior immediately produces access to the desired event or item (Cipani, 1990, 1994; Michael, 1982; Vargas, 1988). For example, an individual is hungry and goes to the refrigerator, opens the door, selects an apple, and eats it. This chain of behaviors directly produces the apple. We would not say that the individual exhibits those behaviors because of the attention someone gives to him or her, regardless of whether such attention is positive, negative, or neutral. Attention would be a tangential consequent event. Getting the apple is the desired reinforcer. This is an example of a direct access function; the sequence of behaviors produced the bite into the apple directly.


Two Access Functions

Two Escape Functions

Direct access (DA)

Direct escape (DE)

Socially mediated access (SMA)

Socially mediated escape (SME)

What are some other examples of direct access functions? Putting the key in the car and turning it (behavior) when one wants to get across town produces the desired result (car starting). Lying down on the bed (when one is tired) involves a chain of behaviors that directly produces the reinforcer (rest). It is important to note that these behaviors produce the reinforcer immediately and directly.

How many people do you know who sing in the car while driving to work? Or sing in the shower? What “motivates” this behavior? Does someone reinforce this behavior with attention? Probably not! Their behavior is maintained as a result of the direct environmental effect produced, that is, the sound (hopefully of a somewhat melodious nature). Particularly in the shower, the sound reverberation can be sensorial reinforcing. Such singing behavior in the absence of an audience is reinforced because it produces an immediate inherent pleasurable sensory outcome that is desired at that time.

Too often, personnel assume that the problem behavior is maintained because of what people do after the behavior (called social mediation). But some instances of problem behavior are not “driven” by the desire to obtain attention, but rather are the result of the effectiveness of such behavior in directly producing the reinforcer. Here is a great real-life example. A client in a residential facility for persons with developmental disabilities jumps out of a wheelchair with some frequency (J. S. Bailey, personal communication, 1989). On one of these jumps, he falls to the ground, and his head begins to bleed. Obviously, the behavioral treatment program needs some adjustment! The facility calls in a nationally recognized expert in applied behavior analysis (Dr. Jon Bailey). Staff claim that the client jumps out of his wheelchair because he loves the medical attention he receives upon getting hurt.

What is the natural result of jumping out of the wheelchair? On some occasions, it does seem to be contusions and abrasions to this person’s body. However, the one result that always occurs is being on the floor (or, conversely, being out of the wheelchair). Every time this person pushes himself out of the wheelchair, he gets the freedom to roam around on the floor. Could sitting in a wheelchair for sometimes 12 to 16 hours a day, every day, be a motivational context for desiring “out of wheelchair” time? Any person who has driven a long time in a car or flown on a transcontinental nonstop flight may be able to relate. The problem with this behavior is not what the client wants but the manner in which he seems to have to access it. If this client was unable to verbally communicate his desire to staff, it would seem plausible that he would take matters into his own hands.

Socially Mediated Access (SMA)

Other behaviors achieve their desired effect through the behavior being mediated by someone else (Cipani, 1990, 1994; Michael, 1982; Sundberg, 1983). These behaviors produce the desired positive reinforcer through the efforts and actions of someone else.

The following scenario utilizing the previous example can provide the contrast between direct access and socially mediated access functions. Previously, the individual wanting to eat an apple (the person is hungry) performed a chain of behaviors that directly produced the apple, hence a direct access function. The same behavioral effect can occur when the individual requests someone standing next to the refrigerator to hand him or her an apple. Note the different manner in which the apple is obtained. In the latter example, the requesting behavior is mediated by another person, and subsequently, the reinforcing event is produced.

Socially mediated access often occurs through some form of vocal request, but it need not be so. A celebrity gestures to his driver, who subsequently opens the door to the limousine for him. The gesture functioned in the same manner as a vocal request, “Henri, the door please.” A child at a residential facility comes to the dinner table, and staff provide him with his snack for the afternoon. Coming to the dinner table is interpreted as “he is hungry.” In some cases, the vocal request may not even appear to be a request. A client with schizophrenia mutters about people stealing her money. Subsequently, after meeting with the facility administrator, she gets a few dollars to spend on candy and soda in the vending machines. It is important to be cognizant of the role staff, care providers, and adults play in the maintenance of this type of problem behavior.

In the previous case involving a client jumping out of a wheelchair, other behaviors might also produce the same function through mediation of the behavior from staff. Instead of jumping out of the wheelchair, let us say the client learned to have toileting accidents in order to get out of the wheelchair (i.e., staff have to clean him up, thus taking him out of the wheelchair). One could derive an equivalence between this existing environmental outcome of a toileting accident with a vocal request, “Can I get out of this?’ being honored by getting out of the wheelchair. Note that the outcome is the same for both behaviors; getting out of the wheelchair. If a client is incapable of such a vocal request due to language impairments, then having accidents becomes functional under that scenario. With socially mediated access functions, people play a major role in the development and maintenance of target problem behaviors becoming functional.

Direct Escape (DE)

Behavior can also produce the direct termination of existing aversive environmental events, serving a direct escape function (Cipani, 1990, 1994). For example, an individual walks into a noisy room, finds the level of noise aversive, and subsequently walks out. Note that the removal of the aversive stimulus (i.e., the heightened noise level in the room) was terminated through a chain of behaviors ending in leaving the room. Walking out of the room produces direct escape. The behavior directly produced the removal of the aversive event involving the noise level. Closing the blinds when the sun is too bright (for you) directly terminates the aversive stimulus (i.e., bright sun light). Taking a shower involves a chain of behaviors that is highly probable for many of us under conditions of being hot and sweaty (after physical exertion) because it directly terminates that condition (being hot and sweaty). These are all examples of chains of behaviors that produce escape (or avoidance) of aversive stimulation in a direct manner. This function is referred to in the behavior analysis literature as automatic negative reinforcement.

Socially Mediated Escape (SME) Behaviors

Escape behaviors can often achieve their effect of removing or avoiding an aversive condition through the behavior of someone else (Cipani, 1995; Iwata, 1987; Iwata, Dorsey, Slifer, Bauman, & Richman, 1982). In the case of the noisy movie theater, an individual verbally protests to the manager of the facility, who then gets the crowd to quiet down. The verbal protest behavior exerted its desired effect through the behavior of another person—the manager. If the individual yells loudly, “Let’s have some quiet in here,” and the room quiets down, the desired result was produced through a vocal request. The result (cessation of noise) was produced through other people becoming quieter as a result of this behavior. Both of these examples involve a behavior that achieves its effect indirectly, through the behavior or actions of someone else. However, if this annoyed person simply leaves the theater, thus terminating the noise, such a behavior produced its effect directly when the chain of behaviors ended in leaving the theater.

Negative reinforcement effects via escape functions can also explain why care providers, parents, and staff respond to their child or client’s behavior in the manner they do. In conducting workshops, I have seen participants begin to realize how much of a role the client’s social environment plays in the rate of problem behavior. Invariably, someone will make the following comment: “Why doesn’t the parent [care providers/staff members, teachers, aides, etc.] see that they are enabling [now you would say maintaining] the child’s misbehaviors?”

Before ascribing a dim view of such people, realize that maintaining contingencies also explain the behavior of parents, care providers, and teachers as well! An analysis of a behavioral function is not just for explaining why clients do what they do. Recall the case of the child who cries when put in the time-out room. As delineated previously, crying in this circumstance is probable because it affects the length of time the child stays in the time-out area. When the child cries, getting out early is more likely. When the child does not cry, getting out early is less likely. That is an analysis of the child’s crying behavior in time-out. But what analysis fits the parent who reinforces (mistakenly) that crying behavior by letting the child out?

When the parent removes the child from the time-out, what environmental effect do you think that produces? Does the child exacerbate the crying upon being let out? No, in fact the opposite is true. The child stops (at some point) crying and whining. As you can see, the child’s behavior affects also the parent’s behavior. If you put the child in time-out, then you fill the room or house with crying. If you take the child out of time-out, the crying stops. What function explains the contingency that results in increasing a behavior that terminates an aversive event? Escape function! The parent’s response to the child’s crying is under control of the presence of the aversive event (crying). When the parent’s response results in the child stopping the tantrum, such a parental response becomes more likely in the future. The parent learns to terminate (escape) the aversive state of the child crying by terminating the time-out prematurely.

In fact, the parent can avoid the crying in the first place by not putting the child in time-out when misbehaving. Instead of placing the child in time-out, mild warnings are issued, but not often followed by time-out. Therefore, time-out becomes less frequent even though the parent “knows” that the child should go to time-out for the target behavior. The parent learns to avoid the aversive stimulus by not producing time-out as frequently as needed. Unfortunately, this does not help the long-term effectiveness of time-out in reducing the rate of the child’s target behaviors. This phenomenon explains why follow through on consequences by some personnel and parents is weak and inconsistent. It would be nice if children and clients made it easy for us to administer consequences for behavior, but unfortunately, they do not.


Can you describe a hypothetical example illustrating an escape function for an individual’s self-injurious behavior? What would be the likely result of a self-injury incident? When would such a behavior be likely? What would the scenario look like if the self-injury were maintained by access to preferred items?

Simulaton Exercise: Type of access or escape functions

This simulation is designed to give the students practice with role-playing examples of the four major functions. Start the simulation with access functions.

With another person (role respondent), identify some item as a desired item (e.g., a pen, coffee cup). In each trial, which starts with a cue from you that it has begun, engage in either a behavior that directly accesses the item (goes over and gets it) or a behavior that gets the item via the role respondent (e.g., request, undesirable behavior such as whining, vociferous demanding, but leave out any profanity—no lawsuits wanted). Remember the socially mediated behavior need not be vocal; the requirement is simply that the role respondent acts on such with the retrieval of the desired item. The class then writes down which access function the scenario portrayed. Repeat several times to obtain practice effect.

The same can be done for escape functions except the aversive stimulus is presented (e.g., role respondent can grab your arm/hand, begin talking to himself or herself, stand close to you, present you with an assumed difficult academic task; just make sure such behavior does not create a “real” aversive event to the person—no lawsuits wanted). You then engage in a behavior that directly terminates such an event (leaving the area would be the most typical) or a socially mediated behavior such as vocal protest, intimidating looks to the role respondent, crying, and so on. The class then writes down which access function the scenario portrayed. Repeat several times to obtain practice effect.


The prior material delineated the characteristics of four different types of functions. All four functions have a relationship between operant behavior and a specific reinforcement effect it produces. But what makes some activity or event function as a positive reinforcer at a particular time? Or what makes the termination of some event of value at a given time? It is clear that what can function as a reinforcer for one person may not function as a reinforcer for another person. For the hypothetical client Milton, who pinches to get staff to take him for a walk, the walk would be termed the reinforcer. Is getting a walk a reinforcer for all inpatient clients with schizophrenia? Obviously not—“different strokes for different folks” applies here. But is a walk always a desired event for this hypothetical client? Again, obviously not. A walk becomes a desired event for Milton (i.e., reinforcer) when he has not had one for a while. Milton may want a walk around 9 a.m., but after having a walk, he does not want one for another 5 or 6 hours. Hence, pinching will cease for a period of time until getting a walk becomes more of a desire on the part of this client. Pinching as a means to get a walk becomes likely only under the conditions where Milton desires a walk.

Realize that the longer the time since Milton’s last walk, the greater the value of a walk. In layman’s terms, he greatly desires a walk. In day-to-day communication this explanation is not a problem. However, when we attempt to engage in a scientific examination and explanation of how Milton comes to “want” or “desire” a walk, we need a more precise terminology that allows us to analyze the environmental factors that lead a person to “want” a particular item or event. One cannot measure “desire” directly, other than through a very poor proxy such as self-report rating measures. But one can measure in a very quantifiable and reliable manner the amount of time that has passed since the last walk.

When Milton wants or desires a walk, it exists in the absence of that event, that is, he is not currently getting a walk. Also at that instance, there has been some quantifiable length of time where he has not had a walk. A state of deprivation with respect to getting a walk thereby exists. If this state of deprivation is sufficient enough to “motivate” him to engage in behavior that has historically produced a walk, such a deprivation condition or state creates or “establishes” the value of the walk. In other words any given behavioral function becomes operable only when an antecedent condition is in effect that establishes a given event as valuable, that is, a functional reinforcer. The term, Establishing Operation or EO, was first coined by Keller and Schoenfeld (1950) and refers to both states of deprivation (which involve access functions) as well states of aversion (which involve escape functions).2

All access and escape functions exist in the presence of an antecedent condition. There are two variables that are inherent in antecedent conditions: motivational and discriminative variables. It is important to consider at this point the role of a motivational condition (i.e., in common parlance, people’s motivation for doing what they do). It is the “driving force” (Cipani & Schock, 2011; Cipani & Cipani 2017) of a behavioral function. Milton engages in behavior that produces a walk at that particular time because the prevailing motivational condition (EO) is sufficient enough to create two effects: (a) establish getting a walk as being of extreme value and (b) make behaviors that have historically been effective in obtaining such an outcome very probable. This is true for everyone. The motivational condition or EO “drives” the function. When the EO is considerable at a given point in time, a function becomes operable, and a historically effective “functional” behavior occurs and thereby produces the desired outcome (getting something or getting away from something). When the EO is weak or nonexistent, a particular function does not exist. Hence any behavior that produces a specific given outcome is not probable. For example, if getting attention becomes valuable (i.e., EO for attention is great), then a behavior that produces attention becomes very probable. When the EO for someone’s attention is weak to nonexistent, such a behavior becomes less likely (unless it also serves other functions).

For example, students raise their hands when they want teacher attention and help with an ambiguous part of a class assignment. The EO for attention/help is of sufficient strength to make such a behavior likely, because the value of teacher help has been made essential. Once the student gets the necessary help/instructions, the “need” to raise one’s hand diminishes for a while (until the assignment once again becomes ambiguous to this student). In the current terminology, the EO has been lessened and abated. In summary, the greater the strength of the EO the greater the likelihood the person will engage in a behavior that effectively produces the particular outcome. The lower the strength of the EO the lower the likelihood the person will engage in such a functional behavior. EOs drive particular relevant functions!

But why do some people desire potato chips twice a day whereas other people can go a week without such a desire (interpret “desire” for EO)? Based on current research, all EOs appear to be unique to the individual person. Each person will differ with regard to the specific level of deprivation or aversion that establishes the value of a particular outcome (be it access or escape function).

Consider the question: What is the point at which food will operate as an effective reinforcer (i.e., food deprivation, increased calorie expenditure)? For some people it is 4 hours after they finished eating lunch. For other people, the point at which food will operate as an effective reinforcer does not occur until the end of the day. One cannot say food will operate as an effective reinforcer when we deny access to food for a 10-hour period (absolute value), nor can one say that it will be a static time period for that individual without considering the level of calorie expenditure for that person. Food deprivation is relative to the individual under consideration and dependent on level of activity. It is also relative to other states of deprivation or aversion that may exist at that time.

Therefore, the point in time when a given individual will engage in behaviors that make access to food highly likely is different for each individual and each situation. However, the extent of the deprivation state necessary for food to become valuable, for a particular person, is quantifiable and can be determined with a high degree of accuracy using one of the various analogue assessment procedures (see Chapter 2).


“Absence makes the heart grow fonder!” Can you explain this in behavioral terms (EO) and use a hypothetical example to illustrate your point?

EOs for Access Functions

EOs relative to access functions typically involve some period of deprivation of that item or event (i.e., deprivation EO). To reiterate, the amount of deprivation necessary to produce an establishing operation is relative to each individual, and not one absolute value for everyone. Table 1.4 presents several examples of EOs for access functions and the more informal, layman’s term used to describe these events in general conversation. In the first example, a person may have eaten breakfast but missed lunch, so by the time dinner is being prepared, the value of food is very high. Therefore, any behavior or set of behaviors that result in obtaining food will be strengthened. The value of food will be abolished when some volume of food has been ingested causing the stomach to expand. In the second example, if a person has not had any fluids for several hours, the value of fluids will be established. Any behaviors that result in obtaining fluids will be strengthened. In this case, ingesting fluids will operate as an abolishing operation.


Establishing Operation

Lay Term

Reinforcer for Some Behavior

1.Deprived of food



2.Deprived of drink


Drink, liquids

3.Ingested salt


Drink, liquids

4.Deprived of physical contact

Wants hugs

Physical contact (specific type)

5.Deprived of attention

Annoying, wants attention

Attention (specific person/type)

6.Deprived of TV

Wants TV


7.Deprived of stimulation, all varieties

Wants stimulation

Stimulation (specific kind)

8.Smoker deprived of cigarettes

Wants to smoke

Inhalation of cigarette smoke

In the third example, consider that bartenders have long understood how to effectively establish the value of fluid intake. Have you ever been to a bar at happy hour? There are typically many food items available for free. Have you also noticed that when you eat the free food, you seem to get thirsty much more quickly than usual? Most of the food items served at happy hour are either high in salt content or very spicy. This, of course, tends to establish the value of fluids, and the discriminative stimulus (SD) for availability of fluids is typically the bartender. The fourth example considers that for most people not having physical contact (or a specific type of physical contact) for an extended period of time will establish the value of physical interaction. Under this condition any behavior that results in some level of physical interaction will be strengthened. It appears that there is an EO for some level of physical contact.

Having no social interactions for a specific period of time will increase the value of any interaction (see fifth example). In regard to attention, the EO may be for a specific person, which means the reinforcer would be attention from the specific person. In the sixth example we consider the case of a child with a history of watching TV every day after school. The television stops working, so the child does not have access to TV after school for a period of about 1 week. The value of TV watching will increase such that, given the opportunity, he will engage in high levels of behavior to get access to viewing TV.

Example seven presents that some level of stimulation appears to be an EO. Deprivation of all types of stimulation for some period of time will increase the value of any type of stimulation. Consider the case of an individual with restricted mobility living in an environment that is deprived of stimulation. Under this condition, individuals will frequently engage in “mouthing” of clothing, hands, or any item they can manipulate. These behaviors produce some level of stimulation, and therefore, under this condition the mouthing behaviors will be strengthened.

The eighth example illustrates that people who smoke certainly value cigarettes, and we can clearly establish the value of a cigarette by ensuring the person does not smoke for some period of time (EO). Under this condition, any behavior that produces access to cigarette smoke will be strengthened. The abolishing operation in this case appears to be levels of nicotine in the person’s body.

In some cases there may be a considerable delay between the EO and the actual occurrence of any behavior related to that specific EO. Here is a perfect example. Some clients in treatment facilities will retaliate against a peer who has assaulted them. This is easily explained when the retaliation is immediate, but how do you explain retaliation that occurs several days later? Clearly the EO establishing the value of retaliation to the other person is already at maximal value immediately after the person is assaulted. What would account for the delay? The most likely explanation is related to the availability of retaliation (inflicting damage to the other person) as a reinforcing event. Specifically, after an assault event, staff are more likely to be vigilant in order to avoid the retaliation. They are more likely to be present when the two clients are in the same area. Staff presence makes the context for retaliation not available, because the staff would prevent the occurrence of damage to the other client. Over time the staff are less vigilant, and eventually the two clients are left alone. Under this condition, assault is highly likely to produce damage to the other client because the staff are unlikely to be close enough to intervene, so the retaliation then occurs. Because this may occur several weeks after the initial assault event, staff may report that it happened for no reason or “out of the blue.”


Explain what a deprivation establishing operation is. How does it affect the value of the specific reinforcer? Give an example of the EO and its effect on a function involving attention.

EOs for Escape Functions

Thus far we have considered EOs only for access functions. EOs for escape functions have the same characteristics as those for access functions. However, the environmental event in this case is related to an outcome that the individual is acting to avoid or terminate (see Table 1.5). The establishing operation consists of a state of aversion, hence an aversive EO.

In the first example, a teacher may present a relatively difficult task to the child. Because the child is not capable of performing such a task or demand, he or she finds the presence of such a task demand aversive (EO). Therefore, a behavior, or set of behaviors, in the repertoire of this child that results in its removal will be strengthened. Realize that what is difficult for one child may be easy for another, hence, difficult task is a relative term.

In example two, a behavior that terminates the presence of a person’s obnoxious behavior will be strengthened under conditions involving the presence of the noxious conditions (EO). Let us say this person’s obnoxious behavior is using foul language. If a person does something that affects this person whereby such language is lessened or eliminated, such behavior will be strengthened in the future when faced with this person’s foul language. But let us say that nothing seems to perturb this individual who goes right on with rude language. Leaving the area then becomes probable. Why? Because it terminates the person having to listen to and put up with such language. In example three, painful stimulation establishes the value of termination of the discomfort (EO). Behaviors that produce escape or avoidance of such painful stimulation will be strengthened. Again, painful is a relative phenomenon, varying according to the individual.

In example four we consider that for many people, rain on the head establishes the value of terminating rain on the head (EO). This group of people will therefore engage in behavior that results in rain no longer falling on their heads. This may include using an umbrella, holding something else over their heads, or simply going under cover. Any of these behaviors have the same effect of stopping rain on the head and will therefore be strengthened. Example five is similar to example one; however, in this case, the teacher may present a relatively long task to the child. Because the child is not capable of performing such a task or demand for that long period of time, the task demand is aversive (EO). Therefore, a behavior, or set of behaviors, in the repertoire of this child that results in its removal will be strengthened.


Establishing Operation

Lay Term

Reinforcer for Some Behavior

1.Presence of difficult task


Removal of task

2.Person acting in an obnoxious manner


Terminating engagement in social situation

3.Presence of pain

In pain

Termination of painful stimulus

4.Rain on the head

Does not want to get wet

Termination of rain on head

5.Request to complete a long task


Shortening or termination of the long task

6.Occurrence of extra pyramidal symptoms (EPS)

Wants symptoms to stop

Termination of the EPS

7.Allergic rhinitis

Wants relief from allergies

Termination or reduction in allergy symptoms

8.Drug withdrawal symptoms

Wants more drugs

Termination or reduction of withdrawal symptoms

9.Alcohol withdrawal symptoms

Wants more alcohol

Termination or reduction of withdrawal symptoms

EPS, extra pyramidal symptoms.

Example six considers the behavioral effects of EPS, which are side effects associated with many antipsychotic medications. These EPS can be quite aversive. The occurrence of EPS will establish the value of (EO) terminating the EPS. The most direct means to terminate the EPS is to stop taking the medications that cause the EPS. This EO is frequently the major factor in medication noncompliance for people diagnosed with major mental illness. In example seven we consider that for people who have allergies and experience the private event of a sinus headache, the value of terminating that stimulus is established (EO) when they are in the midst of a sinus headache. Most have found multiple means to do this: (a) take sinus medication; (b) take a hot, steamy shower; and (c) apply ice packs to the affected area. But suppose you are living in an institution, and you are not verbal and therefore unable to communicate that you have a sinus headache. The most direct way to terminate the sinus pain, albeit briefly, is to press on or hit the sinus area that is producing the pain, thus these behaviors would be strengthened.

Examples eight and nine review how the occurrence of drug or alcohol withdrawal symptoms will establish the value of terminating the withdrawal symptoms (EO). The most direct means to terminate the withdrawal symptoms is to take the drug associated with the withdrawal symptoms, thus strengthening this behavior. The first use of drugs and long-term abstinence is associated with several other EOs. The particular EOs involved are typically idiosyncratic but can and should be delineated when approaching the treatment of substance abuse.

Consider some people with the genetic syndrome called Prader-Willi. They never seem to ingest enough food or calories to abolish food as a reinforcer. For these individuals, the value of food is established anytime they are not ingesting food. For people diagnosed with borderline personality disorder, the value of adult or peer attention is difficult to diminish or abolish, and the value of attention is established very quickly after the last social interaction. Both of these examples are probably best analyzed as difficulties associated with ineffective AOs. In both cases you could reduce the value of the reinforcer (food, attention) by continuous, noncontingent reinforcement, but in both cases this presents pragmatic and ethical concerns.


Explain what an aversive establishing operation is. Using an aversive interaction with a classmate as the presenting condition, give an example of an escape function, by delineating a behavior and its effect.

Abolishing Operations

As mentioned earlier, once the functional reinforcer is delivered for a given behavior, the value of such an event is diminished. The delivery of a sufficient amount of the reinforcer abolishes the existing EO. Hence, this operation is termed the AO. Table 1.6 depicts access functions and the relationship between EOs, behavior, and AOs, using the prior examples found in Table 1.4. In example 1, the delivery of a sufficient amount of food results in the abolishment of the prior state of deprivation. In plain English, once you have eaten enough (relative to each individual), you are no longer hungry. The remaining examples illustrate the same principle of the AO relative to the initial deprivation EO. It is important to note in each example that the AO exists once there is a sufficient amount of the functional reinforcer acquired.

The same is also true for escape functions, with respect to the EO, behavior, reinforcer delivery, and the AO. Table 1.7 illustrates their relationship. In example 1, the removal of the task for some behavior produces the AO—termination of the aversive EO.





Antecedent Discriminative Stimuli

You should now understand the role of motivation: behaviorally defined as the existence of an establishing operation (motivational condition is great) or an abolishing operation (motivational condition is lessened or not existent), and its effect on the function of behavior. However, you may be asking the following question: “Why does a child with severe disabilities hit him or herself when in the classroom to get a cookie (EO present); but points to a picture of the cookie to get one when with the speech therapist?” The answer: People reinforce (sometimes inadvertently) specific behaviors under specified motivational conditions!

Socially Mediated Functions and Discriminative Stimuli. The technical term for the other antecedent variable that affects behavior is the “discriminative stimulus” (SD). The SD is a stimulus associated with the availability of an outcome (given a specific behavior or class of behaviors). While you may be familiar with this term as it relates to instructions, directives, and task demands, let us examine its role in socially mediated functions of problem behavior, where the reinforcer is produced via the behavior of another person. This is important to remember as we examine what part the discriminative stimulus plays in socially mediated functions and what or who represents the SD.

The previous example illustrates that the same desired outcome is obtained with two different behaviors, hitting oneself versus pointing to a picture. Why does the client hit him- or herself in the classroom? Someone has made cookies accessible for that particular behavior a reliable outcome! Why does the person not hit self in the speech therapy room? The speech therapist has not made hitting a behavior that produces cookies; rather a nonvocal communication produces that result. Both the person in the classroom and the therapist “are correlated with” the availability of cookies—but what behavior “works” is quite in contrast. In one circumstance, hitting is the functional behavior; in the other, pointing to a picture is functional. Therefore, the person in the classroom is discriminative for hitting oneself to obtain cookies. This is not the case with the speech therapist. The speech therapist is discriminative for communication responses to obtain cookies.

For socially mediated functions, people constitute the discriminative stimulus (at least part of it). Since the reinforcer is obtained via their actions, people determine what behaviors result in obtaining such. This is true for not just attention and tangible reinforcers, but also for the termination of aversive events. The previous example could just as easily portray a circumstance where the child hits him- or herself to have the classroom instructional staff person terminate a task demand. Such staff persons would be discriminative for escape behaviors involving self-abuse. In their presence, self-abuse results in the task being removed. The speech therapist ends the session when the child points to a picture with the words printed underneath, “break, please.” She would be discriminative for such a behavior serving as the mechanism for escape from the session. In short, a specific function under a given EO is operable with a particular person(s) serving as the discriminative stimulus.

How do you determine what person(s) is discriminative for problem behavior? In the previous example, when the child is desiring a cookie, and is with the teacher, what are you very likely to see? Problem behavior! Hence, it is often easy to determine the discriminative stimulus for socially mediated functions. Persons in the child’s natural environment where the rate of problem target behavior is high are discriminative for such behaviors.

For those of you who use profane language at times, you might reflect on the circumstances where such behavior is more probable. Where you do engage in such language, people and contexts serve as discriminative stimuli for such language. In contrast, being in church or at your professional job involves stimuli that make such language unlikely (hopefully).

Can a person be discriminative for problem behaviors in one setting, but not serve as a discriminative stimulus for such behavior in a different setting? Of course! You probably can recall many circumstances where someone behaves differently in two different contexts, with the same person involved in the provision of the desired reinforcer. For example, a child may engage in mild crying (“whining”) to get a cookie at the grocery store from his or her mother. However, such behavior at home results in being sent to one’s room (without cookie). Even with the same adult, the function of one set of behaviors can be operable in one setting, and yet be nonfunctional in another setting.


Provide a scenario contrasting how an adolescent girl can behave in one fashion to get attention from her parent and then engage in a different behavior to get attention from her boyfriend. Identify the discriminative stimulus for each situation.

Direct Contingency Functions and Discriminative Stimuli. For problem behaviors that serve a direct contingency function under deprivation or aversive EOs (DA or DE functions), people do not constitute the discriminative stimulus condition! What are the characteristics of discriminative stimuli for DA and DE functions? First, there is some context condition in the physical environment that makes the desired outcome available. But this also is often combined with the absence of certain people that would impede the direct access or escape. These two conditions usually sets the stage for an unacceptable behavior that directly produces the desired outcome unencumbered.

If a client wanted an extra piece of cake in the refrigerator, and was not allowed to have such, what are the conditions under which the behavior of grabbing a piece would be effective? Obviously, the refrigerator has to have a piece of cake in it (physical discriminative stimulus present). But also, the staff person or persons who would impede such an effort would have to be away, or maybe “sleeping on the job.”3 For problem behaviors that directly access the tangible reinforcer, when such is not allowed or prohibited, the absence of persons who would interfere with the obtaining of such is discriminative for such attempts, along with the availability of the desired item in the physical environment.

Suppose someone wanted to leave the classroom. The door serves as the discriminative stimulus in the physical environment that makes the chain of behaviors that lead to opening it and walking outside functional. Unless some staff person is blocking the door! The chain of behaviors that would lead to outside will be successful only if the staff person is either not there, or is engaged in a distracting activity, which makes the “escape” viable.

In summary, there are two parts to the antecedent condition for a given function. The EO is the “driving force” for the function, establishing the value of getting something or getting away from something. The discriminative stimulus dictates the manner (i.e., form of behavior) in which the desired outcome will be procured. With those two elements, a specific behavioral function exists.


When considering behavior change at the antecedent level, it is important to be aware that the greater the value of the EO, the greater the strength of the behavior. A person deprived of food for 24 hours will be much more likely to engage in behavior that has historically been effective in procuring food than a person who has just finished eating a full meal. As the value of the EO increases, the frequency (both absolute and relative), duration, and intensity of behaviors associated with abolishing that EO will also increase.

For example, let us say that you have been without a candy bar for 1 day. Given this level of candy bar absence, the EO is not sufficient for you to go out of your way to go buy one. Of course, if one happens to be in a desk drawer of yours, given that state of deprivation you would eat it. But to reiterate, you would not expend a great amount of effort to get a candy bar at this point. However, let us say a week goes by and you (for some reason) have not yet had a candy bar. Additionally, you see someone at work who has just opened one and starts eating it. At this point the state of deprivation is great, and with the added observation of someone else eating one, you scour the neighborhood at work to find a candy bar. In fact, you do not give up your quest to get the candy bar until you finally locate one with about just 10 minutes to go before you must get back to work. Your “desire” (i.e., EO) was so great you were willing to forego some other usual lunchtime activities (e.g., sitting by the nearby lake and “people watching”) to obtain a candy bar! The state of deprivation with respect to any item (or state of aversion as well) does vary across time. When it reaches a point where such is sufficiently deprived (relative to that person and other states of deprivation and aversion prevailing at that time), functional behaviors become more probable, intense, and frequent. Until the AO results!

Decreasing Targeted Behaviors by Reducing the EO Strength

Would you engage in a behavior (of some considerable effort) that obtains something if you have had “your fill” of that item? Let us say a young man with physical disabilities has learned to scream to get some school staff person to come to him and move his wheelchair outside, under a deprivation EO for such. Of course this is not a desirable circumstance, and some intervention is needed. Suppose when examining his screaming bouts for the past several weeks, we note that he screams between two and four times per day. Remember that his screaming occurs under a desire to be wheeled outside the classroom at that point in time. What do you think would happen to the frequency of his screaming bouts if school personnel took him outside about once an hour (i.e., five or six times per school day)? It makes screaming rather useless, that is, nonfunctional. By making outside time more available to him, we reduce the possibility of a deprivation EO occurring with respect to outside time. Hence it is unlikely that this student will be faced with a time when he desires being outside, since the amount of outside time programmed into his schedule is at or above his usual level (2—4 times). He will not find the “need” to scream.

For example, if a child engages in a behavior to get a certain preferred activity, such as computer time, then increasing the child’s time on the computer two- or three-fold over the current access level will do two things. First, such a manipulation will reduce the value of this particular event. Second, it will alter the frequency of the target behavior that serves such function, that is, getting computer time.

In regard to behaviors serving escape functions, there are two methods for making escape from some aversive condition not as necessary (i.e., valuable). These methods are: (a) do not present the aversive EO or (b) alter one of the elements that make such a condition aversive. Here is an example of the first method. A hypothetical student does “not like” long division (such constitutes an aversive EO), and engages in a variety of disruptive behaviors when asked to do such by the teacher. Of course, such behaviors are relatively successful in getting the teacher to terminate such a task and provide an alternate task. Given that information, what would you predict would be the outcome if, one day, the teacher did not present the long division problems but instead started with the task that the teacher would otherwise switch to when encountering disruptive behavior? By not presenting the aversive EO, the teacher would also avoid the disruptive behavior the student typically exhibits.

In summation, one can decrease, if not eliminate, the behavior associated with a particular EO by ensuring that it never becomes valuable. In the case of access functions, this involves providing the reinforcer frequently and at high enough levels such that their value is never established. While this antecedent manipulation does not alter the function of the target behavior directly, it will reduce the behavior in frequency, duration, and intensity, as well as increasing response latency. Table 1.6 provides some examples of this methodology. With regard to target behaviors maintained by escape functions, the method of changing the motivational condition for such behaviors is to eliminate the presence of the aversive event, make it less aversive, or terminate the aversive condition noncontingently (see Table 1.7).

It may also be possible to alter slightly the presenting conditions, removing just the factor(s) that make the condition aversive. Of course, one would have to be quite familiar with the student or client to utilize such a strategy. In the previous example, let us assume that long division is aversive to this student because he or she is not skilled at the procedures of doing such. We observe the student’s struggles with how to divide the divisor into the relevant part of the dividend. One can make this particular instructional task less aversive by teaching the student how to perform the first step or component of long division (dividing a single digit into the first one or two digits of the dividend), and practice that only (until mastery). Such a strategy reduces the aversive aspect of the task and would thereby reduce the problem behavior. Such a manipulation will ensure that the value of escape or avoidance of that task is never established; as a result, the frequency and intensity of the oppositional behavior will be reduced or eliminated (Tables 1.8 and 1.9).


Maintaining Reinforcer

Antecedent Manipulation

Adult attention

Increase frequency/duration of attention

Tangible reinforcer (food)

Increase availability of food throughout the day/time period

Tangible reinforcer (free time)

Increase availability of free time throughout the day/time period

Tangible reinforcer (preferred activity)

Increase availability of preferred activity throughout the day/time period


Maintaining Negative Reinforcer

Antecedent Manipulation

Removal of unpleasant social situation

Avoid presenting such events, or reduce the level of aversion by altering some aspect of the situation

Removal of relatively lengthy task, chore, or assignment

Avoid presenting such tasks, chores, or assignments, or reduce the level of aversion by altering the length of the task

Removal of relatively difficult task, chore, or assignment

Avoid presenting such tasks, chores, or assignments, or reduce the level of aversion by altering the difficulty of the task

Removal of physically aversive stimulus

Avoid presenting such aversive stimuli, or reduce the level of aversion by altering some critical aspect of the stimulus

Increasing Behavior by Enhancing the EO Strength

As should be clear to you by now, a behavior (that produces a given function) is unlikely to occur if the value of a particular outcome has not been established. When attempting to teach a new behavior, it is important to understand that there are two ways to ensure that a reinforcer is at its maximal value. First, you could wait for naturally occurring environmental changes that establish the value of a particular outcome. For example, if you are trying to work on establishing behaviors related to accessing food, you could wait until right before meal times to use food as a reinforcer. Second, you could contrive a situation that increases the value of some outcome as a reinforcer. For example, if you are trying to work on establishing behaviors related to accessing fluids, you could provide very salty food to a person prior to and during training, thereby ensuring that the value of fluids remains at a relatively high level.

A more frequent use of contrived EOs is seen in the teaching of social skills and verbal behavior. You can manipulate the value of social/verbal interactions by establishing the value of a tangible reinforcer and then making access to that reinforcer dependent on a social interaction. This arrangement will establish the value of social interaction. Once the value of social or verbal interaction is established, you can effectively invoke the social behavior.


Reinforcement systems that are designed to simply target appropriate behavior for increase have been referred to as contrived contingencies. The derivation of the specific contingency is contrived in that the functional relationship between problem behavior and its enabling contingency may not have been utilized or deduced. In using contrived contingencies to solve problems, one simply selects a powerful reinforcer, of unknown relationship to the current problem behavior, and delineates a temporal relationship between it and some desirable behavior. Such a contrived contingency hopefully competes with the existing function of the problem behavior.

Nevertheless, contrived contingencies do produce a function, and there are many reasons why they may be preferred as a first stage of an ecological intervention in many settings. A contrived classroom contingency arranges a functional relationship between some target criterion behavior (e.g., a designated percentage of time on-task, designated score on quiz grades) and production of effective reinforcers (Barrish, Saunders, & Wolf, 1969; Greenwood, Hops, Delquadri, & Guild, 1974). Point or token systems entail conditioned reinforcers (points) that are traded in for back-up reinforcers. The designation of potent back-up reinforcers provides the system with the requisite mechanism to influence behavior (Inkster & McLaughlin, 1993). Therefore, such contrived contingencies make student performance on the assignment or material of consequence. Many people refer to contrived contingencies as nonfunctional contingencies. You should now realize that functional is determined by the environment, given the presence of certain EOs. Designating a contrived contingency makes some behaviors functional with respect to the delivery of certain stimuli that are hopefully precious at that moment in time. Let us examine how contrived contingencies produce functions.

Ms. Tally designs a point system that provides points for on-task behavior during intermittent observations of such behavior. She has learned in her behavior analysis courses that contrived reinforcement contingencies have been demonstrated to be effective with many students, including those labeled as unmotivated and not intrinsically motivated. She feels that the term intrinsically motivated describes a certain percentage of students in her classroom, and she wants them to perform to a higher level. She implements a beeper system to increase student on-task behavior (see Cipani, 2008, for details). On-task behavior is often defined as looking at the materials during seat assignments or at the teacher during lesson presentations or instructions. A beep is produced at random intervals. With these random beeps, she intermittently observes the students in her class and records whether they are engaged with their class assignment. When students are demonstrating on-task at the time of the sampling, they earn points. If they are not on-task at that point in time, they do not earn points. Points are traded in later for a variety of back-up reinforcers, such as tangible items and preferred activities.

Ms. Tally implements this program for 4 weeks subsequent to a baseline. As a result of implementing the beeper system, the mean rate of on-task behavior goes from 46% during baseline to 77% for the first 4 weeks of treatment. Further evidence of the efficacy of the beeper system is the substantial improvement on the students’ quiz grades. Why does this procedural arrangement between student engagement and contrived reinforcement contingencies work? In using the beeper system, it was not necessary for Ms. Tally to determine the function of the existing off-task behaviors in order to reduce the level. Why does such a contingent relationship alter student performance in the desirable direction?

The environment often creates, through the class routine involving instructional periods and associated academic tasks, deprivation with respect to student-preferred items or activities. The extent to which the contrived contingencies involving points will increase the rate or level of on-task behavior (or other targeted behavior) depends on two factors: (a) the relative deprivation of the back-up items and (b) the aversive nature of the current tasks or activities. When the back-up items and events are relatively deprived, behaviors that produce such become more probable. In using the beeper system, on-task behavior becomes functional in producing points that provide eventual access to such events. Hence a function exists. If this new function “overpowers” the aversive EO or EOs of the particular task, it will make on-task behaviors very likely with this particular teacher and her imposition of the point system.

Therefore, contrived contingencies may address EOs that are far more powerful in terms of their value-altering effects than the EO for the current problem behavior. For example, a student engages in gazing out the window for minutes at a time to take in the view. We could hypothesize that such behavior may either be a function of direct access to visual sensory reinforcement or direct escape of engagement with instructional materials. In the absence of contrived reinforcers for performance, such behavior occurs several times (of variable duration) in a given instructional period. The teacher may even report that this student gets easily distracted. However, when contrived contingencies for on-task behavior are deployed, via the beeper system, gazing out the window goes down to a markedly lower frequency and also of very short duration. Note that the EO addressed by the contrived contingency overrides the weaker motivating condition for viewing the scenery or directly escaping the task. Looking outside the window does not produce a reinforcer that is as powerful as getting points for back-up tangible reinforcers such as extra free time. As the movie adage taken from Jerry McGuire goes, “Show me the money!”

What about the students who do not need teacher-designed contrived contingencies to persist in engaging in the task or assignment? Some students (often labeled as “good”) continue to perform despite getting more work after they complete their initial assignment. Are these instances of intrinsic reinforcement? While many nonbehavioral educators point to this as an example of intrinsic motivation, they fail to understand the powerful role of other superimposed contingencies on classroom performance. Performing well in class can bring teacher, parent, or other adult attention and, in some cases, peer approval and attention. Grades also provide a source of social contingencies, and some additional benefits from rule-specifying contingencies (“If I get good grades, I’ll go to college.”). Therefore, in these students, because other EOs are present (e.g., attention from teacher, parents, peers), and there are discriminative stimuli for such behavior, their behavior is well understood within a contingency analysis. These students’ performances are sustained because of these inherent superimposed contingencies in a given student’s life.

While such superimposed additional contingencies can explain why some competent students do not need contrived contingencies to maintain classroom performance, what about those who do not perform under such conditions? These students cannot perform at the academic levels that recruit such reinforcers. What happens to learners who do not have the competence to be affected by such superimposed contingencies? Very often, the failure to use contrived contingencies with these students results in a decrement in everyday performance. Their performance in the classroom becomes highly variable. On some days, they perform at admirable levels, and on other days they do not complete their assignments. Unfortunately, when their performance over time fails to recruit superimposed additional reinforcement contingencies, such as good grades and adult approval, they become even more unproductive. When these learners fail to consistently recruit those contingencies, their daily work does not come under control of such consequences.

As a result of this failure to recruit superimposed contingencies, classroom instruction and tasks become an aversive condition. Instruction, task assignments, and other learning activities develop aversive stimulus properties, which become EOs for escape and avoidance behaviors. With the reluctance of the teacher to designate some alternate (performance-based) criterion for terminating tasks (see the case of the wacky contingency, Chapter 3), undesirable behaviors become strengthened through the selective reinforcement of such.


Behavior (operant) that occurs with high probability in certain antecedent contexts (EOs and discriminative stimuli) is functional. This is true for both desirable and undesirable behaviors. There are two major functions: (a) access functions (positive reinforcement) and (b) escape functions (negative reinforcement). Further, the manner in which such functions are produced can be from one of two methods: (a) directly or (b) socially mediated. The following chapters will expound on this concept of specific operant behavior becoming functional under specified antecedent contexts as a result of its environmental effect (i.e., delivery of a reinforcer that produces an abolishing operation).


When mental health personnel view behavior as a result of an internal condition, behavior is often explained as the result of such a condition. Hence, such behavior is to be expected from time to time, and environmental contexts are relegated to an immaterial role. The following case is instructive for several reasons. First, the type of behaviors occurring at such a young age was rather unusual and of a form that intimidated the adults caring for the child. Second, an analysis of the probable function of such behaviors revealed “super-ordinate” contingencies that existed in foster care placement. Such contingencies are invoked when ­foster children of a young age engage in behaviors that make the current placement caretakers no longer willing to care for them.

Description of Case

Polly was a 5-year-old child who had been referred to me for in-home behavior management. She was currently residing at her second foster home, having been detained from her biological mother’s home previously by child protective services (CPS). Polly’s biological mother was reported to have perpetrated several incidents involving physically abusive behavior toward Polly and her other siblings. Her first foster placement ended rather quickly (prior to my involvement) as she reportedly exhibited uncontrollable outbursts and tantrums. Subsequently, the foster parent requested a termination of this placement, and the county department of social service turned her case over to a foster family agency (FFA) for placement. The family with whom Polly was residing at the time of my involvement belonged to this FFA. I did not receive much information besides this brief placement history prior to my first home appointment to meet this family. An initial phone contact with the foster mother provided some basic information regarding Polly’s current problems in her home. Polly was described as an out-of-control child who did not respect adult authority. Further, she could become verbally aggressive and sometimes physically destructive when upset.

Background information obtained from the case manager revealed that Polly was initially raised by her biological mother along with four other siblings. Two of the siblings were children of Polly’s mother and father (although Dad had long since disappeared). Her other half-siblings were each fathered by a different man. The mother at the time of detainment was raising all the children on her own.

When Polly attended preschool and childcare, while still in her mother’s custody, she was expelled as a result of severe behavior problems that the school could not tolerate. She was reported by the preschool staff to hit, pinch, bite, and attack other children without provocation. She would not respond to directions from the teachers, particularly directions to go to time-out for misbehaviors. It was during these times that she would become verbally abusive. While staff noted that she could be pleasant and happy on many days, it was her bad days that made the situation intolerable, thus requiring her expulsion.

When Polly was 5 years old, CPS was notified by Polly’s elementary school of possible child endangerment. A CPS social worker came out to investigate that referral and found an abrasion on Polly’s chin. Her brother was found to have a similar mark on his back. The school also reported to CPS that Polly was often 30 minutes early to school without supervision, often dirty, and with no snack. Polly told CPS that there was no food in the house. As a result of all this information, Polly and her younger siblings were detained and placed in foster homes.

On my first appointment I pulled up to the curb of the foster parent’s house and covertly noted that the foster family lived in a nice neighborhood. I was greeted at the door by the foster mother and shown into the living room. Polly was sitting in a chair at the dining room table engaged in some innocuous activity (I thought to myself—“looks good so far”). Present in the room were the foster father, a social worker from the county, and a social worker from the FFA, who would coordinate the behavioral treatment. I began interviewing the foster parents as to the nature of Polly’s behaviors, that is, under what conditions did she seem to exhibit such behaviors and what was their method of attempting to manage her. The parents painted a picture of a child with a dual personality (I am not implying the diagnosis here, their implication, not mine). Sometimes, she was the best-behaved child, played with peers in an appropriate manner, and was so angelic. But on the other hand, when she was upset, watch out. Her demeanor changed, she was no longer pleasant, and profane language and hurtful words poured out of her mouth. However, during the entire time I was interviewing the foster parents and the social workers, I noted no incidence of such behavior. She continued to entertain herself at the table as per the direction of the foster parents.

As I obtained enough information from the foster parents at that time, I then decided to test the waters. If at all possible, I like to see how children handle commands and instructions from me to give me an indication about their compliance to adult instructions. I gave Polly a few brief requests, which were all responded to with a smile. When I praised her and told her she did an excellent job in following my instruction, her smile went from coast to coast. I then attempted to correct her on a math task I assigned. This is often a good test of a child’s ability to handle criticism appropriately. This “simulated criticism” can usually evoke some level of agitation with children who are disrespectful of adult instructions. She made no objection or dispute of my correction! Over the course of the 15—20 minutes of individual work with her, in the dining room, while all the adults watched on, she appeared to be the antithesis of the child who was described to me. At that point, I just could not picture what someone would have to do in order for Polly to become a child who was out of control. I could not even envision her being in the vicinity of a screaming tantrum, replete with profanity, let alone being the source of such verbiage.

If first impressions are a valid measure of a child’s behavior in their social environment, one might surmise that Polly’s demonic nature was a figment of the foster parent’s imagination. Could the parents possibly be over-­controlling and require perfectionism? How could this innocent child do anything but good! With such great compliance from Polly, the thought of these foster parents being so inept that they could “bungle” parenting such a sweet child ran through my mind but was quickly dismissed. The evidence to the contrary was substantial. In addition to the report from these foster parents, severe behavioral episodes and reactions were reported by Polly’s previous teachers and foster parents.

With no incident of behavior that remotely resembled a troubled child, the foster parents wanted to assure me that this is what they expected for my visit. They indicated that it was Polly’s nature to be angelic with other people. They informed me that she had a mental disorder, which made her act this way in front of strangers, but then she would quickly turn on the parent figure. Could this be the granddaughter of Norman Bates? They handed me some material on reactive attachment disorder (RAD).

Why do these children behave like this? According to attachment theorists (who predominate this field), these children fail to develop attachments because of an early trauma (such as abuse, neglect, or separation). As a result they fail to “bond” with a significant caretaking adult. Because there was no initial bond between the child and mother, they feel no remorse for hurtful incidents regarding other people. According to attachment theorists, a “moral conscience” is developed only as a result of a baby’s attachment to an adult early in life. People who argue this view trivialize the role of social consequences of behavior in teaching children to learn to refrain from saying spiteful things to people. The foster mother reiterated that Polly feels no remorse for persons she hurts or offends. Further, I should not be fooled by the display of good behavior, as these children are the masters at this behavioral deception. She again reassured me that Polly is really disruptive to their family situation, and I should not be fooled by her deceptive appearance on this visit.


Fortunately, that sermon about RAD did not sway me from my objective: to bring problem behaviors under control of more systematic contingencies. In some cases, treatment merely requires a manipulation of contingencies; in others, skill training is added to that component. The child’s caretakers, whether they be biological parents, foster parents, or group home staff, are then trained on-the-job to perform the requisite behaviors in response to the child’s behavior.

I like to target just one problem area at time, bringing desired change in that area before tackling another. I chose to target a daily context that was generating problems in child management: bedtime. Polly was reported to have difficulty going to bed and staying there. She would get up constantly and have to be put back to bed. To get Polly to accept bedtime more readily, the foster parents had tried enforcing bedtime at a predictable time and incorporating low-key activities prior to bedtime. With that regimen, Polly was having less of a problem at bedtime than in the prior placement. However, on nights when she had difficulty, she had severe tantrums and would get out of the bed constantly.

While these were good first steps on the part of the current caretakers to address bedtime problems, I suggested some additional procedures. There should be an incentive for good bedtime behaviors to compete with the desirability of creating havoc in the family household. My plan was to allow Polly to earn a half-hour of later bedtime the following night for good bedtime behavior (the current night). Good bedtime behavior was defined as (a) the lack of tantrum behavior and (b) not getting out of bed once she was placed in bed. She could also lose an hour off of regular bedtime the following night if she got up or had tantrums.

The first two nights of this regimen were not good. The parents estimated that Polly got up about 25 times that first night and a little less the second night. It was fortuitous that the foster parents did not bail out at that point. I had prepared them for this possibility. On the third night, the plan resulted in her complete cooperation and access to the privilege of staying up late the following night. For the entire next week she responded well to the plan, always getting her extra half hour each subsequent night. Parental persistence paid off.

However, these parents gave notice to have her moved prior to my third visit. She was moved to another home about 25 miles away. I was notified initially to discontinue the case because the new foster parent reported that Polly was not a management problem at the time. I noted to myself that a prevention approach was not high on anyone’s list. I was eventually contacted about a month after she was moved, with the presenting problem as arguing behavior (also termed “mouthy behavior,” although the foul language had not yet set in). There were also two reports of major screaming bouts, but the overall impression of the new foster mother was that Polly was manageable and not unlike other foster children she has had.

However, shortly thereafter, and before my first scheduled visit, Polly’s level of tantrums had increased. Of even more significance, her inappropriate language was back in full force. Its form took on a new and ugly twist, one that frightened this foster mother tremendously. She was now threatening to do bodily harm to the foster mother and her toddler grandson, who would come to play at her house (e.g., “I’m going to hurt you” or “I’m going to hit him”). She had not acted on these threats, but I obviously wanted to address this before it reached that stage.

Let us say that the foster mother just told Polly, after a request, that she could not go outside and play (for whatever reason). Polly may not cry like other children. Perhaps her conditioning history taught her that such was not a fruitful approach to getting what you wanted. However, saying, “I hate you, you are a ——— and I am going to ——— you or your baby” was quite the opposite. What would you imagine would happen next? A probable reaction from an adult upon hearing such a threat from a young child is to ask the child why she would want to do that. For example, a parent might say to the child, “Why would you be so mean?” Or, “You don’t mean that, do you?” This would be followed by Polly saying, “Because I hate you,” or, “Oh yes I do.” This subsequently escalates the fear (increases the value of terminating such statements) on the part of the parent, as the child shows that she can be vindictive and spiteful (apply contingencies aversive to the foster parents). Their reaction to this type of behavior promotes the continued verbal assertion on the part of the child that she does mean it, or will do it, or is not sorry for saying that, and so forth. Again, in Polly’s case, this dynamic between child and adult only increased everyone’s belief that she has no conscience and may never develop one and, therefore, is someone to be feared. Of course, such a behavior probably will cut down on any future tendencies on the part of the parent to deny a desired activity or event to Polly.

I perceived my first job in this circumstance as one of teaching the foster mother to place these threats in context. This is not to say that I do not take these threats seriously. It was my contention that attempting to reason with her at that point in time only fuels the fire! What is needed from the parent is a refutation of such behavior as unacceptable, with a consequence for saying such spiteful things. I immediately set up a behavioral program, targeting threats with an early bedtime contingency. Each time she threatened, she was told that such statements were unacceptable in a firm manner (not inquiring as to why she would say something like that). The incident will be marked on the board in the living room. If she had more than three marks at the end of that day, then it was early bedtime that night (hey, it worked before, let’s keep using it).

How did this plan work? The data on the frequency of threats was the following (plan began on 6/7). Please note, that without baseline data being taken, data in Table 1.10 probably represent an immediate overall decrease in this behavior over the previous week.

As one can notice, Polly developed amazing self-control in her ability not to go over three threats in a day (maybe she developed a conscience each day as the day proceeded). Subsequently, the program was then changed to allow only one incident of threatening behavior before the early bedtime contingency was enacted. This produced the desired change in the behavior as weeks went by with no threats to family members at a month’s follow-up. The board had actually been put away, but the foster mother was instructed to initiate the planned contingencies again should threats develop again.

Fix One Problem, Get Another!

However, as is often the case, as one problem gets solved a new problem develops “critical mass.” It began innocuously as Polly was asked by the FFA social worker if she wanted to see her brother and sister. Polly replied in the affirmative. Following that meeting, her behavior slowly escalated to refusal of all requests during the day, in the home. The reason offered (by the social worker) for this acting out contagion was that during the same meeting, Polly had dealt with sensitive abuse issues. Therefore, her behavior was a reaction to this “bringing up” of traumatic issues. In my experience, traditionally trained therapists often attribute the existence of problem behaviors to prior traumatic events, even after many years.

I inquired further, not feeling that this explained why Polly was being defiant. I would rather look for environmental factors that make such a behavioral pattern more functional in the current environment. I was more intrigued by the fact that this escalation coincided with her having a proposed visit with her siblings. Upon receiving this information, I first conjectured that maybe Polly did not want to see her siblings. Certainly acting out would be a functional way to get such a visit postponed (due to her bad behavior). However, I discarded that hypothesis when the foster mother relayed some information about what transpired after the talk with the social worker. She did not think this information was important at the time but now thought that it might help me in my quest to figure this out. The foster mother told me that Polly, subsequent to the meeting with the social worker, inquired over a several day period about the visit and wanted to know when it would happen. My hypothesis regarding her attempting to get out of the visit did not seem plausible. Rather she seemed eager to go. Then it hit me!

With the mention that she might get to see her siblings, Polly, being the 5-year-old that she was, translated that in the following 5-year-old’s English—“You absolutely positively will get to see your brother and sister; we have them en-route right now and they will be arriving in 5 minutes!” If you believed that your visit with your brother and sister was imminent, would you be upset if 15 minutes goes by and nobody arrives? How upset would you be when 1 whole day passes, 2 days, and so forth? Polly knows one powerful way to influence adult decisions. If you do not get your way, then utter profane words at someone in authority.



Frequency of Threats



















Another incident seemed to confirm this hypothesis. One day, Polly asked her foster mother if she could go to her friend’s house after school. The answer she got was not a “yes,” and not a “no,” but a “we will check it out” (i.e., “we will get back to you on that”). Well, once again that got translated into 5-year-old English, which probably went this way—“Oh what the heck. Just take the bus home with him tomorrow and I’m sure his parents will not mind.” The next day the foster mother gets a call from the parent of Polly’s friend, saying that Polly showed up with her son on his bus after school. Subsequent to the foster mother picking her up, Polly engaged in tantrum behavior that day and the following day (remember my hypothesis—when I don’t get my way, I’ll make it so miserable for you that you will have to give in).

In order to teach Polly not to react inappropriately when she wants ­something, she was taught that making a request for some item or activity will usually be honored at some point, provided she complies with some request or instruction from the foster mother. This procedure allowed me to build in a time delay from request to delivery of reinforcer. The type and number of tasks to be done depend on Polly’s request. For example, for a food item such as a piece of chocolate cake, Polly would have to possibly do several small compliant behaviors, such as picking up a few toys, washing her hands, and so forth. For a more extensive request, she may have to do something over a 2-day period. The foster mother was asked to keep track of the number of requests per day that Polly made, with a target goal of five to eight per day. The data for the first week showed that requests ranged from a low of two occurrences per day to a high of five.

The foster mother reported that the strategy was going well. I’m not quite sure she understood the necessity for this approach, and I therefore speculated that it was not being used as often as it should have been. Shortly after reporting that “things” were going well, this foster mother gave social service a 30-day notice on Polly for change of placement. As I suspected, “things” were not going well. The utility of teaching this skill is that if it is reinforced in the social environment (home), then inappropriate ways of getting items and activities will be replaced by this more acceptable and effective method. However, to utilize this, one must implement it at the time that demands are being made. One should not give into demands immediately, that is, without the “work” component, or deny them completely (except for untenable requests).

Case Analysis: What Does Polly Teach Us?

In this case, I was struck by the forecasting of many people in the system. They attributed everything that Polly did, in terms of severe problem incidents, to Polly’s mental disorder. These professionals further prognosticated that such incidents would be repeated over and over again due to her RAD. I observed that this armchair philosophizing had resulted in not getting past theorizing and into the needs of solving the problem behavior. The parents and I were successful on several fronts in the short term when I looked past the predictions of what one could expect from a child with such a disorder. In contrast, I focused on what we would do when she engaged in such inappropriate behaviors. Apparently, children who are diagnosed with RAD can benefit from a contextual analysis of their problem behaviors.

One of the consequences of severe behavioral incidents is the child being moved from one placement to another. Foster children with severe disruptive and challenging behaviors have frequent changes in their residential placements. This occurs from a policy standpoint because the current care provider is not required to house the child. If her behavior is deemed too difficult for the provider to handle, she is moved to another placement. Often this is sanctified in the name of “finding the perfect fit.” But could behavioral contingencies be at work in some of these cases where a child has six placements in 3 years, mostly as a function of the provider giving notice to terminate placement as a result of behavior?

It seemed to me that Polly would continually misbehave until she either got what she wanted (caregiver gives in to appease her) or she moved to a new home (where maybe she would get a better deal this time around). The possibility that the change in placements might be a functional reinforcer did not strike me until I began reviewing the records of her prior placements. The pattern revealed that Polly would behave for a short honeymoon, where the new foster parents probably wondered why the previous parents had so much trouble managing this poor innocent child. However, they soon began gasping for air as this short honeymoon was followed by a downhill spiral of bad behavior and mean remarks.

Why would circumstances start off well and then turn south? Let us speculate that during the honeymoon period, her new foster parents are probably more lenient and likely to provide her with many of the items and activities she wants (hence, no need to act out). However, after a while, the parents begin to tighten the reins, and Polly starts to have problems. Over the long haul, when Polly did not get what she wanted, her misbehavior and hateful comments toward the caregivers would escalate and thereby force her case manager to find a new home for her because the current parents would no longer tolerate her behavior. I remember saying to the clinical supervisor of the FFA that we need to draw a line in the sand. The next placement needs to deal with this cycle of behavior without throwing in the towel at the end and moving her. The only way we are going to solve her problems in the long run is to not allow her misbehavior to alter her placement. What the system has taught Polly (at a very young age) is that extreme disruptive behavior, over long periods, makes you move to another house. From a behavior analysis perspective, the system needs to redirect such contingencies to not support severe misbehavior.


✵In the Behavioral Functions Involving Positive Reinforcement (Access Functions) section of this chapter, a number of hypothetical examples of a functional ­behavior-analytic approach are provided. Examine the case of Milton. Why does Milton pinch other clients? Does staff have something to do with that behavior?

✵Illustrate how a problem behavior (specify the topography or form of behavior) can serve both an access function and then how the same form of behavior can serve an escape function.

✵Provide an example of a direct access function. Now provide an example of a socially mediated access function.

✵If food is an effective reinforcer, what behavior would be involved in a direct access function (i.e., describe the specific form of the DA behavior)? What might a socially mediated access function look like?

✵If getting one’s hand stuck in a glass jar is an aversive event, describe what behavior would be involved in a direct escape function (hint: glass jar is breakable)? What about a socially mediated behavior that would also hopefully result in the hand being removed from the jar?

✵Discuss what the establishing operation for an access functions is. Then contrast the EOs for access functions with EOs involved in escape functions.

✵Delineate how a problem behavior of some topography can be functional under states of deprivation and also under states of aversion. Provide one example of each type of EO (deprivation state vs. aversive condition) and how the behavior produces a contingency that abolishes that EO.

✵What makes a reinforcer a reinforcer for access functions?

✵Explain how a sufficient state of deprivation with respect to length of time playing computer games might be different for a 10-year-old versus a 66-year-old person.

✵What antecedent condition must be present for an escape function to be viable?

✵Read the Contrived Contingencies Produce a Function section of this chapter. What is the relevance of this section of the book?

✵Provide your thoughts on the case of Polly, the 5-year-old child, described near the end of this chapter.


1.I prefer to use the term direct (to be synonymous with the term automatic) to reflect the behavior’s direct effect on the physical environment.

2.Dr. Jack Michael at Western Michigan University proposed a precise terminology that designated the role of motivative variables as (a) an antecedent variable and (b) separate from the role of a discriminative stimulus (Michael, 1988, 1993). Motivating operations (MOs) are environmental events that affect an organism’s behavior by altering the reinforcing or punishing effectiveness of some environmental change (consequence) and the frequency of occurrence of the behaviors that have in the past been associated with the occurrence of those consequences. They differ from the layman’s terms motivation, desire, drive, or want in that they are quantifiable and verifiable changes in the person’s environment that affect the value of a particular outcome for that person. The first articles regarding these variables referred to motivative variables as EOs (Michael, 1983, 1988); however, more recent articles and text have labeled these variables MOs (Laraway, Snycerski, Michael, & Poling, 2003; Michael, 2007). MOs can be divided into two distinct operations (see Table 1.3): (a) EOs refer to the process by which the value of a particular outcome is increased; and (2) abolishing operations (AOs) refer to the process by which the value of a particular outcome is decreased. MOs affect the current rate of behavior by increasing (EO) or decreasing (AO) the value of the outcome associated with that behavior.

3.Clients or students can therefore become very “cognizant” of circumstances where staff/adults who would be successful in encumbering their direct access/escape seem to be unable to perform such a duty at that point in time.


Bailey, J. S., & Pyles, D. A. M. (1989). Behavioral diagnostics. In E. Cipani (Ed.), The treatment of severe behavior disorders: Behavior analysis approach (pp. 85—107). Washington, DC: American Association on Mental Retardation.

Barrish, H. H., Saunders, M., & Wolf, M. M. (1969). Good behavior game: Effects of individual contingencies for group consequences on disruptive behavior in a classroom. Journal of Applied Behavior Analysis, 2, 119—124.

Cipani, E. (1990). The communicative function hypothesis: An operant behavior perspective. Journal of Behavior Therapy and Experimental Psychiatry, 21, 239—247.

Cipani, E. (1994). Treating children’s severe behavior disorders: A behavioral diagnostic system. Journal of Behavior Therapy and Experimental Psychiatry, 25, 293—300.

Cipani, E. (1995). Be aware of negative reinforcement. Teaching Exceptional Children, 27, 36—40.

Cipani, E. (1998). Three behavioral functions of classroom noncompliance: Diagnostic and treatment implication. Focus on Autism and other Developmental Disabilities, 13, 66—72.

Cipani, E. (2008). Triumphs in early autism treatment. New York, NY: Springer Publishing.

Cipani, E, & Cipani, A. L. (2017). Behavioral classification system for problem behaviors in schools: A diagnostic manual. New York, NY: Springer Publishing.

Cipani, E., & Schock, K. (2011). Functional behavioral assessment, diagnosis, and treatment: A complete system for education and mental health settings (2nd ed.). New York, NY: Springer Publishing.

Cipani, E., & Spooner, F. (1997). Treating problem behaviors maintained by negative reinforcement. Research and Intervention in Developmental Disabilities, 18, 329—342.

Cipani, E., & Trotter, S. (1990). Basic methods of behavioral intervention. In E. Cipani & A. F. Rotatori (Eds.), A research annual: Behavior modification in special education (Vol. 7, pp. 137—201). Greenwich, CT: JAI Press.

Greenwood, C. R., Hops, H., Delquadri, J., & Guild, J. (1974). Group contingencies for group consequences in classroom management: A further analysis. Journal of Applied Behavior Analysis, 7, 413—425.

Inkster, A., & McLaughlin, T. F. (1993). Token reinforcement: Effects for reducing tardiness with a socially disadvantaged adolescent student. British Columbia Journal of Special Education, 17, 284—288.

Iwata, B. A. (1987). Negative reinforcement in applied behavior analysis: An emerging technology. Journal of Applied Behavior Analysis, 20, 361—378.

Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Richman, G. S. (1982). Toward a functional analysis of self-injury. Analysis and Intervention in Developmental Disabilities, 2, 3—20.

Iwata, B. A., Vollmer, T. R., & Zarcone, J. H. (1990). The experimental (functional) analysis of behavior disorders: Methodology, applications, and limitations. In A. C. Repp & N. N. Singh (Eds.), Current perspectives in nonaversive and aversive interventions with developmentally disabled persons (pp. 301—330). Sycamore, IL: Sycamore Publishing.

Keller, F. S., & Schoenfeld, W. N. (1950). Principles of psychology. New York, NY: Appleton-Century-Crofts.

Laraway, S., Snycerski, S., Michael, J., & Poling, A. (2003). Motivating operations and terms to describe them: Some further refinements. Journal of Applied Behavior Analysis, 36, 407—414.

LaVigna, G. W., Willis, T. J., & Donnellan, A. M. (1989). The role of positive programming in behavioral treatment. In E. Cipani (Ed.), The treatment of severe behavior disorders: Behavior analysis approaches (pp. 55—84). Washington, DC: American Association on Mental Retardation.

Lennox, D. B., & Miltenberger, R. G. (1989). Conducting a functional assessment of problem behavior in applied settings. The Journal of the Association for Persons with Severe Handicaps, 14, 304—311.

Michael, J. (1988). Establishing operations and the mand. Analysis of Verbal Behavior, 6, 3—9.

Michael, J. (1993). Establishing operations. The Behavior Analyst, 16, 191—206.

Michael, J. L. (1982). Skinner’s verbal operants: Some new categories. The Analysis of Verbal Behavior, 1, 1—5.

Michael, J. L. (2007). Motivating operations. In J. O. Cooper, T. E. Heron, & W. L. Heward (Eds.), Applied behavior analysis (2nd ed., pp. 374—391). Upper Saddle River, NJ: Pearson.

Neef, N. A., Shafer, M. S., Egel, A. L., Cataldo, M. F., & Parrish, J. M. (1983). The class specific effects of compliance training with “Do” and “Don’t” requests: Analogue analysis and classroom application. Journal of Applied Behavior Analysis, 16, 81—99.

Sundberg, M. L. (1983). Language. In J. L. Matson & S. E. Bruening (Eds.), Assessing the mentally retarded (pp. 285—310). New York, NY: Grune & Stratton.

Vargas, E. A. (1988). Event-governed and verbally-governed behavior. Analysis of Verbal Behavior, 6, 11—22.