The Cipani Behavioral Classification System

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

The Cipani Behavioral Classification System


✵Students will be able to identify the four major diagnostic categories of problem behavior, and define each in terms of the establishing operation (EO) and functional relationship between behavior and its relevant abolishing operation (AO)

✵Students will be able to identify the two Direct Access (DA) functions, and define each in terms of the EO, and functional relationship between behavior and its relevant AO

✵Students will be able to identify the three Socially Mediated Access (SMA) functions, and define each in terms of the EO, and functional relationship between behavior and its relevant AO

✵Students will be able to identify the four Direct Escape (DE) functions, and define each in terms of the EO, and functional relationship between behavior and its relevant AO

✵Students will be able to identify the four Socially Mediated Escape (SME) functions, and define each in terms of the EO, and functional relationship between behavior and its relevant AO

✵Students will be able to identify the factors involved in deciding the function and diagnostic category of problem behavior, that is, reliable relation between problem behavior and relevant outcome under a given EO, as well as the efficiency of such behaviors relative to other behaviors in producing such an outcome

✵Students will be able to delineate how a given topography or form of behavior can be multi-functional

Chapter 3 Behavior Analysis Certification Board (BACB) Task List

4th edition

5th edition

✵I-05 Organize, analyze, and interpret observed data

✵I-06 Make recommendations regarding behaviors that must be established, maintained, increased, or decreased

✵I-02 Define environmental variables in observable and measurable terms

✵F-6 Describe the common functions of problem behavior.

✵F-9 Interpret functional assessment data.

This chapter presents a function-based diagnostic classification system for target problem behaviors: The Cipani Behavioral Classification System (BCS). There are four major categories in this system, previously delineated in this text in Chapter 1 and historically (Cipani, 1990, 1994; Cipani & Cipani, 2017; Cipani & Schock, 2007, 2011). These are: (a) Direct Access (DA) 1.0, (b) Socially Mediated Access (SMA) 2.0, (c) Direct Escape (DE) 1.0, and (d) Socially Mediated Escape (SME) 4.0. The Cipani BCS delineates 13 different classifications of behavioral functions under these four major functional classification categories. Each of the 13 individual categories contain either a different behavioral function under a specified establishing operation and/or a different manner in which such a function is produced (direct vs. socially mediated).

The Cipani BCS is a classification system for behavioral functions; it does not categorize forms of behavior or “symptoms.” How is a function-based BCS differ from a more traditional classification system for problem behaviors? The following section identifies the basic characteristics of a function-based classification system.


The characteristics of a function-based diagnostic classification system are the following (Cipani, 1994):

✵Diagnosis of behavior problem characteristics, not child characteristics

✵Prescriptive differential treatment derived from a differential diagnosis

✵Assessment data collected provides information on context variables, not just rate of behavior

✵Assessment phase concludes with diagnosis phase, in which a function-based category is selected that best fits the problem behavior’s putative function under the specific EO

Diagnose Behavior, Not Client

A function-based diagnostic classification system examines the contextual nature of the problem behavior. It does not presume that the exhibition of behavior is driven by characteristics inherent in the client or child. This sharply contrasts with the current psychiatric approach to diagnosing client behavior (e.g., Diagnostic and Statistical Manual of Mental Disorders [5th ed.; DSM-5]). In a function-based diagnostic classification system, the form of the behavior (in many cases) does not dictate a particular function.

Let us say you have three different children with whom you are involved as a behavioral consultant. Each child engages in a topographically (form) different set of target behaviors. In the traditional psychiatric diagnostic system, each child may receive a different diagnosis because the form of the problem behaviors is different. Because the behaviors or symptoms are different, their presumed cause is assumed to be different. In contrast, a function-based classification of the problem behaviors may reveal that the problem behaviors exhibited by these three children are similar in function, even though topographically dissimilar. Therefore, the classification of these behaviors function using the Cipani BCS could be the same. Using this hypothetical example, let us presume that the functional behavioral assessment obtained data that indicates that all three children’s behavior produces the same reinforcer (i.e., access to tangible reinforcers such as preferred activities). Hence, the function-based diagnosis for all three sets of problems might be subsumed in the same major category. Therefore, despite the obvious individual differences between these children, the prescription for behavior-analytic treatment for this problem area will be similar in composition.

Prescriptive Differential Treatment

As just alluded to, a function-based diagnostic classification system also has implications for differential prescriptive treatment. The behavioral intervention designed takes into consideration the function of the problem behavior(s). It is the case that different topographies of behavior displayed by a given client can produce the same behavioral contingency if their function is the same. Here is an example.

A child diagnosed with a DSM-5 criteria of Oppositional Defiant Disorder is referred for a number of behavior problems. These problem behaviors include aggressive behavior toward residential staff, noncompliance, and running away from the facility. Aggression, noncompliance, and running away are topographically dissimilar behaviors. Does that mean that different behavioral contingencies should be invoked, depending on which behavior is exhibited at a particular time? Not in the least! If all three behaviors are found to serve the same environmental function, the treatment program would specify the same contingency across the occurrence of any of these behaviors. In other words, irrespective of which behavior occurs, the contingency it would produce (according to the functional treatment plan) would be the same. However, if noncompliance and running away serve a different function than the child’s aggressive behavior, there would be one treatment contingency for aggression and a different contingency for the other behaviors. The specifics of the behavioral treatment are driven by the function of behavior, not by the traditional diagnosis (such is actually irrelevant for functional treatment).

The same behavioral intervention would also apply if different clients or students engage in different forms of problem behaviors, but such behaviors serve the same function. Here is an illustrative example. A child named Susanna, diagnosed with conduct disorder, hits the teacher. A different child named Billy, diagnosed with oppositional defiant disorder, throws a tantrum during class and refuses to follow even simple directions. A third child, Raul, who is diagnosed with attention deficit hyperactivity disorder (ADHD), cries and throws a tantrum when he is prevented from running out of the classroom. Professionals who use a traditional diagnostic classification system presume that each of these three children exhibit different types of behaviors because of their different diagnostic classification. Susanna hits her teacher (and others) because of her affliction with conduct disorders. Billy throws a tantrum because he has oppositional defiant disorder. Raul cries and throws a tantrum because he is impulsive and is incapable of delaying gratification because of his ADHD. If there are three different disorders, should there not be three different treatments, that is, one for ADHD, one for conduct disorders, and so on? Our answer: Not if the different types of behaviors serve the same function (Cipani, 2014)!

Let us say that Raul’s tantrum behaviors and Susanna’s hitting behavior occur when they are denied access to a preferred activity. Both problem behaviors often result in getting the desired activity for each child. Given the same function, the behavioral treatment would be the same. If Billy’s topographically dissimilar problem behaviors are found to produce escape from the task, then the treatment would be designed that addresses that function. But it would be different because of the different function involved (i.e., access vs. escape), not the hypothesized mental disorder! Research studies in applied behavior analysis have failed to demonstrate that certain behavioral procedures work only for children with conduct disorder, or oppositional defiant disorder, and not for those with ADHD, and vice versa.1

In summary, classifying problem behaviors according to environmental function does make a crucial difference in the design of a functional behavioral treatment. In contrast, differential diagnosis using the traditional psychiatric Diagnostic and Statistical Manual of Mental Disorders does not prove fruitful in determining functional behavioral treatments based on syndromes of behaviors.

Assess Context Variables

It should now be clear that a useful Functional Behavioral Assessment is concerned with more than just determining the rate of the target behavior. Assessment is driven by the need to determine the environmental factors that are present when the problem behavior occurs, that is, the social and physical environmental context. The effects of both antecedent and consequent conditions of the problem behavior need to be considered. Understanding the role of the client’s antecedent motivational condition at the time makes for a clearer picture of why certain consequent events function as reinforcers at those times. Understanding how certain people may be discriminative for such functions is also important (i.e., socially mediated functions).


I was involved in a case consultation years ago where a student who was attending a nonpublic school for emotionally disturbed children was being discussed at an interdisciplinary team meeting. Both school and mental health professionals were in attendance at this meeting. Although no actual behavioral assessment data were reported on specific target behaviors, the school and residential staff indicated that his behavior had worsened. They ascribed it to a litany of possible reasons. The reasons ranged from wrong diagnosis (e.g., “I don’t believe he is schizophrenic. I think he is bipolar!”) to blaming his dysfunctional family and his home visits. About 45 minutes went by without any pertinent discussion about specific target behaviors in the classroom. As the meeting was winding down, one of the teachers asked, “OK, what do we do when he acts up?” Somebody volunteered what appeared to be a solution, “Let’s use time-out.” Why was time-out being recommended? Was it important to comprehend that he was from a dysfunctional family, according to some experts? Are students who come from dysfunctional families best treated with time-out? Was it because time-out works best with manic-depressive children (but not, apparently, schizophrenia)? Nothing in the prior discussion had any relation to discussing the behavioral reasons for this proposed treatment. Nor was the rate of target behaviors presented. What was also missing was an analysis of the context under which these target behaviors occur. Further assessment of the contextual nature of the behavioral problems was required.

Assessment Phase Concludes With a Differential Diagnosis Phase

In the early development of the field of applied behavior analysis, the design of behavioral treatments often did not include a diagnostic phase. The collection of behavioral data often would lead directly to the formulation of a behavioral treatment regimen. In part, this occurred because the word diagnosis was associated with the mental health diagnostic system, which proved useless in the design of behavioral treatments. Given the lack of utility of a traditional diagnosis in prescribing functional behavioral treatments, many early behavior therapists/behavior analysts have often worked with only two stages of service (i.e., assessment phase and subsequent design of treatment).

We believe that the Cipani BCS serves an important intermediate step. Such an intermediate phase between assessment and intervention would provide for a more guided and deliberate approach to behavioral treatment selection.


For people naive to behavior-analytic formulations, it may appear that one simply identifies a consequence for a selected target behavior. The selected consequence should be capable of ameliorating the level of the target behavior problem when applied as a contingency. Therefore, the only technical skill needed is to pinpoint the referred problem in observable terms and follow it with an effective consequence. For example, if a referred client is disruptive in an enclave-work environment, first one defines the disruptive behavior in observable terms. Then the professional identifies a consequence to follow the target behavior and specifies the treatment contingency. If this contingency does not work, the professional would select another consequence and design another behavioral treatment contingency.

Functional behavior-analytic treatment stems from hypotheses about the function or purpose of the target behavior. These hypotheses (classifications or categories of function) are generated from the collection and examination of functional behavioral assessment data. A function-based diagnostic classification of the problem behavior results from the clinician reviewing the information gathered during the functional behavioral assessment process. Such a process is analogous to procedures used in other science-based fields.

For example, in medicine, during an office visit, the physician gathers information about your current medical condition by asking you a series of questions, called a “diagnostic interview.” She or he then obtains other information, possibly from a physical examination as well as individual tests run on medical equipment. The physician then analyzes all the data obtained relevant to your condition and hypothesizes about the cause of your current medical problem. The physician then makes a differential diagnosis. This diagnosis subsequently allows the physician to prescribe a treatment based on that diagnosis. This treatment prescription makes “good sense” given this diagnosis. The treatment for a diagnosis of flu is different from a treatment prescription for a diagnosis of whooping cough. We believe it is just as important in many circumstances to determine the function of the problem behavior to ensure that the behavioral treatment being prescribed is functionally related to the problem behavior.

The remaining text in this chapter will delineate the 13 functions in the Cipani BCS (see Table 3.1) and provide examples of each classification function. For diagnostic criteria for each of these classification categories for use in school settings, the reader is enjoined to read Cipani and Cipani’s (2017) diagnostic manual for the Cipani BCS.


Cipani (1990, 1994) previously delineated four major diagnostic categories for classifying the environmental function of problem behaviors. The previous editions of this text utilized the same framework in the delineation of the function-based classification system (Cipani & Schock, 2007, 2011).

The following major category functions constitute this classification system:

(1.0) Direct Access

(2.0) Socially Mediated Access

(3.0) Direct Escape

(4.0) Socially Mediated Escape

Within each major category, subcategories are offered. Each subcategory contains the basic characteristics inherent in the major category. The subcategories within each major category are:

✵DA 1.0 Functions

○DA 1.1: Immediate Sensory Stimuli (specify type and location of stimulation)

○DA 1.2: Tangible Reinforcers (specify tangible or class of tangible accessed)

✵SMA 2.0 Functions

○SMA 2.1: Adult Attention (specify which adults/staff)

○SMA 2.2: Peer Attention (specify which peers)

○SMA 2.3: Tangible Reinforcers (specify tangible or class of tangible accessed)

✵DE 3.0 Functions

○DE 3.1: Unpleasant Social Situations (specify situation[s])

○DE 3.2: Lengthy Tasks/Chores/Assignments (specify length of task)

○DE 3.3: Difficult Tasks/Chores/Assignments (specify task that is difficult)

○DE 3.4: Aversive Physical Stimuli/Event (specify aversive physical stimulus)

✵SME 4.0 functions

○SME 4.1: Unpleasant Social Situations (specify situation[s])

○SME 4.2: Lengthy Tasks/Chores/Assignments (specify length of task)

○SME 4.3: Difficult Tasks/Chores/Assignments (specify task that is difficult)

○SME 4.4: Aversive Physical Stimuli/Event (specify aversive physical stimulus)



Category: DA 1.0

Given a deprivation EO, the abolishing operation (i.e., delivery of positive reinforcer) occurs directly through the problem behavior or at the end of a chain of behaviors. In much of the professional literature, this category of behaviors has been referred to as automatic reinforcement (Vaughn & Michael, 1982). Our approach is slightly different from the currently in vogue use of the term automatic reinforcement. The use of the term direct access (and the latter term, direct escape) removes some of the possible misinterpretation of the term automatic. It borrows such terminology from Skinner’s verbal behavioral writings in describing verbal and nonverbal behavior (Skinner, 1957; Vargas, 1988).

We believe there is an advantage to the use of the term direct, in contrast to automatic reinforcement. Using the term direct is defined as behavior that contacts the nonsocial contingency (i.e., physical environment) directly, within a 0.5-second time interval. In other words, the reinforcing event occurs within 0.5 second of the terminal response occurring. If a behavior is maintained by an event that occurs subsequent to this time frame, it is probably not in this category. Further, not all DA functions involve stereotypic behaviors that produce sensory reinforcement as the maintaining contingency (see subcategory 1.2).

There are two subcategories serving DA functions: DA 1.1 Immediate Sensory Stimuli and DA 1.2 Tangible Reinforcers.

DA 1.1: Immediate Sensory Stimuli. Probably one of the greatest advancements in treating difficult problem behaviors exhibited by persons with severe disabilities, such as stereotypic behavior, is the understanding and demonstration that such behaviors may not be socially mediated (Caudery, Iwata, & Pace, 1990; Iwata, Dorsey, Slifer, Bauman, & Richman, 1982; Mason & Iwata, 1990). Problem behaviors in this category produce immediately (often within 0.5 second of terminal response) the sensory event that maintains such behavior. For example, persons who engage in frequent repetitive movements such as hand flapping often do so because of the sensory result such repetitive behaviors produce. Flapping their hands in a certain fashion produces a kinesthetic result that automatically creates its own built-in reinforcer.

Here is another example. Try rocking back and forth in a chair, and note the rhythm produced. Such a chain of behaviors produces a built-in sensory event. This built-in event can function as a reinforcer for the “rocking in the chair” behavior. Such an automatically produced result becomes more desired by persons who cannot create many forms of sensory stimulation due to their cognitive impairments. In contrast, nonhandicapped people are able to watch TV, read a book, listen to music, go on swings, and engage in other solitary activities on occasion. But for nonhandicaped people, occasionally engaging in such activities does not consume one’s entire purpose for the day. However, for persons with disabilities, such stereotypic behavior pervades the entire day and conflicts with alternate essential daily activities and behaviors.


SMA or DA lecture


After reviewing the narrated PowerPoint lecture entitled, “Is It SMA or DA?” delineate the test procedures that can be used to isolate the function of behaviors that appear to be DA, but could also be SMA functions.

For the select target behavior of hair twirling, include in your paper the following (in this order):

1.Delineate the procedures for the test condition involving DA function (behavior that produces what contingency).

2.Delineate the procedures for the test condition involving SMA function (behavior that produces what contingency).

3.Generate hypothetical data in a table or graph that displays data for three session for each test condition (six total sessions) in a multielement design, Then explain in text what the results indicate, that is, what function seems verified by the hypothetical data.

Table DE 3.1 depicts the relationship between the sensory event being in a relative deprivation state (see first column A, Establishing Operation) and the occurrence of the self-stimulatory behavior (second column). Note that the maintaining contingency or abolishing operation is the direct result the behavior produces (third column). Any effect of the behavior on the social environment is inconsequential. The staff or teacher may admonish the client or child after the behavior. However, such a social event is not the purpose of this behavior, even though it may be a reliable contingency of the behavior. It is the sensory result of the stereotypic behavior that will make such a form of behavior occur again when a sufficient state of deprivation is present. Sensory events that can be produced directly through various behaviors can include auditory, visual, tactile, gustatory, and olfactory stimuli.

Many ritualistic (stereotypic) behaviors appear to produce their own immediate kinesthetic reinforcer. Such behaviors often occur independent of the reaction from the social environment. Addictive behaviors, such as smoking and drinking, may also be maintained as a result of their immediate sensory effect. Although this may not describe how such behaviors were developed in the first place, it can explain why such behaviors are still maintained even when the original purpose (EO) is no longer operable. Certainly, for smokers at certain age ranges, social reinforcement and peer acceptance of smoking has diminished tremendously in the last several decades. Yet some people continue to smoke despite the absence of peer reinforcement (and known health risks). Why? For smokers who began the habit years ago, the act of smoking often has its own built-in reinforcer. Table 3.2 provides some questions to consider when entertaining DA to immediate sensory stimuli as a hypothesis.

In Question 1, you should consider whether a one-to-one relationship exists between the behavior and the sensory event. What is the desired sensory event (Question 2)? In some cases, it may be easy to determine the probable sensory reinforcer. For example, someone who sings in the shower is probably reinforced with the auditory result of vocal production in an enclosed area. However, in other cases, it may be difficult to discern what automatic sensory event is the reinforcer. If necessary, an experimental analysis that presents conditions where the sensory effect is ameliorated or eliminated may be the most effective manner of determining such.


1.Is there a reliable (every single occurrence) relation between the specific behavior and immediate production of the hypothesized desired sensory event? What is the specific form of the behavior?

2.What seems to be the auditory, visual, tactile, kinesthetic, or gustatory sensation that maintains contingency?

3.What evidence supports the hypothesis that a sensory event is the maintaining contingency? Does this behavior occur across a variety of situations when the client is in a deprived state relative to the sensory event? In other words, does the client engage in this behavior (usually) irrespective of context? Will it occur in the absence of people?

Question 3 addresses the evidence that points to a hypothesis of DA to sensory reinforcement. If the problem behavior is sensory reinforced, it will often occur across most contexts (possibly excluding contexts where punishing consequences have affected the rate). For many clients who engage in self-stimulatory behavior, they engage in such behaviors irrespective of context. If a client engaged in hand flapping only in the context of certain staff, and not others, such a pattern of stereotypic behavior would seem to point to an SMA function. Perhaps some examples of subcategory DA 1.1: Immediate Sensory Stimuli will clarify DA functions representative of this category.

Talking Is Good? I observed a client at a residential facility for persons with disabilities talking loudly to herself while pushing a laundry cart across a parking lot. I looked around and did not see any staff or other people. How can anyone talk or engage in conversation when there is not an audience with which to interact? A traditionally trained mental health professional may surmise that such vocal behavior is directed at an imaginary person. They contend that the client is carrying on a conversation with someone in her mind, who is responding back to her. Although one cannot see what is transpiring inside her cortex, these professionals assume the existence of such on the basis of the carried on conversation. Hence, the client might be hypothesized to be experiencing auditory hallucinations during those incidents.

Is there an alternate explanation? Can such a behavior be maintained in the absence of an audience? Vocal discourse may produce its own reinforcer, that is, auditory (hearing one’s voice) or kinesthetic effect on vocal cords. Certainly, this phenomenon is present in infants who babble for long periods, testing out their newfound ability to generate a sensory effect by vocal production. As I observed this person pushing the laundry cart by herself, that certainly seemed a plausible hypothesis.

How can one determine whether sensory reinforcement is maintaining the client’s vocal behavior? If this behavior occurs when staff are not physically present, and such behavior does not result in their subsequent attention, its function does not seem to be socially mediated. In contrast, if particular staff attend to her in certain ways upon hearing her talking to herself, then subcategory DA 1.1 is probably not an accurate diagnosis for this client. Further evidence of a socially mediated function would be the inability of other behaviors that are in the client’s repertoire to get staff or adult attention when it is available.

Failure to diagnose this as a behavior maintained by sensory reinforcement when indicated may lead to the design of an ineffective treatment strategy. If the vocal discourse produces a DA function, a treatment strategy that involved “ignoring the behavior” would have no effect at all on the rate of this behavior. Intervention of problem behaviors within this category should be aimed at “weaning her off” of sensory reinforcement or bringing it under control in certain contexts that are not as public (see Chapter 4).

The Voices Make Me Laugh. Here is a similar real-life circumstance. A man, for no apparent reason, would occasionally just burst out laughing as if he had just heard the funniest joke in the world. During an intake at a psychiatric center, he told the professional conducting the assessment that he heard voices in his head. These voices would tell him funny stories and jokes. As a result, this individual was diagnosed with schizophrenia, with “hearing voices” a prime reason for such a diagnosis.

Does he really hear voices? Many mental health professionals would say “yes,” explaining that such voices are a product of his hypothesized neuro-chemical deficits. His mind makes him actually hear voices. Of course, no one can absolutely posit that he does hear voices because the only person who can accurately determine such is this man. The only data regarding this phenomenon is his self-report, which has obvious problems with validity. Is it possible that such a behavior, laughing independent of social input, could be a behavior maintained by reinforcement? As he became a client at the Community Re-Entry Project at the University of the Pacific, Keven Schock and his colleagues would have to answer that question.

How would an analysis of environmental function proceed with such a behavior? First, it was not their intention to actually try to figure out if some internal auditory stimuli were present (currently impossible to deduce). The primary goal was to determine whether the behavior could be socially mediated. To rule out staff (or peer) attention as a possible maintaining contingency, data was collected on the laughing incidents under two conditions: (a) how often he burst out laughing when there were people close by and (b) how often he laughed when he was too far away from people for them to notice him laughing and interact with him. If adult attention were maintaining such a behavior, one would expect to see him engage in this laughter when he was near people, but not involved in social interactions. In addition, if he did engage in such a behavior when away from people, such behavior would recruit their attention.

Therefore, they examined two related sets of data under each condition. First, they examined the number of times he laughed and the total number of minutes he laughed. They found that in this person’s case, if he was left alone and not interrupted, the probability of this behavior was far greater than in the presence of people. In addition, it was noted that such laughter was not frequently resulting in staff attention. In fact, he was perfectly content to be alone while bursting out into laughter. Given this information, it seemed unlikely that his outbursts of laughter were maintained by social attention. This caused them to suspect that for this person, laughing itself was reinforcing.

Could laughing without someone providing a comical situation be a behavior that produces reinforcement in the absence of people? It is similar to the process that many of us go through related to singing in the shower or talking to ourselves. When we sing in the shower, there is generally no one else present to provide attention. Singing in the shower also does not terminate or avoid some aversive situation. Further, most of us would not say that a voice in our head told us to sing in the shower. Therefore, we are left with the only possible function being self-stimulation. So whether it is laughing, singing, or talking to oneself, try to discriminate where you do these self-stimulatory behaviors, but once you are alone, go ahead, I will not tell!


What would you surmise would be the conditions under which laughing at inappropriate times would be maintained by social attention? How would you conduct a trigger analysis of such an antecedent EO (i.e., lack of attention)?

Time-Out Does Not Work? A landmark study in the field of applied behavior analysis involved a study of the effectiveness of time-out procedures with stereotypic behavior (Solnick, Rincover, & Peterson, 1977). It provided evidence that some behaviors produce their own reinforcer. A 6-year-old girl with autism engaged in tantrum behavior. The initial assessment did not reveal the function of this behavior. (Remember this was 1977, and the state of practice and research was not yet investigating behavioral function of target problems.) The researchers tested the effectiveness of time-out on tantrum behavior during instructional sessions. In contrast to a baseline, the results were surprising. Time-out did not decrease tantrum behavior. How could that be?

The researchers noted that when this girl was sent to time-out, she engaged in stereotypic behavior, which consisted of weaving her fingers in a pattern. Such behavior was obviously discouraged during instruction. Hence, time-out was the perfect opportunity to engage in such a behavior without impediment. Solnick et al. (1977) then conducted a second study to empirically demonstrate why time-out did not produce a change in tantrum behavior. In the follow-up study, time-out was still a contingency for this girl’s tantrum behavior during instruction. However, the efficacy of time-out was investigated under two different conditions. In one condition, the stereotypic behavior occurring in time-out was left unfettered. In a second experimental condition, such behavior was prevented by immediately restraining her hands upon the initial hand weave. Therefore, in this condition, the hypothesized sensory effect desired was prevented.

The results proved an important point. If time-out resulted in free access to stereotypic behavior (and sensory reinforcement), then the time-out procedure was ineffective at reducing tantrum behavior. However, when such a sensory effect was eliminated, contingent time-outs reduced tantrum behavior. In 1977, these researchers uncovered the possibility that stereotypic behavior was probably automatically reinforced via its sensory effects in many clients with such a problem.

Pica: The Classic Case of Sensory Reinforcement. A male client sticks inedible things in his mouth and gnaws on them (commonly called pica). Many naive personnel may ascribe such behavior of a client to some form of social attention. When asked why the client does that, a staff person responds, “He knows that it upsets me in that I have to go and get it out of his mouth. He likes to see me work.” Putting aside for the moment the employee’s work habits, the purpose of the behavior often has nothing to do with staff. One would not ascribe a baby’s gnawing or mouthing a particular object as a behavior intended to upset his mother or care provider. The same may hold true in this case.

If pica behavior in a particular client is producing its own sensory reinforcer, what could that possibly be? Mouthing of inedible objects produces oral stimulation (probably similar to gum chewing for some people), a sensory event that is the direct result of the behavior. If pica (in this particular case) is serving a DA function, then oral sensory stimulation might be the reinforcer. One should see that such behavior occurs at a consistent rate when left unabated. Attempts to stop it often are met with the client finding contexts in which access to the reinforcer is “under the radar” of the staff. In other words, the client becomes sneaky and stealthy in engaging in this behavior.

In some cases, pica may be socially mediated. However, the form of the behavior is of such a nature that sensory reinforcement should be initially considered when reviewing behavioral assessment data. To rule out social mediation of this behavior, examine the relationship between engaging in pica and SMA to attention via a trigger analysis. Concurrently, also examine the possibility that pica incidents produce SME from relatively aversive events. For example, when the client engages in pica, is the aversive event removed? Is pica a more efficient behavior in getting such a result than other behaviors?

I Like to Bang Against Plexiglas. A hypothetical male adult client with developmental disabilities slaps his hand against the Plexiglas window. Data reveal that this behavior occurs multiple times a day, and he targets the Plexiglas window rather than the stucco wall next to the window. Staff may try to keep him away from the window, but he continues to seek opportunities when they are not as vigilant. When they are not looking, he runs to the Plexiglas window and strikes it with open hand. Staff attribute this client’s destructive behavior toward property to the adult attention he receives upon its exhibition. The staff will scold him and redirect him back to his couch. They will often report, “He likes negative attention. We tell him to stop, but that seems to be what he thrives on, since he goes and does it again at every opportunity.”

If such were true, then removing attention would result in the extinction of window slapping from this client. However, removal of such staff attention does not decrease this daily pattern of behavior. In fact, it only seems to facilitate ad lib access to such and exacerbate the problem. Could it be that such behavior is maintained by the direct result it produces? When this client hits the window, a sound of some decibel level is produced. If such a sound is unique to hitting this particular Plexiglas window, it may be this sensory reinforcer is the controlling variable. Maybe it is not the sound that is created, but rather, it is the unique kinesthetic feel of hitting Plexiglas with one’s bare hand. Hitting the wall would not produce the same sensation, which explains why he targets the window. Either way, a desire for attention is not the “driving force.”


Describe the scenario under which such a behavior as hitting Plexiglas would be an SME function? What would staff have to do when he bangs Plexiglas?

What About Kissing and Sex? Distinctive circumstances that fit this category are behaviors that result in sexual (sensory) excitation. Behaviors such as kissing, fondling, and the act of intercourse involve directly contacting a sensory event that follows such behaviors. The DA contingency function involves the behavior/action producing the putative desired reinforcer within a split second; for example, when the lips make contact, the sensory event is produced. This relationship may produce the maintaining contingency, under the relevant EO of deprivation of such stimulation (i.e., a sufficient state of deprivation is relative to each individual). If the functional reinforcer (maintaining contingency) is sensory excitation, such instances of behavior are best categorized as an immediate sensory stimuli DA function.

But such a sensory result is obviously rendered by another person. Why is such a behavior-environment relationship not considered a socially mediated function? In this behavior—reinforcer relationship, the stipulation that the behavior is maintained by its direct relationship with the reinforcer makes such instances a DA function. The other person is not delivering an event or activity that has no direct relationship to the behavior. For example, if upon a kiss, the other person gave the target individual a pizza, the kiss has an SMA function under the deprivation EO related to pizza access. You can say that this hypothetical person kisses someone to get a pizza when he or she wants one. But kissing at some length and intensity serves a DA function when the most powerful existing deprivation EO is a sexual sensory stimulation.

Of course, SMA functions with attendant EOs can also explain some/many incidents of such behavior when occurring under a social/sexual relationship. Someone’s attention and emotional affection can serve as reinforcers (as well as obtaining gifts, etc. as possible tangible by-products of a relationship), given the relevant EOs. Particularly with high school students (but certainly could apply to older persons as well), having a “boyfriend” or “girlfriend” can produce peer attention/approval and peer status. Therefore, in order to determine the function of a given act, (e.g., kissing someone), the relevant EO would have to be identified.

DA 1.2: Tangible Reinforcers. When you want a drink of juice, you may ask someone to get it for you if you are in his or her home. But when you are in your own home, you walk to the refrigerator, pull out the juice bottle, pour it in a cup, and drink. Such a chain of behaviors is reinforced, under the motivational condition of wanting juice (EO), by ingesting the liquid refreshment. For many people, DA to tangible reinforcers is available to them when they engage in a sequence of acceptable behaviors, such as walking to the refrigerator and retrieving a desired food or drink item. In contrast, some clients who reside in residential facilities with other people cannot simply go get something out of the refrigerator whenever they desire something. Facility staff persons may be under orders to prevent ad lib access to food and drink items by clients. Hence, access to such becomes restricted. As mentioned, simply walking to the refrigerator to get a food item is encumbered by staff, but the desire (deprivation EO) still remains in effect! Perhaps running to the refrigerator when staff are not in the immediate area may then develop as a more functional behavior. Subsequently, a referral is made for someone who runs in the facility and attempts to steal food.

Table 3.3 illustrates the contextual conditions for DA 1.2: Tangible Reinforcer. This DA function is preceded by the existence of a sufficient state of deprivation with respect to an item, activity, or event (first column). This deprivation EO thereby establishes the item as valuable at that point in time. Therefore, whatever behavior(s) is capable of producing the desired item becomes more probable. With respect to this DA function, a chain of behaviors that can result in obtaining the desired item or activity fits this requirement better than other behaviors (next two columns).

Once the item is obtained to a sufficient degree or amount, the prior state of deprivation is abated and the value of the item diminishes at that point. The particular chain of behaviors therefore becomes functional under this deprivation EO and becomes more likely in the future. Note again that any social consequence that may follow this behavior is tangential and is not the maintaining variable or function. The staff may scold the child, place the child in time-out, or engage in other discipline practices. Such practices have nothing to do with the purpose of the behavior. Further, in light of the efficiency of the chain of behaviors to produce reinforcement, such attempts to “correct” the behavior may fail.

If a client wants a piece of chocolate cake and is told “not till after dinner,” pilfering a piece of cake undetected becomes an effective manner of getting cake. Do you have clients who eat their food at mealtime and then begin grabbing the food off their neighbor’s plates, sometimes undetected? If they are adept at pulling this off reliably, such a chain of stealth behaviors becomes strengthened while at the group mealtime. Do they do this for your “negative” attention? Probably not. Your comments regarding their behavior are tangential and inconsequential to the true purpose of this behavior. They would probably be elated if you would turn around and ignore them! They would then have a greater chance to devour a sufficient amount of someone else’s dinner without interruption.

Table 3.4 presents some questions to consider in evaluating a behavior as a DA 1.2: Tangible Reinforcer diagnosis. First, the target problem behavior must have a reliable and frequent relation between the desired tangible reinforcer and its occurrence (Question 1). If it is a chain of behaviors (often the case), it is the last response step in the chain that produces the tangible reinforcer.




1.Is there a reliable, somewhat frequent, direct relation between the problem behavior and getting a desired item or activity? Describe the chain of behaviors.

2.Are the target behaviors more likely to produce the desired items/activities than appropriate behaviors? Are other attempts to access these items or events thwarted? Are the DA problem behaviors thwarted? If so, how often is the client successful relative to other behaviors that may be socially mediated?

3.Does the behavior occur when there exists a sufficient state of deprivation for that client?

4.Are there any special contexts (e.g., during shopping trips, while at home) that make these behaviors more likely?

In addition to the reliable relation between the problem behavior and direct production of reinforcer, you should note whether other behaviors are more or less successful in obtaining the desired event (Question 2). The ability of a chain of behaviors to directly produce the tangible reinforcer under sufficient motivational conditions determines the probability of such behaviors, especially when examining how successful (or not) other behaviors are. Question 3 involves examining the antecedent condition to determine if the relevant state of deprivation seems to exist at the time the behavior occurs. Question 4 attempts to discern if there is any special context for such behavior, such as shopping trips or other occasions where a special tangible reinforcer is available.

Breakfast Is Down the Corner. Stealing is a great example of a behavioral pattern that can be maintained by DA to the desired tangible items. It also is often maintained by social peer attention, as is often the case with juvenile delinquents. Knowing which motivating operation (MO) is in effect can lead to a more functionally derived treatment strategy. A foster family I was involved with had several children, all coming from the same mother. One of the boys (Roberto, who was 9 years old at the time) was stealing money and various things from family members as well as from classmates at school. This was apparently a behavioral pattern he picked up early in life when he was with his biological mother. I was told that when the children wanted breakfast, they were told (facsimile of conversation), “The mini-mart is down the street, get to it.” His biological mother felt her children should fend for themselves, hence his early exposure to shoplifting. Once this state of affairs was uncovered, they were removed from their mother. But you can imagine that with insufficient consequences for being caught, and continued practice, Roberto became quite adept at being a frequent usurper of other people’s possessions.

Prior to my involvement, the foster father had tried many strategies to deter stealing. These included pleading with him, discussing society’s prohibition against stealing, appealing to his better judgment, trying to induce guilt and shame over stealing others’ possessions, and grounding him. It would be essential to design a plan that made stealing items not functional (in terms of keeping items). Additionally, getting desired items should be addressed via appropriate channels. If the behavior is maintained because of the items it directly produces, then a program that addresses that function should designate an alternate venue for getting desired items. Reinforcing the absence of stealing would seem to be the way to go in this case.

The plan I came up with involved planting items around the house in conspicuous spots to monitor stealing. I called this the planted item technique, borrowed from researchers Switzer, Deal, and Bailey (1977). Each day the father would place several items, including money, in designated places (unannounced to Roberto). This allowed his father to systematically track stealing by checking each place.

Roberto was informed of the plan the night before it was to go into effect. If all the items remained in their place at the end of the day and there were no other reports of stealing, Roberto received $0.50 for the day. However, if something was missing, the punishing consequences involved the following: (a) return item(s) stolen, (b) lose the stipend amount for that day, and (c) pay a penalty equal to double the value of the item(s) taken. Note that this plan had consequences for stealing as well as for not stealing. I believe the father also threw in early bedtime as well.

As you can imagine, his rate of stealing went down. Stopping this behavior had a profound impact on Roberto’s relationship with both of his foster parents. Probably one of the nicest outcomes of changing Roberto’s behavior happened on one of my visits to the home. His foster mother reported to me an incident in which she was so particularly proud of him. Roberto had returned some planted money to her, saying he found it (planted item) and that she must have lost it. Now that is how you develop a moral compass, a conscience in a child who lacked an upbringing that instilled such values. About a year and a half later, Roberto was still reported to not have a problem with stealing.

Food Scavenging! Some clients in residential and inpatient facilities have a target behavior listed on their habilitative plan as scavenging. In the case of students with severe disabilities, such may be listed on their individualized educational plan (IEP). Selecting items off the floor and placing them in one’s mouth is a chain of behaviors that can serve a DA function. The food ingestion would seem to be the terminal reinforcing event for the chain of behaviors involved in scavenging. Such scavenging may become more likely after people have finished eating. Again the form of the behavior may change over time as staff become more adept at preventing access to food on the floor by closer vigilance of clients. Successful retrieval of food on the floor may require a quick and stealthy performance to avoid staff detection.

The same behavioral phenomenon exists with some clients who ingest cigarette butts that are thrown on the floor or ground after being smoked. They scavenge areas looking for cigarette butts and will resist staff attempts to physically refrain them from picking up the cigarette butt off the floor. If you believe that this scavenging behavior is socially mediated, try ignoring such attempts to scavenge in a single test session of an in-vivo experiment.2 If the rate stays high or increases, you obviously have not removed the maintaining contingency. You may also receive a “thank you” from the client for allowing unrestricted access to the desired event.

Category: SMA 2.0

Problem behaviors that serve this type of access function obtain the reinforcer through the actions/behavior of another individual. Social attention and interaction from teachers, peers, staff, or parents are all examples of functions in this category. Obtaining preferred food items, toys, or activities can also be SMA functions if such are obtained through the behavior of another person (not directly as previous DA 1.2 category). Within this major category are the following subcategories: SMA 2.1: Adult/Staff Attention; SMA 2.2: Peer Attention; and SMA 2.3: Tangible Reinforcers.


No, not to the extent that there is any empirical evidence supporting such a contention. For example, to say that persons with intermittent explosive disorder are more likely to have aggressive behavior that is functioning to access adult or peer attention is in opposition to the content of this book. The social environment of each individual determines how certain behaviors affect other people as well as the physical environment. One person with this disorder may yell and scream while at work. Other employees leave the person alone at these times. The function may be avoidance of social interaction at times when it is not desired. Unfortunately, frequent exhibition of such behavior will inevitably result in being fired (long-term consequence). In another person with the same psychiatric disorder, the function of verbal and sometimes physically aggressive behavior may be in the context of a spouse, for example, when the spouse talks on the phone to friends and does not pay enough attention to him or her. Such behavior may be maintained by recruiting desired attention.

The role of a traditional mental disorder diagnosis in a function-based diagnostic classification system is irrelevant at best. While such a diagnosis may be useful in other treatments such as medication, its utility in a behaviorally based system is nonessential. It could often be counterproductive if it sways professional personnel from examining behavioral function.

SMA 2.1: Adult/Staff Attention. “He does it for my attention! He even likes negative attention.” Not all problem behaviors function to access attention, as some people would have us believe. But in some circumstances, it is true. Problem behaviors that successfully access teacher, parent, or care provider attention are strengthened when the child or client is in need of such attention. Concurrently, other behaviors in the child or client’s repertoire that are less effective or efficient in producing such desired events become weakened. For children, attention from parents, teachers, care providers, or staff at facilities can serve as the function for behavior (both appropriate and inappropriate) under a deprivation EO with respect to such attention. For clients in facilities, staff attention is sometimes the maintaining variable in target undesired behavior. A descriptive analysis for this subcategory is given in Table 3.5.

Table 3.5 illustrates that behaviors that serve this function occur when the EO for attention is great (column A). This sets up the conditions for the client to be “motivated” to obtain such (given someone who is discriminative for providing attention for some displayed behavior). The occurrence of the behavior (column B) at some frequency or duration produces the desired form of attention (column C), thus making it functional under those antecedent conditions.

The form of behavior that results in adult attention is determined by the specific social environment. For example, in one situation, a smile from a man may evoke eye contact and a smile back from an interested female walking past this man. However, the same smile from this man has no effect on another woman next to whom he sits down. Later on, the same behavior from this man results in a frown of disgust from a married woman in the restaurant. If the smile produces a fair number of acceptable social responses, this man is more likely to smile than exhibit other behaviors when seeking someone’s attention. As a side note, certain elements/stimuli of the social context may develop discriminative properties over this man’s behavior for someone’s attention over time. For example, he may learn to smile only when the female shows initial interest (i.e., makes eye contact with him).

Problem behaviors maintained by adult attention can take many forms, from innocuous minor behaviors such as giggling, to behaviors that cause great disruption, such as severe tantrums, aggression to others, and running away. For attention to be the function, the problem behavior should reliably produce teacher attention under the relevant deprivation EO, and this temporal relationship should be observed. For example, in the face of a child wanting attention, the problem behavior becomes more effective or efficient at getting the adult’s attention than other behaviors, either desirable or undesirable. Table 3.6 specifies factors to examine when considering an adult attention hypothesis.




Contingency (AO)

Absence of attention for a sufficient period of time

Behavior occurs (at some frequency and/or duration)

Adult attention (delivered in some form)

AO, abolishing operation; EO, establishing operation.


1.Is there a reliable, somewhat frequent relation, between the problem behavior and teacher, staff, or parent attention? What is the form of the behavior?

2.Are the target behaviors more likely to produce attention than other more acceptable behaviors?

3.Does the behavior occur in the absence of attention and a sufficient state of deprivation (relative) exists?

An attention function involves a reliable relation between the problem behavior and access to desired attention (Question 1). Although the target problem behaviors may have to occur at some frequency and duration, such problem behaviors are more successful at getting attention than other more acceptable behaviors (Question 2). Further, the target problem behavior occurs in the absence of attention, where a sufficient state of deprivation exists (Question 3). Let us examine possible scenarios of problem behavior representative of this diagnostic category.

I Want Your Attention, NOW! A hypothetical student, Dolly, is referred for consultation as a result of whining and tantrum behavior in a primary elementary grade class. The teacher reports to you that this student is simply immature for her age. Dolly spends part of her day in general education and the remainder in a special education resource room. The general education teacher states that these tantrum behaviors can occur unexpectedly; “whenever Dolly is in a bad mood. You never know when she is in a bad mood!”

You schedule a 50-minute consultation visit and prepare to identify the context under which tantrum behaviors occur. Using direct observation and determining what EO seemed to be present at the time of the “tantrum incident,” you view the following in a session involving teacher lecture and independent seat work. Tantrum behavior did not occur during teacher-delivered instruction. Rather, it occurs during seat work. Dolly would work for a while without any problems. However, when tantrums did occur, the context seemed to involve the following. Dolly would raise her hand to obtain teacher attention. Unfortunately, such requests frequently go ignored for some period of time because the teacher is often working with someone else. Hand raising is not an effective or efficient manner to get the teacher’s attention at these times. This is not making a social judgment, but rather an empirical observation. This situation also leads to a strengthening of the EO for attention relative to other needs. Getting teacher attention becomes an even higher priority with Dolly when this happens. However, when Dolly hits her desk with her hands while concomitantly whining and complaining, the teacher first tells her to stop. However, shortly thereafter, the teacher comes to her desk to “find out what all the fuss is about.” With this offering of help by the teacher, Dolly then proceeds to calm down and begin her seatwork (and the teacher provides her help and encouragement).

When this student wants the teacher to come over to her, what are her options? Raise her hand? Selecting that option results in a longer protracted wait. Bang on the desk? That option produces faster results; hence, that form of behavior becomes strengthened under the conditions of wanting teacher attention.

In this case, there is another question one has to ask and answer. Why is teacher attention in some form of deprivation in the general education class, but not in the resource room (tantrum behaviors not reported to be a problem in this setting). Observation of the two classrooms will reveal the reason. You realize that in this student’s resource room there may be 6 to 11 children at any one time. Therefore, needing teacher attention is not as lengthy a wait as in the general education class. In the general education class, 25 other students have needs as well. Dolly’s access to teacher attention is markedly less. Hence she engages in behaviors that are hard to ignore, but unfortunately will result in her loss of placement in mainstream settings.

I Am on the Phone! How many times have you seen (or been the recipient of) a child screaming, “Mom, I need to talk to you!” while Mom is on the phone or engaged in conversation? Interrupting behaviors occur because they are more successful under those conditions than other more desirable behaviors. The child may tug on Mom’s shirt, but to no avail. However, a loud scream produces several responses from Mom. Let us look at a scenario depicting what transpires that makes screaming an adaptive response for this child when Mom is busy with someone else.


Mom, I need you (in conversational tone while pulling on her shirt).


Wait just a minute; I am on the phone with your sister. She is at the dentist.


(after several minutes go by): Mom, I need to get some juice. I am losing my voice from lack of liquid.


Hold on, I will only be another minute.


(several minutes go by, child screams): Well I will just get it myself. I have to do everything myself. Nobody cares about me!


OK, I am coming (and terminates conversation).

As you can see, the child escalates his or her demanding when minor forms of demands for Mom’s attention are not fruitful (which is very often the case with phone conversations). Ignoring these minor forms then makes the screaming bout at the end more likely, with Mom terminating the conversation shortly and attending to the child. The child’s level of screaming under conditions where attention is wanted has just been strengthened. Of course, many more scenarios similar to this play out over time to develop a reliable contingent relation between the form of screaming and accessing Mom’s attention.

A Pinch Here, a Pinch There. An adult male client with severe mental retardation who has no vocal speech would pinch people as they passed by him. It was easy for me to discern who this client was upon entering the building without anyone pointing him out. As I entered the building I could see him sitting at a table. Nothing striking there, right? Upon observation, adult staff at his day treatment program would near his table area and then move in a semicircle away from him in order to get to the other side. Now you know why it was easy to figure out who this client was without someone pointing him out.

Occasionally, even with staff winding a long arc around his area, he would be successful in pinching people. If someone got too close, he would dart out of his seat and pinch that person. Of course, new persons on the floor had to be warned about such a behavior in order to engage in the protocol for avoiding him. As you could imagine, this “imposed circle of avoidance” made attention even more desired from this client’s perspective. As staff became more successful at avoiding him, the rate of attention from staff became less and less. The more successful the staff persons were at avoiding him on a given day, the greater the deprivation EO became, making attention even more valuable.

To compound matters, staff members were instructed to engage in the following contingency upon being pinched by this client. They were to immediately say “No pinching,” and then proceed to his hand, shake it, and say, “This is what we are supposed to do.” The program designer probably thought that this would teach him how to initiate acceptable social interactions with staff. Instead of pinching, he would offer to shake hands. Wrong on that count. Let us examine all the behavioral contingencies in place for this client. When he simply sits in his seat, people avoid him so that he will not pinch. Nothing in the program design mentions to catch him when he is not pinching and prompt the acceptable behavior! But, when he does pinch, he gets some brief interaction with people in this correction procedure. Pinching successfully ensures him staff attention as the staff adhere to this program. Pinch away!

“I Will Kill Myself!” When someone in a mental health facility makes this statement, his or her environment is guaranteed to change. Protocol requires that a number of staff interactions, interviews, and assessments be conducted to determine the dangerousness of such a verbal statement. Such traditional assessments often revolve around determining whether the person is depressed to such a level that killing oneself is an option he or she might entertain. Of course, all instruments that could be used for committing suicide are made inaccessible. Staff members are intensely vigilant in these circumstances to ensure that such items are not attainable. Such efforts are certainly mandatory in the case of protecting life.

Could an analysis of behavioral function be of utility in separating out those persons whose life is in immediate danger from others? What possible behavioral function could such threats serve? Could such threats to kill oneself be under control of social reinforcers? If so, why would some people need to go to this extreme (stating they are going to kill themselves) to get such reinforcers? Although not discounting the possibility that some people obviously commit suicide because they feel (at that time) life may not be worth living, a functional evaluation of verbal statements may prove useful in designing treatment.

Let us take the case of someone who has just lost a spouse through an abrupt death. In a state of depression, the person contacts a mental health crisis center. Contingent upon saying that suicide is an option as a result of his or her grief, the person immediately interacts with many different professional people (e.g., doctors, nurses, social workers, front-line support staff, and therapists). In a time of need, these professional people are very supportive and caring, and the client possibly begins to reevaluate his or her life in new ways. As this person’s mental health is evaluated as improving, the client sees these people less and less over time. For some people that may not be a problem because they have other people in their social network to return to. People with friends and relatives can be provided the care and support one needs to face difficult life-altering circumstances. But what about those people who have lost the one friend they had? To whom do they turn? If there is no one left, or their relatives live far away, they return to emptiness (in terms of social network). For these people, getting “better” results in a significant decrease in social interactions (adult/staff attention). The better they are, the less contact they have with people.

It is important to address why a functional evaluation is just as essential as traditional diagnostic evaluations of persons in this circumstance. A traditional view begins with the assumption that verbalizing the statement “I feel I might kill myself” is a symptom of depression or low self-esteem. Given that view, the professionals involved will “pull-out all stops” to make this person less depressed. They will probably provide antidepressant medications and increased professional services (including individual and group therapy). What this regimen establishes has both short- and long-term ramifications. First, after receiving such services, the person does feel better, that is, less depressed. That is the short-term result of such statements, which is essential. Additionally, such statements proved to be very effective in recruiting social interaction when desired. If social interaction and attention is less available for the client once released from the inpatient setting, the EO for attention develops over time. Hence, the longer the client goes without social interactions while in the community, the more likely such statements can occur. This process, in effect, establishes a cyclic pattern of improving and worsening. The functional utility of such statements in accessing attention, as well as the person’s loss of attention from professionals as he or she gets better, needs to be considered if we are to develop more effective interventions for people diagnosed with mental illness.

If mental health systems take such a possible function into account, would it replace the current system? Absolutely not! People who say they are going to commit suicide need attention! It is currently impossible to determine with great reliability and precision who is likely to follow through with this threat and when. However, what a functional analysis will add to the comprehensive treatment of these people is the concurrent development of a social network for them when they leave the facility. Perhaps developing conversational skills as well as social trips to local community centers while they are inpatients would be therapeutic. Making new friends and contacts would be a treatment objective. Professionals would spend as much time in developing these clients’ new circle of friends as they would in triaging the presenting complaint. In short, giving such a person many reasons to live is the best long-term approach, one facilitated by a functional analysis of the maintaining contingencies.

Will Steal for Social Attention. A 12-year-old female living in a foster home was referred to Keven Schock. The psychiatric staff were considering putting her on psychotropic medications due to her diagnosis of kleptomania. The referral was to determine if she could control herself in this matter or if it was purely neurological (hence the need for medication as the only effective solution). This young woman would steal daily, often items that were of no real value to her. Among the items reported to have been stolen were a broken pencil, other students’ school work, a kitchen egg timer, and even a pair of dentures! Now most people would assume that the reinforcer for stealing is the DA to the desired item (i.e., DA 1.2: Tangible Reinforcer).

In this case, that diagnosis did not fit. The items she stole were not items used by her. They served no purpose in her life. So why might someone steal items that were of no use to him or her? Keven asked one of the staff persons, “What does she do with these items?” The staff person’s reply was, “She leaves them on her desk in her room. I think she must be psychotic or something. She does not have any sense to hide them. She leaves them where she knows I will find them.”

Could it be that this person desires to get caught? But why would someone want to get caught? Is such a purpose best explained by invoking an existential crisis theory? When Keven asked staff what happened when she got caught, the reply was illuminating. “I punish her by giving her long lectures on how bad it is to steal, but it just isn’t working, I think she has low self-esteem and wants to be punished” (there you go, an existential crisis). Perhaps she is a masochist!

Can an analysis of behavioral function support another explanation (one rooted in some form of reality)? Perhaps she desires attention, in whatever form. Consider this hypothesis. If she had minimal ability to initiate interactions, she might be forced to rely on another behavior to interact with people. Her stealing did not seem to be a function of access to tangible reinforcers. It also did not occur under conditions that would promote escape or avoidance behavior. Rather, it was her most effective way to access adult attention.

SMA 2.2: Peer Attention. This diagnostic category encompasses problem behaviors that gain desired peer attention, particularly when peer approval is a powerful reinforcer, that is, presence of deprivation EO. This function exists more frequently in school settings with adolescents (e.g., class clown phenomenon) but can be operable with other clients where peer attention is a desired event.

In order for peer attention to function as a maintaining contingency for target problem behavior, peer attention has to be available. Given that children, adolescents, and adults can transmit information regarding the display of behavior, peers can either be physically present (e.g., in the classroom at the time of the behavior) or be able to be informed of the incident. But peer attention is not solely a function that is endemic with adolescents. It can also be a powerful motivating condition for adults as well. For example, some adults engage in behavior that recruits peer attention or approval as well (e.g., man or woman who “brags” about sexual conquests to buddies, or brags about other phenomena that recruit approval from peers). Other examples for adults can include the “necessity” to buy certain types of items such as cars, clothing, and so on to achieve or maintain some form of social status (i.e., peer attention).

Behaviors that reliably access peer attention become strengthened, particularly in social environments where peer approval is a powerful reinforcer. Problem behavior maintained by peer attention might be hard to detect because the approval or attention of peers may not occur immediately. In fact, with older children, some problem behaviors are reinforced by peers at a later time when they find out about the misbehavior (e.g., aggressive or disruptive behavior).

This function involves a reliable relation between the problem behavior and the access to desired peer attention. Other more acceptable behaviors are less likely to result in peer attention. Although the target problem behaviors may occur sometime before certain peers become aware of the incident or behavior, the delayed result still maintains the behavior. Teacher or adult attention may occur much closer to the problem behavior but may not be a factor in its maintenance (deprivation EO for adult attention not operable). Very often, an incorrect hypothesis about such behaviors in this subcategory being maintained by adult attention leads to an ineffective strategy. The following examples illustrate this point.

Everyone Loves a Clown (Except the Teacher). An example of problem behavior maintained by peer attention is a hypothetical junior high school special education student, Billy. He is referred for a functional behavioral assessment for his verbally inappropriate behavior toward teachers. Following is a sample of what transpires during an incident of inappropriate verbal behavior during a prealgebra class.


Who can tell me what a linear equation is?


(somewhat covertly, although within earshot of other students) What a bore.


Billy, do you have something to contribute to this discussion?


I said what a core concept! (other students start giggling)


Perhaps you would like to teach this core concept.


What for! (class breaks out laughing, Billy is sent to the principal’s office)

It is not difficult to surmise that such behavior is resulting in peer attention. When Billy makes such comments to the teacher, he gets the attention and approval of his fellow classmates. For a trip to the principal’s office, he now becomes the most popular kid in the class. He has peers wanting to hang out with him during recess and at lunch. Such peer attention and approval can often override the power of designated discipline strategies such as dismissal from class. It is important to diagnose these problems as SMA 2.2: Peer Attention in order to provide effective intervention.

It is important to note that while some people may view the teacher’s attention as the maintaining variable, such is not the case. Without peer attention that followed teacher attention, the level of the problem behavior would be much weaker. This is not to say that the teacher’s response to Billy’s initial smart remark did not provide the perfect forum for this student. It set the stage for peer attention to become more pronounced.

Suppose the teacher attended to other behaviors of this student that do not result in peer attention, such as raising his hand to be recognized. Would such have an effect on inappropriate verbal behavior? Many people, naive to an understanding of behavioral function, would vehemently respond with “Yes.” From what you know now, why would a change in this student’s behavior probably not happen?

Class Clown: It’s Elementary. A hypothetical second-grade girl engages in a number of inappropriate behaviors, such as sticking her tongue out at the teacher (when the teacher is not looking), dropping her pencil, and making faces. Despite the fact that the teacher admonishes her for engaging in such behaviors, the peer reinforcement for such behaviors is far stronger than the admonishment she receives. She becomes the most popular student on the playground during recess. Subsequently, the rate of these behaviors is maintained at a high level as a result of their ability to get peers to attend and laugh with her.

SMA 2.3: Tangible Reinforcer Hypothesis. We have all seen children at the supermarket whine, cry, and throw a tantrum. Are they possessed by demons? Many people will comment that they probably want something from their parents. Very often, they are right. They have assessed the motivational condition correctly. What is also true is that this set of behaviors has been successful in the past in obtaining such desired items or activities. As a result of the adult’s mediation of such undesirable behaviors, they become more probable, and more appropriate behaviors that are ineffective at getting the desired item become less likely. In contrast to the DA category, DA 1.2, the client’s target behavior does not directly produce the item. The undesirable behaviors (at some level or duration) are mediated by a parent providing the item contingent upon the behavior. The delivery of this reinforcer makes such target problem behavior much more likely when the EO for a given item in future shopping trips at the store becomes strengthened.

Therefore, target problem behaviors that function to access a desired tangible item, object, event, or activity under a deprivation (EO) are strengthened (see Table 3.7). Further, other behaviors (perhaps more desirable) in the child’s or client’s repertoire that are ineffective or inefficient at producing such events become weakened.

In many cases, target behaviors functioning in this subcategory are usually misdiagnosed for problems maintained solely by adult attention (SMA 2.1). This can easily happen if one simply watches the first socially mediated event following the target behavior. For example, let us say that a student in a special day class for mild disabilities refuses to line up for adaptive physical education (PE). As a result, the teacher quickly moves until in close proximity to the student and may say something. If one assesses that the first social event that occurred subsequent to the behavior is the maintaining variable or function, one would classify this problem as an SMA 2.1, given a reliable relation between these events. However, after some protracted discussion, the teacher bribes the student to get in line by allowing her to carry her favorite doll while she goes to PE. Now you know the whole story!




Contingency (AO)

Absence of item or activity for a sufficient period of time AO, abolishing operation; EO, establishing operation.

Behavior occurs (at some frequency and/or duration)

Item or activity delivered in some form by someone

The problem in accurately diagnosing behavior problems in this category is that the delivery of the desired event or item will also coincide with adult attention. This is always true because access to the item is socially mediated. However, one should recognize that adult attention is an incidental variable and that access to the item, activity, or event is the “driving force” for the behavior. This can be tested in an in-vivo hypothesis test comparing two conditions. In half the sessions, present only attention and praise for this student when the child gets in line. In the other sessions, provide the desired item if the child gets in line within a certain time. If the item is the power behind the maintaining contingency, guess what condition will result in the child getting in line more frequently?

The problem behavior may have to occur over some period of time before the adult provides the event or activity. In the case of children, the care provider or parent may initially attempt to ignore such behavior and not “feed into it.” The parent refuses to give the child what he or she wants. Such initial attempts at ignoring the behavior do not result in the problem behavior ceasing. Rather, the problem behavior escalates in form and duration. As a result of the child’s behavior worsening, the adult, under pressure (and wanting to now stop the behavior at all costs), gives the child the object or activity. It is essential to detect such a shaping of different exacerbated forms of problem behavior and durations by the parent’s response to both the initial form as well as the more severe form.

Sometimes the presentation of the tangible reinforcer is done in a manner that looks like the teacher, parent, or care provider is reinforcing appropriate behavior (e.g., “If you’re quiet for 5 seconds, then I’ll let you go outside and play.”). The child engages in appropriate behavior and subsequently is given the desired item. However, one should realize that the entire chain of problem behaviors has led to the client’s access to the item. It is not simply a matter of what behavior occurred prior to the delivery of the item! Table 3.8 presents questions to pose when considering this diagnostic subcategory.

This function involves a reliable relation between the problem behavior and the access to the desired item or event (Question 1). Other more acceptable behaviors are less likely to result in the desired item or activity (Question 2). Although the target problem behaviors may have to occur at some frequency or duration, they are more successful at getting the desired item or event than other more acceptable behaviors (Question 2). To reiterate, attention and other events may occur in the interim between behavior onset and delivery of socially mediated tangible reinforcement. Do not assume the events that are temporally closer to reinforcement are maintaining the behavior! Rather, in some cases, it is the ultimate delivery of the tangible reinforcer that is the driving force behind the client’s behavior. Let us examine possible examples of problem behavior serving this diagnostic category.

Elopement. A female client was frequently running naked down the driveway of a residential facility. Staff reported that this occurred at least twice a week. Several staff would follow her and try to talk her into coming back. Obviously being naked will result in adult attention, both from staff and passersby. But such attention alone did not seem to explain this periodic nude display in the neighborhood. The staff indicated that this client was perfectly capable of accessing attention from staff, multiple times during a given day. Further, on a streaking day, her access to attention would be comparable to other days, suggesting that such was not in a deprived condition at the time of the behavior.


1.Is there a reliable, somewhat frequent relation between the problem behavior and SMA to the tangible item or event? What is the form of the behavior?

2.Are the target behaviors more likely to produce access to the item or event than appropriate behaviors?

3.Does the behavior occur under a sufficient state of deprivation, that is, relatively protracted absence of item, event, or activity?

Keven Schock was tasked with determining what happens that makes her engage in such a behavior. She obviously returned to the facility at some point in time. Perhaps how the staff “bribe” her back might provide a clue. Keven asked how they got her to return to the inside of the facility (and put clothes on). Their reply was that they had to offer her a soda if she returned.

Can the access to soda be so powerful that it makes this client shed her clothes and run down the driveway naked? It must be that access to sodas via other means (i.e., behaviors) is pretty difficult. Apparently, simply running out of the facility with clothes on did not set the conditions for getting a soda. The following is a facsimile of an interview with staff to get at this information.


What else does she do to get a soda?


Oh, she is not allowed to have sodas. She is on a restricted diet. She asks for sodas all the time and we tell her that she cannot have one, as it is not on her diet plan.


What does she do when she gets a soda?


Oh she guzzles it down, belches, and smiles.

Can you see that access to a soda, under the rigorous diet plan, becomes a deprived event? Further, acceptable behavior such as requesting a soda is not functional. Hence, the streaking behavior becomes functional under the motivational condition where getting a soda is highly sought by this client. Solving this problem now becomes easy. A quick conversation with the dietitian resulted in her being able to have one soda per day. The results of such a strategy confirmed my hypothesis. Once she had one soda each day, she did not streak down the driveway.

Milton, the Pincher. The stated relationship between Milton pinching people and getting a walk would make pinching more probable when he wants a walk. Hence, pinching behavior would be diagnosed with an SMA 2.3 category: Tangible Reinforcer. We would contend that pinching other people is a functional behavior when Milton desires a walk. Also, unfortunately other more appropriate behaviors are not more functional in Milton getting a walk.


In Milton’s case, how would you determine that the walk is the maintaining contingency and not simply attention? How would you conduct an in-situ treatment hypothesis strategy for attention-maintained behavior to separate these two variables that occur together (i.e., in order to get a walk Milton has to have the staff attend to him)?

I Don’t Like What I Am Wearing! Some students with severe disabilities may soil their clothes, even though they are capable of independent toileting, to get a change of clothes (i.e., change to preferred clothes). Other students learn to rip or tear their clothes for the same purpose (i.e., they will be given more preferred clothing articles to wear once their old clothes are torn). As long as the student gets new clothes contingent on these behaviors, ripping or soiling clothes will continue to occur under conditions where the student desires a change of clothing. Although the teaching staff’s behavior is understandable, it certainly presents you with a plausible function for this behavior. If you diagnose the deprivation EO to be one of wanting to obtain more preferred clothes, the suitable classification is SMA 2.3; Tangible Reinforcers.

I Am Hungry Now. A hypothetical female adult with profound mental retardation exhibits a frequent and sometimes intense form of aggressive behavior toward staff, particularly before lunch. She will hit and attempt to bite them. An examination of the function of such behavior may reveal that such behavior may be maintained by staff providing specific edible items to the client upon such behavior (once she is calmed down). They may even state, “If you calm down, I will give you a cracker or candy bar.” Although some staff may not realize that such a contingency is strengthening aggressive behavior, it is the entire chain of behaviors that is reinforced, not simply “calm behavior.” In order to get to a condition where a candy bar becomes available, aggression must occur first. If aggression does not occur, one cannot proceed with “calming down” after the fact! If this person is calm without engaging in aggression, it is highly likely that no candy bar will be rendered.

To complicate matters, such aggressive behavior may not eventually result in the desired item every time it occurs. The staff may report to you that they do not give her candy or crackers when she is aggressive (forgetting the times when it is used to calm her down). Of course, you realize that delivery of candy and crackers does not have to occur with every incident. Rather, some intermittent delivery of food can maintain such behavior, especially if other behaviors are even more unsuccessful than her aggressive behavior. When she wants the candy item, her best bet is to bite someone!

He Likes to Be Restrained. This real-life case illustrates how a chain of events that begins with a client being restrained for aggressive behavior can end in a tangible reinforcer. Often with clients who frequently engage in behaviors that terminate in restraint, personnel will claim they “like” being restrained. But an analysis of the events that transpire antecedent and subsequent to the restraint may be an eye-opening revelation as to true function.

In the mid-1980s, I was hired as a consultant for a training program for people with developmental disabilities. This program was set up for some clients just coming out of the state institution in a wave of depopulation taking place during that time period in California. These clients had some of the most disruptive and challenging behaviors of the persons leaving the state institution. This agency wanted to serve such persons (given a higher rate of reimbursement for handling such difficult behavior problems). It was not uncommon in this program for a client to have to be physically restrained as a result of potentially dangerous behavior. The restraint would occur at a point where the client’s behavior had escalated so that danger was imminent. The staff were well-trained in this procedure and executed it in a safe and efficient manner when needed. Over time, some clients needed to be restrained only every once in a while. In contrast, other clients were involved in a prone containment just about every day.

One client in particular was brought to my attention. His rate of episodes where he had to be restrained was quite high. The facilities behavioral specialist reasoned that restraint should always function as a punisher, because no one in their right mind could actually like that being done to them. Therefore, the specialist’s explanation of his behavioral episodes, involving aggression and property destruction, was that he was just “going off.” Unfortunately, when he got to this level of aggression and property destruction, there was no alternative except to restrain him. I thought, “Could there be an alternate compelling function, with a different, more powerful EO involved? What could he possibly be getting out of the restraint process so that accessing this tangible item or event would override the aversive aspect of the restraint?” The answer? Coffee!

Here was a typical scenario. As reported by staff, he would be restrained when his state of aggressive behavior reached the danger point. Prior to the restraint, directives to engage in more appropriate behavior were fruitless. It was as if he wanted to be restrained. Two or more staff persons would put him in a prone containment in a safe manner, preferably on a soft surface. As he calmed down, they began a process of releasing him progressively, until he was standing erect again. They would then have him sit in a chair. As a method to calm him down further, he was given lots of attention from staff, soothing comments, and coffee. The proof that coffee was the “smoking gun” was the following piece of information I obtained after viewing this phenomenon with my own eyes. Prior to the incident, if he asked for coffee, he was told he could not have any until lunch because he just had some for breakfast. Now do you see the competing differential consequences? Act up, get restrained; 15 to 20 minutes later, get coffee. Don’t act up, but ask nicely, no coffee.

Potato Chips: Bet You Can’t Have Just One. A hypothetical female client with schizophrenia in a day treatment program engages in property destruction. How does such a behavior relate to getting potato chips? The road to extra potato chips is long and winding, but the payoff is at the end. During snack time, after this client has finished her food, she will sometimes attempt to get up and get more snack items, particularly when the snack item is potato chips. This attempt is rebuffed by staff (i.e., DA behavior is ineffective). They get in front of her and direct her verbally, and sometimes physically, back to her chair. She may then scream and holler, in which case staff move her away from the dining table. She then knocks items off the adjacent table or any other surface that is in need of redecorating. As a consequence for such behavior, she is placed in time-out. Looks like none of these behaviors result in extra potato chips, right? Suppose that after the time-out, sometimes staff give her a half a bag if she was quiet during her time-out (as an incentive not to create havoc while in time-out). Now you see the whole chain of events and how such a behavior as property destruction can serve a food-accessing function. One has to get to time-out in order to have the possibility of payoff. Whatever behaviors result in staff giving her a time-out will become more probable when there is a sufficient state of deprivation for potato chips to produce an EO.

I Like Massages! Keven Schock was working in a brain injury treatment center and was asked to assist with the case of a preteen girl who was being physically restrained daily. In interviews with the staff, they reported that she would typically come home from the on-grounds school, enter the residence, and begin to engage in a series of escalating behaviors. She would verbally taunt the staff and other residents. The escalation of severity of this problem behavior eventually involved self-injury or attacking other residents and staff. At that point, the staff felt they had no other option than to physically restrain her to stop her from injuring herself or others. In Keven’s review with staff of the restraint events, there were no clear antecedent events that were reliable and consistently correlated with the escalating chains of behavior. Likewise, there were no consistent consequent events other than the physical restraint. The function of such dangerous behaviors was unclear to Keven at this point.

Keven started treatment with staff implementing noncontingent reinforcement (NCR) for social attention and tangible items/activities. While the staff were walking home with her, they provided continuous attention; when she got home, the attention continued; she had free access to tangible items and activities. Keven also implemented noncontingent escape (NCE) for all demands in the home; hence staff made no demands for her to complete any tasks once she reached the residence. In other words, Keven removed most of the deprivation and aversive EOs that could possibly be involved. Unfortunately, the treatment outcome was not as comprehensive as hoped for. The effect of this treatment intervention was to prolong by about an hour the need for a physical restraint for dangerous behavior.

In reviewing the restraint events, the staff noted that she would struggle against the physical hold for 5 to 10 minutes; thereafter she relaxed, seemed calm and actually appeared to be happy. After the restraint event she would be cooperative and not have any major problems until the following day when returning home from school. Then the same routine began again with the sequence ending in restraint. Did she “like” restraint, or perhaps some aspect of the physical contact that occurred with the restraint? Keven decided to test the hypothesis that the physical contact was the maintaining event. On the way home from school there was a physical therapy gym. Keven convinced the agency physical therapist to provide her a deep tissue massage prior to her returning home. Keven and the staff were pretty sure we were on the right track during her first massage when she was smiling broadly during the massage. The hypothesis was confirmed when she returned to her home and did not engage in the series of behaviors that resulted in restraint. There were no more restraint events on days she received deep tissue massage.

Category: DE 3.0

The behavioral function of this major category involves particular behaviors that directly terminate (or completely avoid) an aversive event. For example, you feel an itch on your leg, which produces a certain level of discomfort. This antecedent context sets up the motivational condition for some behavior to relieve the aversive state—the itch. You scratch that area in a certain manner that causes the minor discomfort to disappear for a while (termination of aversive event). Scratching the itch brought about its removal directly.

DE functions can often be operable when clients are involved in situations or activities involving task demands. A client or student is presented with a nonpreferred task or demand. This aversive event thereby creates a sufficient state of aversion (i.e., aversive EO) which establishes the value of its removal. Leaving the task area directly produces at least a temporary cessation of the demand or task. What staff do after the behavior will determine whether such behavior will result in temporary postponement of the work task or permanent removal. If staff begin to prevent the client from walking away from the area, another “escape” behavior may take its place. The client may learn that in order to be successful in getting away, one cannot simply walk away because staff will prevent the client from leaving. Perhaps running from the area and down the hall may prove to be more successful in escaping the task. Hence, the client has now turned into a “runner” in order to make it more difficult to be impeded in escape from the task.

Other contexts that may generate DE behaviors are aversive social or physical environments. Many people find noisy areas an intrusion, and given sufficient exposure to such, seek to leave the area by walking away. Leaving the noisy area for a quiet area is sometimes an option for many of us. However, clients or students leaving a noisy area without requesting permission from staff or teachers can constitute a problem in facilities and educational institutions. Such unauthorized leaving will result in staff or teachers preventing such a behavior, thus rendering it ineffective. But for this particular individual, the aversive nature of the noisy area still prevails. Hence the aversive EO will still set up the motivational condition for a function. What behavior takes its place is up to the staff and/or teachers. If an alternate acceptable behavior is delineated as the mechanism to produce escape, then a management problem is averted. Unfortunately, when no decision to select such an alternate behavior is made, other equally undesirable behaviors take its place and may prove harder to obstruct.

Crowded or odorous environments can constitute aversive situations for many people. Problem behaviors such as leaving or running away from the scene can be effective in producing the direct termination of the aversive event. But such behaviors can cause a significant management problem for staff who work with clients who should be directly supervised at all times.


A hypothetical male student with severe disabilities gets out of his seat during group instruction time and sits next to the computer. He is immediately brought back by staff. Further, to prevent such an episode from ­recurring, the teacher places an aide right next to this student so that the aide can stop the escape from the group activity early in its genesis. Over a brief period of time, this student learns the futility of just getting up and ­walking out. Hence, such DE behavior becomes less likely.

However, the aversive condition has not been diminished. He is still faced with the task. His motivation to engage in some behavior that results in escape from this task is still relevant. He then hits another student during this group instruction time. The teacher immediately directs the aide to guide him to the time-out area to teach him “We do not hit in this class!” With such a contingency, the rate of this student’s hitting peers during group time increases to a daily occurrence. Is this teaching him a lesson? Yes, it is! When you want out, hit someone.


What reinforcement operation explains why the teacher in this hypothetical scenario may continue to use time-out despite its function? Hint: Thank about what happens when this student is put in time-out and about the length of time he spends in that area away from other students.

The subcategories under DE functions (in addition to the other category) are DE 3.1: Unpleasant Social Situations; DE 3.2: Lengthy Tasks/Chores/Assignments; DE 3.3: Difficult Tasks/Chores/Assignments; and DE 3.4: Aversive Physical Stimuli/Event.

DE 3.1: Unpleasant Social Situations. This diagnostic category encompasses problem behaviors that directly terminate the client or child’s engagement or interaction in unpleasant social situations (an aversive EO for that particular individual). Here is an illustration of the DE function. A client is at a work station with a number of other clients. One of the other clients engages in screaming and other tantrum-type behavior. This client runs to the corner of the room, and then eventually goes out the door, leaving the area. The motivating condition for this set of behaviors was an aversive social situation, that is, the other client screaming. This client terminates the aversive condition directly by leaving the area. Perhaps such a behavior had been learned over many years of similar circumstances where screaming clients often follow up such bouts with another disruptive behavior, that is, they start hitting other people.

DE behaviors that can occur under unpleasant social situations can include leaving the facility, home, or classroom without permission; running away; or moving to another area. Many DE behaviors often involve the client’s response to the administration of a discipline procedure. For example, a child does something wrong during outside recess and is taken to the fence area as a time-out procedure. While unsupervised, the child leaves. Leaving the area would not be considered a behavior that is maintained because of adult attention! Rather, its ability to directly terminate the time-out (aversive EO) is what makes it functional and more probable in the future. Perhaps the child’s presence in the area should be more closely supervised.

The following circumstances can constitute aversive social situations (Cipani & Cipani, 2017):

simple compliance with adult requests, criticism (with respect to behavior, work performance, or criticism of appearance by peers), adult disapproval statements, social interactions (with a certain peer or adult, having an argument, having a large number of people in a room, having just a few people in a room, lengthy conversations) and threatened or implemented intended punishment consequences for behavior (removal to time-out, removal of privileges, fines involving points, etc.).

Table 3.9 illustrates that the establishing operation for this subcategory of problem behaviors is the presence of an aversive or unpleasant situation (first column). One must take into account that aversive is a relative term for that individual client, at that point in time. What is aversive or unpleasant to one person may not be to another. Further, the same situation may be less aversive to an individual on Monday than it was for the same individual last Sunday.

Such an antecedent context sets up the motivational condition for behaviors that escape (or avoid) such aversive events. The behavior becomes functional in that circumstance when it is successful at directly escaping or avoiding that situation. Note that in categories where the escape of the aversive situation occurs directly as a function of the behavior, the social consequences that follow such behavior are tangential and inconsequential. Table 3.10 presents some questions that are relevant for discerning such a diagnosis of DE 3.1: Unpleasant Social Situations.




1.Is there a reliable, somewhat frequent, direct relation between the problem behavior and termination of the undesired social situation? What is the form of the escape behavior that directly terminates the aversive event?

2.Are the target behaviors more likely to produce escape from the situation than other more appropriate behaviors to terminate a social interaction?

3.Does the behavior occur when the aversive social situation is presented (or about to be presented)?

Running Away From the Facility. A hypothetical client living in a group home facility for female adults with mental disorders and disabilities wants to go to the mall on Saturday. In this facility, clients must earn outside community privileges on the weekend through a point system. If a client earns a certain level of points by Friday, by behaving appropriately, following staff directions, and completing designated chores, she can earn a trip to the mall with the rest of the clients on Saturday afternoon. Further, other weekend community activities are also available to the clients who earn the prerequisite number of points. However, if a client does not earn the required points for that week, that client has to stay back while the remaining clients go on their community outing. To make such a consequence even more unpleasant, the client staying in the facility has to do extra, nonpreferred chores during this time.

When this client does not earn enough points, she leaves the facility on Friday night or early Saturday morning. In the past 4 months, she has left the facility unauthorized five times. Four of these five occasions were on weekends where she would not have earned community privileges. What is the reason for this potentially dangerous behavior? If she does not earn the points needed, she resorts to leaving on her own and avoiding chores and staying back. The primary consequence of leaving is simply to be brought back to the facility when found. Can you see how such a behavior will continue to occur under these conditions (i.e., not earning enough points)? As she becomes more skilled at navigating life outside the facility on her own, other contingencies become operable. In addition to escaping an aversive condition, she also enjoys some activities that are available only outside the facility. Such powerful results make leaving the facility more likely each weekend, whether she earns enough points or not.

DE 3.2: Lengthy Tasks/Chores/Assignments. This diagnostic category encompasses behaviors that postpone or terminate the child’s or client’s engagement with an instructional task, chore, or demand. The Cipani BCS contains two DE categories relating to instructional tasks or chores creating an aversive EO. Determining that an escape function can exist for such tasks does not reveal the actual aversive element or feature. The question must be raised: Why does such a task create an aversive condition? To address such a query, this manual examines two possibilities: (a) relative duration or length of task and (b) relative difficulty of task. The former element is inherent in this category, while the latter category is addressed in the next function: DE 3.3.

In this category, the value of escape is established because the task or chore is lengthy in duration. Lengthy is a relative term. What may be lengthy to you may be of short duration to me. Such a context sets up the motivational condition for behaviors that escape (or avoid) such conditions. The behavior becomes functional in that circumstance when it is successful at DE or avoidance of that instructional condition. The social consequences of such behavior are inconsequential. In many cases, such social responses following the behavior are often misdiagnosed as the primary function of the behavior. If escape is produced directly, what the teacher does or says after this result is not the “driving force” behind the behavior. Table 3.11 presents questions to pose when considering this diagnostic subcategory.


1.Is there a reliable, somewhat frequent, direct relation between the problem behavior and termination of the client’s involvement in the task or chore? What is the form of the escape behavior?

2.Are the target behaviors more likely to produce escape from the instructional tasks or chore than other more appropriate behaviors?

3.Does a “wacky contingency” exist (thus setting up an aversive EO)?

In this subcategory, the child (or client) is capable of performing the assignment or task, but the duration of the instructional activity required is above his current “comfort zone.” For example, a student can perform at adequate levels addition and subtraction problems involving carrying and borrowing. If given only five of these to do, no problem behaviors emerge. He or she finishes these and then proceeds to something else (hopefully more preferred). However, when given 50 to do, the student gets out of the seat, digs into his or her backpack looking for something, and performs other such avoidance behaviors. “Daydreaming” during seat work time is an example of a behavior that directly produces disengagement with the instructional material.

The professional attempts to analyze the antecedent conditions inherent in such target behaviors and also looks for contexts where assignment completion occurs. Examining how effective such possible DE behaviors are in the presence of lengthy tasks allows for a more accurate prescription to this problem. Manipulating the difficulty of the task is not the answer. Some form of altering task length is one component of a functional behavior analytic strategy.

Many teachers would report they reinforce appropriate behavior. However, the controlling effects of lengthy instructional periods may make termination of instruction a contingency that is more powerful than the use of praise and teacher attention as contingent events for appropriate behavior. Lengthy instructional periods, both in general and special education, would seem to be a prevalent context for problem behaviors, as the following example illustrates.


One might assume that appropriate behavior (e.g., staying on task, doing one’s work) is routinely reinforced in the classroom. When asked, most teachers would reply, “Of course I reinforce good behavior.” But what do they mean when they say, “reinforce good behavior”? Let us look at a typical series of events in the classroom and examine whether reinforcement is delivered for appropriate behavior, such as assignment completion.

At 9 a.m., the teacher hands out the math assignment to all the children in the classroom. The math period on the daily lesson schedule is slated for 9:00 to 10:00 a.m. At 10:00, the students are slated for a 10-minute recess period before the advent of another instructional period. Let us examine closely what happens to two different students, Jacque and Jeanette, in this hypothetical math class during a day in September. Both students are given similar assignments and have similar abilities. Both students also would rate math work as a less preferred task to other more interesting activities.

At 9:00 a.m., the teacher hands out the math assignment sheets, which consist of 25 calculation problems, and requests that the students begin working on their assignments. Jacque, disliking the assignment, says to himself, “I’m not sure I like this but I’ll work hard and get it done. Then I’ll get something else to do.” Jacque proceeds to work hard on the math assignment. Jeanette has similar thoughts, except she tells herself, “I’m not ready to do this yet so I think I will go sharpen my pencil and get my papers organized and get ready for the next period, which is more to my liking.” Subsequently, Jeanette does a minimal amount of work on the assignment, looking for every opportunity to do something other than the math problems.

Shortly after the period has started, Jacque has completed his work and goes up to the teacher and turns in his paper. The teacher remarks, “Jacque, what a wonderful job you’ve done. This is an exceptional paper. I’m so proud of you, and I know your parents would be proud of you also. Now let’s see how you do on this one (as she gives him another sheet).” (In the teacher’s mind, she is telling herself—“that’s the way to reinforce him for his work; I really gave him lots of praise. He’ll probably want to do a really great job on that new math sheet.”) Meanwhile, the teacher notes that Jeanette has not even finished the fourth problem on the assignment. She scolds Jeanette, and tells her that if she doesn’t finish her math assignment that she’ll have to do the same sheet over again tomorrow (the teacher notes to herself— “I’ll not praise Jeanette for her performance.”).

One may be tempted to laugh or discount this as an unusual situation, but it is all too frequently the case across many classrooms. Why is this a wacky contingency? The teacher certainly praised Jacque for completing his assignment! Can one say that reinforcement occurred for appropriate behavior? To answer that question, one has to examine multiple consequent events for completing the assignment. When Jacque finished the assignment, he received praise and more work! Note that the completion of the assignment results in more work. Jacque does not receive any additional reinforcer or incentive for completing his work, but rather is given more work (a nonpreferred task). Also, note that Jeanette’s failure to complete the assignment merely resulted in the work being postponed until the next day’s math period. Both children get the same amount of recess. Therefore, recess is not a contingent event that is based on assignment completion. Rather, it is time based, that is, when it is time for recess, apparently everyone goes, irrespective of the amount of work done.

Under these conditions, what can we predict for the future? Jacque may learn that finishing work usually results in more work. One can imagine what will happen to his work production over time. He eventually begins to complete his assignment less frequently. Jeanette learns to put up with scolding and disparaging remarks but still gets recess like everyone else. She is described by her teacher as a student who doesn’t care about reinforcement (because she obviously doesn’t do her work to receive the teacher’s praise).

What is the wacky contingency? Complete your work—get more! Fail to complete your work—get less! Fix that and you will be very successful in having many students become more adept at completing their assignments in school.

I Need a Break. A hypothetical female student with severe disabilities is given a task, which she completes. On days when she is given small amounts of work, followed by a break, the staff report she is not a problem. However, on days when there is more work, staff report that she more often lays her head down on the table.3 Staff ascribe such behavior to the possibility that she did not get enough sleep the prior night. She lies there for 5 to10 minutes and then gets up. Some days, she lays her head down a little longer. School staff want to speak to her mom about her sleep habits. Unfortunately, the information relayed does not seem to support their contention. Rather, she is more likely to engage in that behavior when the instructional task lasts longer than 20 minutes. Would you say she has learned how much work she wants to do in one stretch?

Making Many Beds. Some adult clients have problem behaviors when required to complete a chore. One should not assume that the difficulty is with the procedural aspects of the chore. The client can perform the requisite steps of the chore; hence difficulty is not the feature creating an aversive event. It is the length of the chore. Let us say a hypothetical client was given a position at a hotel as a maid. This was perceived as a “good fit” because the residential staff reported that she was able to make her bed every morning.

Several weeks after her start, she was let go for not performing her duties. Although she was able to make a few beds, and made them up to standard, the problem was in the lack of completing all her assigned rooms. Realize that if you work for a hotel in supported employment, you must make many beds. This duration mismatch created an aversive condition for that client. She would hide from the supervisor or fall asleep on one of the beds. Taking her own extended breaks directly escaped the lengthy task, and also got her fired. To say that she engaged in avoidance behaviors because she wanted to get fired would be fallacious. She would be perfectly happy having to do only a few beds a day and still be employed. The easiest explanation is not always the correct one.

DE 3.3: Difficult Tasks/Chores/Assignments. This diagnostic category encompasses problem behaviors that directly terminate the task or chore, with such escape responding generated by the difficulty of the task or chore. The aversive state is created because the task or chore is relatively difficult. Difficult is a relative term. What may be difficult for you may be easy for me (and vice versa). Such a context sets up the motivational condition for behaviors that escape (or avoid) such aversive events. The behavior becomes functional when it is successful at directly escaping or avoiding difficult tasks or instructional demands. Note that the social consequences that follow such behavior are inconsequential in terms of strengthening the behavior (column C). Table 3.12 presents questions to pose when considering this diagnostic subcategory.

Much passive off-task behavior is DE from task engagement. Such off-task behavior does not constitute a problem until its duration exceeds expectations of the classroom. When the duration of “taking a break from work” exceeds expectations, teacher admonishment occurs. Therefore, if longer breaks from work are desired by the student, some problem behavior that removes the student more permanently from the classroom is needed. To reiterate an earlier point, students who are faced with difficult material will often engage in disruptive escape behaviors that are socially mediated. The more disruptive, the more likely the student will get removed from the instructional situation by an adult.

Instructional Mismatch. In school settings, instructional tasks become aversive EOs when students are not capable of performing certain academic tasks accurately or fluently. Children who are given academic tasks that are way above their current level face this on a daily basis. For example, giving a child a seat work task involving decimal problems may be appropriate for his or her grade level. However, if he or she is capable only of adding and subtracting single-digit numbers, an aversive context is generated. This content level mismatch creates the conditions for escape or avoidance as a powerful motivating condition. This is a sure prescription for developing an aversion to math time, that is, constantly exposing the child to material with which he or she has very little chance of succeeding. In these cases, the child may engage in behaviors that avoid attending to the instruction or task, such as daydreaming, reading comic books, and so on. If these behaviors are frequently impeded by teaching staff, then other more disruptive behaviors evolve. With that evolution comes the referral for a behavior management program.


1.Is there a reliable, somewhat frequent, direct relation between the problem behavior and termination of the relatively difficult task? What is the form of the escape behavior?

2.Are the target behaviors more likely to produce escape than other more appropriate behaviors to terminate the client’s engagement in the difficult task? On what type of reinforcement schedule is it? Is the client thwarted in his/her attempts to directly terminate his/her engagement in the task? How long or often do such behaviors have to occur before they are successful?

3.Does the problem behavior occur in the presence of difficult tasks, chores or instructional assignments (i.e., aversive EO)?


Explain how instructional variables and problem behavior are inextricably intertwined. When students fail repeatedly with their class assignments, what condition does that set up?

DE 3.4: Aversive Physical Stimuli. If you mistakenly put on clothes that are too tight, you would remove these clothes and put on other clothes. Such a chain of behaviors is functional in that it produces the termination of the aversive event directly. Suppose you are prevented from changing clothes because personnel see such a behavior as your wanting to “control your environment”? Would you resort to behaviors deemed maladaptive? If you are faced with this daily, you probably would. Except that such behaviors are not necessarily maladaptive, but rather functional if the result is a change in clothes (see Table 3.13).

What are some other examples of DE behaviors terminating an aversive event? Upon hearing a loud noise coming from your smoke alarm, you may generate a chain of behaviors that eventually results in disabling the connection between the battery and the smoke alarm (assuming there is no fire). It is the last step that is performed that produces the direct cessation of the noise, hence negative reinforcement. People often leave an area that smells, put a coat on when it is cold, take a break when one has been working too long, and so forth. Although such behaviors are certainly understandable, with some clients or students, the manner in which they escape the aversive physical situation is unacceptable.

The differentiation between this subcategory and the previous categories is the imposition of a discrete physical stimulus that the client finds aversive (i.e., not a social situation). The following are some examples of problem behaviors in this subcategory.

Nude Clients. Some clients in residential and educational programs periodically or frequently engage in ripping or tearing off their clothes. Although the suggestion that such behavior is maintained by social attention is pertinent, that does not always turn out to be a correct diagnosis. What might be an alternate maintaining contingency for some clients? If the clothes the client is currently wearing are less desired, or if the client is used to some other fabric (e.g., cotton, silk), the motivational condition exists for escape behavior. Ripping off one’s clothes produces a direct effect on clothes removal, for example, fabric touching skin is no longer existent. If staff have difficulty finding other clean clothes, all the better (from client’s perspective).

Although tearing off clothes can result in adult attention (whatever is said or done), one must not always assume that such attention is the maintaining variable. Although nudity will certainly result in attention, the primary function is DE from the nonpreferred clothes. If the client was allowed to change into new clothes, he or she probably would not run around the unit nude. An analogue test for this hypothesis would be to allow the client to change into new clothes when signaling the desire (may have to be taught) for several weeks. This will remove the aversive context of wearing nonpreferred clothes for this period of time. Then for the next several weeks, institute a second condition: do not reinforce the request to change clothes. This will create the aversive EO again. If tearing off the clothes and nude appearance are low in the first condition while very frequent in the second condition, you have your answer.




Contingency (AO)

Presence of aversive stimuli

Behavior occurs (at some frequency and/or duration)

Aversive stimuli removed

AO, abolishing operation; EO, establishing operation.

Misdiagnosing this problem as an adult attention function (subcategory SMA 2.1) would lead to an ineffective treatment. Attending to this client when clothed and not attending when nude would do little or nothing to solve the problem. In fact, it may lead to longer durations of nudity because ignoring would elicit no response on the part of staff to get the client dressed!

In cases involving clients who rip off their clothes, why does an alternate, more acceptable behavior not occur, such as a request for different clothes, or changing clothes in a private area? Such behaviors may not be present in the repertoire of the client (see Chapter 4 on classification categories of replacement functions). It may also be the case that such behaviors are in the repertoire of the client but not reinforced by staff. Suppose a client is capable of taking clothes off appropriately in a private area, but staff block attempts to undress. Consequently, such an acceptable behavior is not functional. Similarly, requesting different clothes that are not as uncomfortable falls on deaf ears. Given these impediments to other behaviors, ripping off one’s clothes becomes an effective action to remove the aversive stimuli.

Head Bang From a Toothache? A young man with multiple disabilities was unable to speak and had a history of engaging in mild head-slapping behavior. The staff were instructed to simply interrupt him when he engaged in self-abuse and redirect him to his assigned task. They guessed that he was doing this to obtain attention because he kept doing it. Their next plan was for such behavior to be placed on extinction, that is, removal of staff attention.

Unfortunately, the intensity of the head slapping increased. If their hypothesis was correct, the removal of attention should have resulted in a decrease in this behavior. Concurrently, another behavior that would be more successful in getting attention would have increased. Such did not occur with this treatment regimen (wrong diagnosis!). The young man proceeded to bang his head to the point that his mouth was bleeding and the side of his face was swollen. Another hypothesis needed to be entertained. Upon being brought into the case, Keven Schock immediately sought medical attention. Upon examination of the mouth area, it was discovered that this young man had developed an abscessed tooth. It is very likely that this abscessed tooth was quite painful. The face slapping was hypothesized to have directly alleviated some of the pain caused by this abscess, at least temporarily. When dental treatment was rendered, two subsequent phenomena resulted. First, the treatment successfully reduced the abscess. Equally important, the self-abusive behavior ceased completely.

Category: SME 4.0

SME problem behaviors remove or postpone aversive events, similar to the DE functions. It contrasts with the DE functions in that the child or client’s behavior does not directly terminate the aversive event. Rather, the aversive event is removed through the behavior of another person, for example, parent, staff, care provider, teacher, or peer.

Within this category are the following subcategories—SME 4.1: Escape of Unpleasant Social Situations; SME 4.2: Escape of Relatively Long Tasks or Chores; SME 4.3 Difficult Tasks/Chores/Assignments; and SME 4.4: Aversive Physical Stimuli/Event. Note that the same events that served as aversive EOs for DE behaviors can also serve as the motivational conditions for SME. To reiterate, the difference is in the manner in which escape or avoidance is produced. A tantrum cannot directly remove a child from an unpleasant social situation (its only effect is on sound wave production in the area). Rather, it can serve as a communication to a parent to mediate the situation. Upon hearing the child throw a tantrum, the parent “interprets” such as a message that the child is unhappy in the current situation and thereby removes him or her. When the child appears more pleased at getting away from this situation, the parent “stores this information” away for the next tantrum.

SME 4.1: Escape of Unpleasant Social Situations. This diagnostic subcategory encompasses problem behaviors that terminate the child’s engagement or interaction in relatively unpleasant social situations (unpleasant relative to that child). The antecedent motivational conditions match the DE subcategory of the same circumstance, unpleasant social situations. These would not usually be instructional conditions unless the social circumstance itself (e.g., child sits next to someone who teases) makes the instructional setting aversive, not the instruction. The presence of the aversive social situation sets up the motivational condition for the client to engage in escape or avoidance behavior that is socially mediated. The form of such behaviors is determined by how successful and reliable such behaviors are at getting a person to terminate an already existing event or postpone the advent of an upcoming aversive event. Please refer to the DE 3.1: Unpleasant Social Situations section of this chapter for a delineation of circumstances that would be representative of this category. Table 3.14 presents the questions to ask when considering this diagnostic subcategory.

Hiding. A boy, about 5 years old, was referred to Keven Schock because he had been hiding in dangerous locations in his house (once he was found in the oven). What would be the circumstance that would generate such hiding behavior? Through interview and descriptive analysis, the antecedent for such behavior appeared to be his foster mother attempting to discipline him. When he had engaged in some unwanted behavior, his foster mother would call to him (from another room). She asked him to come to her so she could reprimand him.

As you can imagine, her tone of voice would be different than if she simply wanted to talk with him. Her tone of voice was the cue for this child. Upon hearing her command for him to come, he would find the nearest place to hide; this could be inside kitchen cabinets, under furniture, or in the garden shed. The process of searching for him was so tedious and aversive that it altered the mother’s response. By the time she found him, she was so relieved, she would sometimes hug him instead of reprimand him. His behavior of hiding was quite effective in escaping any further, more intense, reprimand from his mother.

Spinning. Keven received a referral of a middle-aged man in a locked psychiatric unit with a diagnosis of schizophrenia. He had an unusual way of getting the doctors and nurses to quit talking to him. When they approached and began to interact with him, he would simply stretch out his arms and begin to spin in circles. Anyone within arms’ distance would be struck. This behavior was quite successful in getting not only the doctors and nurses to avoid him, but also all of the other patients.


What FBA data would indicate that spinning behavior serves a DA 1.1: immediate sensory stimuli function? If a trigger analysis found that he spun around only 30% of the total times staff approached him, what would you set up as a test of a diagnosis of DA 1.1?


1.Is there a reliable, somewhat frequent, relationship between the problem behavior and termination of the undesired social situation? What is the form of the escape behavior?

2.Are the target behaviors more likely to produce escape from the situation by an adult, staff person, teacher, or parent than other more appropriate behaviors to terminate a social interaction?

3.Does the behavior occur when the aversive social situation (EO) is presented (or about to be presented)?

Shut Up and Leave Me Alone! A male client diagnosed with paranoid schizophrenia had been living in a locked psychiatric facility for 15 years (Schock, Clay, & Cipani, 1998). He had a lengthy history of hitting other people, primarily in the context of their initiating and maintaining an interaction with him. To say that his regard for small talk was not very high is an understatement! The collection of baseline data was not pursued due to the dangerousness of the behavior. However, an analysis of the possible function of this client’s aggression from records and interview seemed to indicate that hitting the other person resulted in his being left alone. The termination of the social situation as a result of hitting someone was hypothesized as the maintaining variable. Instead of saying, “Please leave me alone!” he would hit the other resident, which then produced the desired result of being left alone.

Developing an alternate behavior (instead of aggression) that would hopefully have the same function was required before he arrived at his new residential facility. His continued persistence at ending conversations by hitting people would probably make his stay at the community-based facility a short one. Prior to his arrival, Keven Schock and his staff decided to develop and reinforce a new behavior that would terminate a conversation and not involve hitting. The staff taught him to use the phrase, “Shut up and leave me alone!” They considered a more acceptable phrase, for example, “I do not want to talk to you anymore,” or “Please leave me alone.” However, those forms of protest may not have had the desired effect in the current facility. Residential staff persons might consider such a verbal request as a possible sign of needing help and continue trying to pry the client for information. Subsequently, he would be “forced” to hit that person to be left alone. Hence, more acceptable protests would be unable to reliably terminate the interaction, making such protests less functional than aggression. The selected target statement seemed like there would be no misinterpretation of its intent on the audience. Data over a 6-month period bore this out. The target verbal statement occurred regularly, while aggression did not.

Back to Level 1. Aggressive behavior can be an SME behavior, particularly when it occurs as a chain of events under conditions where a punisher is administered by staff for a client’s previous target behavior. For example, let us say a young boy at a residential facility does something wrong (i.e., a proscribed target behavior) and is moved down one level on the point system. As a result of being told he has been moved back to Level 1, he profusely swears at the staff person and begins kicking things, slamming doors, and throwing books against the wall. (Note to reader: Do not take this as our tacit approval of level systems of reinforcement. They are not our preference in behavioral intervention.)

Why does he engage in such a tirade? Such behaviors do not appear to be functional. The staff do not grant him his wish to go back to his previous level despite his verbal tirade. Perhaps what transpires after the staff intervene provides the explanation. Suppose someone in higher authority, hearing the commotion, meets with the child to calm him down. In the process of engaging in conflict resolution tactics, this authority figure makes a deal with the child regarding his return back to Level 2. What effect does such a negotiation have on the future probability of these reactions by this child to loss of a level on the system? Can such a negotiation “strategy” be the maintaining contingency for this child’s behavior when he is punished by staff and loses Level 2 privileges? Would this child have been offered a special deal had these raucous behaviors not happened? I think he will be more likely to act out in future situations where the only manner of ameliorating the consequence of his behavior is to knock things around.

This scenario may be quite familiar to people who work in facilities and group homes. It is sanctified as a clinical procedure and goes by the name of crisis management. What looks like a good way of handling an explosive situation in the short run (because the child stops intense tantrum behavior) is not beneficial in the long run because, as a result, these children have more frequent crises requiring the clinical expert to come in and render the situation manageable. This is done by negotiating the degree or amount of the consequences to be levied. Of course, such negotiation ameliorates the escalation of the child’s disruptive behavior at the time, but what is never studied or considered is the long-term effect of these frequent negotiations. How effective do you think the originally designed consequences will be in decreasing the initial target behaviors over several weeks or months? It would depend on how superb a negotiator the child is. The child may end up getting something instead of losing something!

Avoidance Functions. Many problem behaviors that occur under noninstructional conditions can be a response to the administration of a discipline procedure. In this manner, the occurrence of the target behavior may be strengthened more for its avoidance capability than for its immediate escape function. For example, a child does something wrong during recess and is taken to a time-out area. The child yells at the teacher, “I’m not going!” As the teacher escorts him or her to time-out, the child attempts to fight with and resist the teacher, hitting and kicking the teacher, resulting in an exacerbation of consequences. The child is then sent home for assaulting a teacher on the playground. Why do such behaviors occur while the child is being escorted to time-out? Aggression in this context gets reinforced if the adult is less likely to implement time-out for fear of stirring up a situation like the last time-out. When aggression, or the threat of it, conditions the social environment in that fashion, it becomes an effective avoidance behavior under threat of discipline.

SME 4.2: Lengthy Tasks/Chores/Assignments. This diagnostic category encompasses behaviors that postpone or terminate the child’s or client’s engagement with an instructional task, chore, or demand. The analysis in Table 3.15 depicts the relationship between the antecedent context, the behavior, and the contingency produced, which makes the escape behavior more probable under similar antecedent contexts. To reiterate from the prior discussion in this chapter, it is necessary to determine the reason for an instructional task being aversive. To address such a query, this manual examines two possibilities with respect to SME: (a) relative duration or length of task and (b) relative difficulty of task. The former element is inherent in this category, while the latter category is addressed in the next function: SME 4.3.

The value of escape is established because the task or chore is lengthy in duration (first column). Lengthy is a relative term. What may be lengthy to you may be of short duration to me. Although some of us may not mind cutting a small front lawn that takes only 15 minutes, we may not be comfortable with a lawn that is 4 acres (that’s why riding lawn mowers have a certain market share). In this subcategory, the child is capable of performing the assignment or task, but the duration of the instructional activity required is too long. Under this motivational condition, the problem behavior occurs. As a result, staff remove, ameliorate, or eliminate the task presented. Again, as with all socially mediated behaviors, the direct result of behavior is inconsequential in terms of stimulus control. The maintenance of the behavior is not determined by such a contingency being reliably produced.

Another point needs to be made in diagnosing this subcategory. Many professionals naive to behavior analysis believe they can determine the purpose of the child’s behavior by asking the child (or client) “Why do you do that?” Any professional report that uses the client’s statement to such a question as the sole basis for a diagnosis of behavioral function should not be taken seriously. Although the history of psychotherapy is one of accepting client self-report at face value, such a methodology cannot be relied on in a science-based approach!

Here is an example of the lack of utility of such questioning. When these children are asked why they don’t complete their work or assignment or engage in inappropriate behaviors, they often respond with “I am bored.” Personnel then believe that through this self-revelation, the causative agent of the problem behavior has been discovered. This contention is ill-conceived on several grounds.



First, behavior analysis requires that controlling variables be identified through reliable and valid measurement. The reliability and validity of the client’s self-report cannot be ascertained. Asking the client why he or she did something does not give us a “window” into the person’s motivation. One can certainly question the correspondence between what someone says and what someone does. The field of psychology, which relies on self-report as a valid measure of motivation, is often ineffectual in changing behavior, particularly when there is a weak correspondence between saying and doing. To understand motivation, one must study and analyze the observable, reliably measured events surrounding the behavior. Hence, an FBA will always require that someone’s motivation to behave in a certain fashion be understood from a scientific measurement of variables in the environment. Second, this explanation does not lead one to understand how off-task and disruptive behaviors are functional. In fact, if boredom was the causative agent behind the child’s performance, then the treatment should be providing more challenging work. I will take bets on how well that works in getting the child to perform better in class.

Examining how effective such possible escape behaviors are in the presence of lengthy tasks allows for a more accurate treatment prescription to this problem. In this subcategory, manipulating the difficulty of the task is not the answer. Some form of altering task length would be a primary component of a functional behavior analytic strategy. The “Case of the Wacky Contingency” presented earlier illustrates why such a context is existent in many school classrooms across the country.

This subcategory can also be diagnosed with tasks or chores outside of a classroom setting. Some clients have problem behaviors when required to complete a chore. It is not a difficult chore for this client; the client can perform the requisite steps of the chore. Rather, it is the length of the chore that establishes the value of escape. One may be able to wash several dishes in a sink at home. However, if you work for a restaurant, you must wash many dishes, pots, and pans. This duration mismatch thus creates an aversive condition for that client, which sets the stage for behaviors that temporarily or permanently escape from the task. It is to be distinguished from the next behavioral function in that the child’s ability to perform the task (accurately) is adequate, but the length of time is the factor making escape from instruction a powerful motivating operation. The questions in Table 3.16 should be posed when considering this diagnostic subcategory.


How would you conduct an analogue assessment to determine whether the problem behavior occurs during instruction because of length of assignment? Instructional difficulty?


1.Is there a reliable, somewhat frequent, relationship between the problem behavior and termination of the lengthy task(s), chore(s), or instruction? What is the form of the escape behavior?

2.Are the target behaviors more likely to produce escape from the situation by an adult staff person or parent than other more appropriate behaviors to terminate a social interaction? On what type of reinforcement schedule is it? How long or frequent do such escape behaviors sustain before the aversive task, chore, or instruction is terminated by someone?

3.Des the behavior occur when the task, chore, or assignment is relatively lengthy for the individual?

Hitting the Table and Social Mediation. A hypothetical student with severe disabilities will hit the table during instruction on a regular basis. Your observations may reveal that such behavior, particularly the lengthier and more intense it is, results in the teacher removing the student to another area. Although the short-term result is one of stopping the behavioral episode (which encourages the teacher’s belief of being on the right path), the long-term effect of such action is disastrous! The teacher is unwittingly developing such disruptive behavior as functional when the student does not want to continue in the instructional session. You have observed that hitting the table becomes highly likely when the instructional session goes beyond 6 minutes. Especially in light of the fact that there is no definitive criteria for instructional session length, hitting the table with some force inevitably occurs at some point in the teacher’s lesson. Although some people ascribe such behavior to the child’s intellectual disability, you now know better. You realize that the social environment has a big part in maintaining and probably exacerbating self-injury as an SME behavior.

What would be an analogue or in-vivo test to determine if length of task is the controlling variable in this case? In alternating fashion, the teacher would present one of two conditions during the instructional period in which this behavior occurred. In the first condition, on alternating days, the student would receive a short assignment. In the second condition, the student would receive a lengthy assignment (other days). An analogue test would mimic the controlling variables by removing the task, contingent on the exhibition of the problem behavior. The task would be terminated for a short period of time (e.g., 1 minute). If the results demonstrate that problem behaviors are prevalent in the second condition, especially when compared with the first condition, you have identified the aversive EO.

“I’m Bored.” A young girl is referred for property destruction, tantrums, and oppositional behaviors in the class. Your interview with the teacher and observations in the class reveal that when the student is given a lengthy assignment, she initially whines and complains, stating “I’m bored.” However, another more challenging task of similar length is met with this same complaint. When she is redirected back to work, she engages in more severe objections (tantrums and oppositional behavior) and eventually gets mad enough that she throws objects at the floor and wall. This results in the teacher removing the child to the corner until she can behave properly and do her work! Note the effect of the child’s throwing-objects 'margin-bottom:0cm;text-align:justify;text-indent: 18.0pt;line-height:normal'>Going Gandhi. A young female teenager, in and out of home placement could do any task or chore you wanted her to do. However, her actual performance left something to be desired. The group home staff constantly reported that she did not finish her chores. They ascribed it to her oppositional and lazy attitude, claiming she had a personality disorder. Her opposition would not be in a form involving aggression or destructive behavior. She did not yell or scream. Keven Schock observed the staff in one interaction with her. They were attempting to get her to get up, clean her room, take her medications, and come eat breakfast. She got up and went to a chair in the living room, and then she refused to move, refused food, refused medications, and refused to talk to anyone. She simply sat there, in protest, quietly like Gandhi.

An analogue test was set up to determine what could be the possible reason for not completing chores that were easy for her to perform under direction. In one condition, Keven assisted her in a simple task of room cleaning. If he gave her one-step tasks, she completed the task. All she seemed to need was one-step commands and contingent verbal praise after completing each step. However, when given a multiple-step direction, she did not complete the task. This was true even when offered praise and tangible reinforcement. During these tests it appeared that she was simply unable to complete the task. However, the same task could be accomplished if it was broken down into its unit components and performed one step at a time. Lengthy tasks were a “no go” at this time. Breaking them up into small units, and doing one at a time, solved the problem.

Enough Is Enough. A young girl with severe disabilities is given a task, which she completes. She gets out of her seat, but at that time, an instructional aide tells her to go back to her seat so that she can work on another task. She reluctantly complies, finishes part of it, and then attempts to leave her seat again. As she leaves her seat, the aide admonishes her to finish. She hits herself at this point, and the aide comes over to her and tells her to stop. She hits herself again, right in front of the aide. The aide then moves her to another area of the room with a different activity. If you were this child, what would you do the next time you felt you did enough work on a particular task or activity and wanted to change activities?

I Need a Break, My Brain Is Swelling. A mainstreamed fourth-grade boy in a regular class is given a reading assignment in a social studies text. The teacher has reported to you that this child will frequently throw a tantrum during class assignments. The teacher has suggested that maybe he does not belong in this class because it is too difficult for him. Your observation reveals that he reads about three pages during a reading assignment. He also is able to answer the questions correctly. He then decides to take a break and gets out some toy cars he brought from home. The teacher sees him playing, admonishes him, and gives him another reading assignment. He argues with the teacher, with no success, and then throws his social studies book on the floor. The teacher sends him to the principal’s office for defiant and disruptive behavior. The teacher decries, “That will teach him not to play in my classroom.” The child goes to the principal’s office and sits in the seat outside of the office with one of his cars in his pocket. You probably could fill in the rest of the story!

The 20-Minute Child. A child hits the desk to temporarily terminate the presentation of an instructional task. This scenario illustrates that the hitting-the-desk incident terminates instruction by the teacher, removing the instructional demand and subsequently “dealing” with the problem behavior. Note that the aggressive behavior is negatively reinforced by the teacher in that another stimulus condition is presented in the place of the task, for some period of time. Yelling and screaming can also accompany aggressive behavior intended to remove aversive events, often occurring before or during the aggressive acts.

Aggressive behavior results in the teacher removing the child to the corner until the child can behave properly and do the work. Note the effect of the child’s throwing-objects 'margin-bottom:0cm;text-align:justify;text-indent: 18.0pt;line-height:normal'>SME of Independent Seat Work. Some children who are deemed to have minimal impulse control when using a more traditional diagnostic classification, are found to throw tantrums and kick objects, desks, and other children (who are close by) when it is time to do seat work. Or they verbally refuse to complete the assignment. Note that turning to the minimal impulse control explanation does not give the slightest hint about what to do behaviorally. Perhaps medication interventions are apparent with this explanation. However, a behavioral intervention for minimal impulse control does not exist!

A functional explanation is far more useful in offering a route for intervention. If these behaviors serve to throw the teacher off track by dealing with these behaviors instead of pressing on with instruction, such behaviors can be maintained by serving an SME function. Although some students may be fine when working under close supervision, they cannot function adequately when more independence is required. The treatment strategy would attempt to do two things: (a) not allow such tantrum behaviors to postpone or ameliorate the seat work task requirements and (b) select another more appropriate behavior that replaces the tantrum behavior to terminate seat work and get close supervision. Perhaps after completing some independent seat work, the child can get his or her chair moved closer to the teacher’s area (hey, a contingency, what a good idea).


There are two escape functions dealing with instructional contexts in the Cipani BCS. This assignment will require you to make the relevant distinction between these two categories. Please address the following points in order:

✵What is the EO for an escape function due to difficult material?

✵How is that different from an escape function due to the “wacky contingency?”

✵Why might it be important to discern the reason for escape behavior as either duration or difficulty with special needs students who exhibit challenging behaviors in mainstreamed or inclusive settings?

SME 4.3: Difficult Tasks, Chores, or Instruction. This diagnostic subcategory encompasses problem behaviors that terminate the task or chore, with such escape responding motivated by the difficulty of the task or chore. In the presence of difficult tasks, chores, or assignments, the target problem behavior occurs. It is maintained in its particular form by its ability to escape or avoid such relatively difficult tasks.

In school settings, the child’s engagement with the instructional task is aversive, primarily because the child is not capable of performing the task accurately or fluently. This lack of skill sets the stage for escape from or avoidance of those instructional situations and conditions. This function occurs frequently with children who are given academic tasks that are way above their current level. This is a sure prescription for developing an aversion to school work, that is, constantly expose the child to material with which the child has very little chance of succeeding. In these cases, the child may engage in the problem behaviors to avoid all school work because such work is often something at which he or she is unsuccessful.

The questions in Table 3.17 should be posed when considering this diagnostic subcategory.

Self-Injury Often Escapes Instruction. Very often, problem behaviors, such as self-abuse, can function to remove instructional tasks or materials because the teacher has to stop instruction to deal with the behavior. Imagine seeing a young girl hitting herself repeatedly when faced with a task demand. Contingent upon this flurry of hits, the teacher attempts to hold her hands, thereby preventing her from hitting her face. But one must note what stimulus change gets affected in addition to physical contact. Instruction stops! As a result of the ability of the child’s self-abuse to terminate an instructional demand, hitting her face becomes more probable in the future when she is presented with the same or similar tasks.4

What would be an analogue or in-vivo test to determine if difficulty of task is the controlling variable? In alternating fashion, the teacher would present one of two conditions. In half of the sessions, the student would receive easy material. Such easy tasks would be determined by previous student performance demonstrating mastery levels of performance. For example, if the student is currently reading at the third-grade level as identified by achievement tests, an easy assignment would be first- or possibly second-grade reading texts.

In the other half of the sessions, the student would receive a difficult task or assignment. This could probably be accomplished by assigning reading material from the everyday work she does (third-grade level). An analogue test would mimic the controlling variables. Therefore, in both conditions contingent on the exhibition of the problem behavior, the task would be terminated for a short period of time (e.g., 1 minute). An in-vivo test would involve the same two conditions, except the break would be contingent on completing the assignment in whatever condition at the time, that is, difficult versus easy. If difficulty is the factor, the data should show a differential result between the two conditions.


1.Is there a reliable, somewhat frequent, relationship between the problem behavior and termination of the difficult task, chore, or instruction? What is the form of the escape behavior?

2.Are the target behaviors more likely to produce escape from the situation by an adult staff person or parent than other more appropriate behaviors to terminate the task? On what type of reinforcement schedule are they? How long do such escape behaviors sustain before the aversive or relatively lengthy task, chore, or instruction is terminated by someone?

3.Do the problem behaviors occur during tasks, chores, or assignments that are relatively difficult for the individual (an instructional mismatch)?

With some children who engage in self-injury, the form of their response and intensity defies one’s sensibilities. Such episodes of protracted hitting to the face and body of severe intensity perplex both lay people and professionals. In watching an episode of protracted length and intensity, it is not uncommon for even professionals to claim that the child is biologically driven to such episodes, that is, the child’s brain is responsible, and the child is not responsible for the actions. The overriding contention is that such behavior cannot be explained by any learning conditioning history.

Can a conditioning history be the explanation for such severe forms of self-injury? Unfortunately, minor forms of self-abusive behavior are often exacerbated unintentionally over time. Over the course of several or many months, the severity of the child’s self-injury (e.g., head hitting) may intensify as a result of the parents or teachers trying to ignore the minor forms of self-abuse. These severe forms of self-injury become even more powerful in their ability to alter the environment. In addition to producing an escape from the aversive context, the possibility of such behaviors also impacts the likelihood of task demands being made (avoidance function as well). As repeated attempts to ignore some forms of self-injury are carried out by the teachers and/or parents, the requirement for the intense forms to occur and produce escape is more evident. At some point in this longitudinal process, the young girl’s repeated self-injury has left scars. Further, she has become too unmanageable for her current educational placement, and her placement is changed to a more restrictive setting. In many cases, over time, her social environment becomes more respectful of such a destructive capability that very few (if any) demands are placed on her. Also over time, her tolerance for such painful blows has been enhanced, as a result of this gradual incremental approach to powerful strikes to the face and body. If you simply observe the child after years of “practice” altering her social environment, you might conclude that she is driven by some innate mechanism to abuse herself. (There is a disorder, Lesch-Nyhan, that is hypothesized to automatically produce self-injury, but most self-injury cases have been shown to be operant behaviors.)

Difficult Tasks. Other students engage in a variety of problem behaviors that escape difficult tasks. A student may initially complain about difficult material and subsequently engage in a number of verbal complaining statements such as, “This is too hard, I can’t do this. I don’t know how to do this.” If the teacher does not provide relief by presenting an alternate activity or making that task less difficult, the complaints may increase in intensity, or new problem behaviors give rise, for example, the student gets out of his or her seat or engages in disruptive behavior with other students. Of course, such behaviors are immediately effective in temporarily halting the student’s engagement with the assignment. It is important to note that the student’s initial complaining of the difficulty of the task was not reinforced. Therefore, the student had to engage in other (more serious) behaviors in order to produce the desirable result: escape from the difficult assignment.


These examples and the nature of this diagnostic category bring up an interesting question related to the current vogue of placing many special needs students in the core curriculum as a preferred practice. What this function presents as a probable outcome is the following: Presenting instruction and assignments that are grade levels above a student’s current level of functioning is often a prescription for serious behavior problems. In light of that, does a decision to place a student on the general education curriculum affect how he or she will behave in class? You bet! Instruction and behavior are intertwined. When students are incompetent at the material, such is an existent condition that breeds behavior problems, irrespective of where the child is served. Hopefully this makes people reevaluate the “mainstreaming at all costs” philosophy. Looking good and doing good can be two distinct entities!

Difficult Chores. Difficulty can also be the factor if the task is not accomplished in a fluent manner. For example, I used to dread changing my car’s oil back in the 1970s (before there were plenty of shops and garages that do such for a reasonable price). The aversion of the task was not in the lack of accuracy on my part. I eventually got the new oil filter and oil into the engine. It was the fact that it took over an hour to do such, with errors that had to be corrected along the way. Such a job for someone who is accomplished at changing the car’s oil probably completes it within 10 to 15 minutes. Both accuracy and fluency are variables involved in judging the difficulty of a designated task or chore. Students who use their fingers or manipulatives can perform multiple digit addition/subtraction problems. But it takes them a far greater amount of time to do 20 problems than students who have memorized the facts. Such a tedious approach to this assignment makes it aversive; memorizing the facts would result in these assignments no longer producing a motivational condition of significant aversion (hopefully math educators are reading this!).

Let us say a client in a female residential treatment program engages in loud tantrum behavior every morning when she is asked to make the bed. Such behaviors do not occur for afternoon chores such as taking out the garbage, or even vacuuming, which she likes. However, making her bed is always a headache for her and staff, who have to prod her through it. One program staff person suggested that she might have problems with bed making because it was a chore done in the morning. Perhaps she is not a morning person! They then changed the time she could make her bed to the afternoon. This solution had no effect on the rate or severity of the tantrum behavior. In examining the client’s performance of this task, she often required corrective prompts on many of the aspects of making a bed, such as the corner tuck and straightening out the bed sheet. Bed making took her twice as long as it should have because she frequently had to correct the errors she made. Could this be why bed making is so aversive to her? It is not that bed making is a morning task, rather it is the difficulty of the bed-making chore. Once the client becomes more adept at making the bed (accuracy and fluency), the behavior problems will decrease. As difficulty decreases (due to effective teaching), so do problem behaviors.

SME 4.4: Aversive Physical Stimuli/Events. The questions in Table 3.18 should be posed when considering this diagnostic subcategory. This function involves a reliable relation between the problem behavior and the escape or avoidance of the aversive physical stimulus (Question 1). Other more acceptable behaviors are less likely to result in escape (Question 2). Although the target problem behaviors may have to occur at some frequency or duration (Question 3), they are more successful at terminating the undesired condition than other more acceptable behaviors (Question 2).

Here’s Something You Cannot Ignore! A previous example illustrated how some clients rip off their clothes to directly terminate an aversive stimulus. In this example, let us stipulate that staff are fairly efficient at chain interruption, thereby blocking direct attempts to rip off clothes and receive new clothes. Similarly, requesting different clothes falls on deaf ears. However, when this hypothetical client urinates in his or her pants, guess what happens? The client is cleaned up and gets new clothes. Because staff cannot allow the client to remain in soiled clothing for very long, what behavior is functional in getting new clothes? Soiling one’s clothes!


1.Is there a reliable, somewhat frequent, relation between the problem behavior and removal of the aversive stimulus? What is the form of the escape behavior? What is the relative aversive physical condition?

2.Are the target behaviors more likely to produce an SME from the physical stimulus than other more appropriate behaviors? On what type of reinforcement schedule are they? How long or frequently do such escape behaviors sustain before the aversive stimulus is terminated by someone?

3.Does the problem behavior occur in the presence (or impending presence) of the aversive physical stimulus?


How does a professional evaluate the different sources of data collection to come up with a diagnostic classification of the target problem behavior? There are no hard and fast rules. In Chapter 2, we presented and rated the accuracy of the different methods of collecting functional analysis data. Therefore, depending on which data source is available, the professional evaluates the data, looking for evidence of a certain function.

It seems to us that in some cases, diagnosing a problem behavior is an ­iterative process. You may initially surmise that screaming, in the case of a client who refuses to do a chore, is SME 4.2: Lengthy Tasks/Chores/­Assignments. However, after an analogue assessment, the data do not bear that out. There were insignificant differences between long- and short-duration chores. Further, evidence brings you to suspect that adult attention is the culprit. Using a descriptive focused analysis reveals that such screaming brings hugs and praise from the staff as well as an occasional tangible reinforcer if the client promises that to be good and do the chores. It also reveals that other behaviors are not successful in getting hugs. Hence, the differential diagnosis more pertinent is either (or both) SMA 2.1: Adult Attention or SMA 2.3: Tangible Reinforcer.

As you can see, it is a lot like playing detective. Unfortunately, there is not an algorithm for determining functions of problem behavior. Although analogue assessments, trigger analysis, and in-vivo hypothesis testing will reliably give you the best information, we believe it is up to the behavior-analytic clinician to put the evidence together and come up with a best fit hypothesis.


The diagnostic categories in this chapter illustrate how different behaviors can serve different purposes. However, just as different behaviors can serve the same function, so too can one behavior (or topographically similar) serve different functions. Under some antecedent conditions, a particular behavior can serve a given environmental function. However, under other antecedent conditions, the same behavior serves a different function.

Perhaps a chart can illustrate how the same behavior in form can serve different environmental functions (Table 3.19). Let us use a hypothetical female client in a day treatment program for adults with varying types of developmental disabilities. This client engages in aggressive behavior under several different contexts, with different motivational conditions present.

Aggression can exert such a powerful effect on the social environment that it can be the most efficient behavior in some people’s repertoire under a variety of EOs. It can be efficient at producing adult attention in some circumstances, whereas it can also be adaptive when the person wishes to access some tangible reinforcer under conditions of deprivation. Further, it occurs under varied aversive contexts and effectively functions to escape or avoid such.

In Table 3.19, note that one function of this client’s aggressive behavior is SMA 2.3. In the morning, she hits staff when it is time to put her clothes on. The aggressive behavior toward staff occurs after she and they have argued for some time about wearing the clothes selected for her by the morning staff. Hitting occurs when staff attempt to get her to dress, following some period of arguing, with the selected clothes. On some occasions, when the client hits, some staff will let her choose which clothes to wear if she promises not to hit them again. The differential diagnosis for this client’s aggressive behavior would be SMA 2.3: Tangible Reinforcer. When she cannot select her favorite clothes to wear in the morning, she is more likely to hit staff. Further, hitting is a behavior that intermittently results in getting the desired clothing. Aggression appears to serve a socially mediated function under this condition.



This client also has some frequency of being noncompliant with staff around chores. The contingency for noncompliance is a brief time-out. Aggression also occurs in this context. When placed in the time-out chair, she will sometimes leave the time-out chair to hit the residential staff. In other cases, staff get hit by her on the way to the time-out chair. Under the conditions of time-out, aggressive behavior is illustrative of subcategory SME 4.1: Escape From Unpleasant Social Events. The client’s aggression will be maintained when being placed in time-out given two factors: (a) such behavior temporarily postpones the advent of the time-out (as they struggle to get her to the chair) and (b) such behavior reduces the likelihood of such a social consequence, that is, time-out, being used in the future (avoidance function).

For this particular client, the intervention for the target problem behavior will depend on the context that is existent when aggression occurs. If the context is the absence of a desired item (i.e., SMA 2.3), the contingency for aggression will be different than if aggression occurred in response to the implementation of a time-out. Such functional treatment strategies will be delineated in Chapter 4.

Diving Onto the Floor. Roy was a 48-year-old man who had been residing in locked psychiatric treatment centers for several years prior to his admission to Keven Schock’s facility. He had also briefly lived in several different care homes, none for longer than 3 months. He had numerous problematic behaviors, including making delusional statements, elective mutism, and catatonia. One of the primary reasons for his discharge from the group homes, however, was diving on the floor and refusing to get up. On some occasions, he injured staff by diving into them.

Once he arrived at Keven’s facility, the staff and Keven observed that sometimes he would dive to the floor and pretend to be asleep. In these incidences, he would lie on the floor while ignoring staff requests, until staff left the area. He would then immediately get up and engage in his usual activities. Certainly such incidents seemed to be escape from some aversive event that staff were presenting. However, there were other times when he seemed to desire their attention. On these occasions, once the staff left the area he would get up, run full speed, and dive headfirst into the nearest staff person’s feet. In one scenario, it appeared that the diving on the floor functioned to escape demands, but in the other scenario it appeared to function to access staff attention. Except for the result of the latter being diving into a specific person, the topography of each type of scenario was that of diving on the floor.

Keven ruled out the possible physical result of the behavior (i.e., the physical sensation obtained when skimming across the floor), as the controlling contingency. Keven and the staff focused on an analysis of the socially mediated effects of the diving behavior. Hence the A-B-C data in Table 3.20 reports only socially mediated effects.



Keven and the staff repeated the A-B-C descriptive analysis of the diving behavior across several days. On some occasions Roy would simply lie on the floor for a short time, then get up and resume his activities. However, during some of the observations, after lying on the floor for a short period of time he would get up, run full speed, and dive into a staff person. Diving into a person was always preceded by diving to the floor, but not every event of diving to the floor was followed by diving into a person.

The descriptive observation data Keven and his staff collected appeared to support two different functions for diving to the floor. One function appeared to be escape from an uncomfortable event or unpleasant task. For example, requests by staff to have his blood pressure checked, requests to attend therapy in groups, and requests to talk with a social worker, doctor, or anyone he did not already know all routinely resulted in a dive to the floor. Under this set of conditions he was likely to get off the floor quickly if the staff left the area and did not repeat the prompt.

The other function appeared to be to access staff attention. The staff noted that the behavior of diving into a staff person revolved around the number of prompts given to get up. If staff repeatedly prompted him to get up and attempted to talk with him while he was on the floor, he would sometimes get up and return to his previous activity. However, if staff delivered a single prompt to get up and then left the area, he seemed more likely to get up and dive into people.

Given the need to quickly and very accurately determine the function of this behavior, they tested the staff observations by conducting an experimental analysis. They developed three test conditions. One condition was in effect each day. The conditions were presented sequentially until each had been in place four times. One trial consisted of exposure to all three conditions. In order to expedite the assessment process, they collected data on two different behaviors in each condition. The first behavior was diving to the floor. The second behavior was diving into people. This design allowed them to simultaneously test both behaviors across three conditions.

In the first condition, staff tested the effects of task demand. In this condition the value of staff attention was kept at low levels by providing staff attention to Roy about every 15 minutes independent of any behavior on his part. The value of escape from task demands was established by a second staff making requests for Roy to complete tasks that had been associated with diving to the floor during the A-B-C data collection. In this condition, if Roy dove to the floor the staff making demands would leave, and the staff providing attention would continue talking with him as if the behavior had not occurred. A high rate of diving to the floor, and not diving into people, would support the contention that the function of the behavior is escape from relatively unpleasant tasks. Conversely, a low rate would tend to rule out that it functioned to produce escape from relatively unpleasant task demands

In the second condition, Keven and the staff tested the effects of deprivation of staff attention. In this condition, the value of escape from task demands was kept at low levels by ensuring no requests or demands were presented to Roy during the session. The value of staff attention was established by staff avoiding any unnecessary interaction with Roy. In this condition, if Roy dove to the floor the staff would continue to ignore the behavior and still make no demands. If he attempted to dive into staff, the behavior would be blocked and Roy would be redirected to another area. A high rate of diving into people would provide support for an attention function for this topography.

In the third condition, they set up a control condition for both attention and unpleasant tasks by ensuring that the establishing operation for both was minimal or nonexistent. In this condition, the value of escape from task demands was kept at low levels by ensuring no requests or demands were presented to Roy during the session. The value of staff attention was kept at low levels by providing staff attention to Roy continuously, independent of any behavior on his part. If Roy dove to the floor, the staff would continue to provide noncontingent attention and still make no demands. If he attempted to dive into staff, his behavior would be blocked and redirected to another area, and the noncontingent delivery of attention would continue.

A low rate of both diving behaviors would indicate that the motivating variables were a deprivation of attention (diving into people) and unpleasant tasks (diving on the floor). If such were the case, Keven would have possibly revisited the hypothesis of a direct reinforcement function. Table 3.21 shows the rate of occurrence of diving to the floor and diving into people as a function of each test condition.

A review of the data for condition one indicated that although he did dive on the floor, he did not dive into people. Because they had reduced the value of staff attention by providing it noncontingently in this condition, they could be reasonably sure that diving on the floor did not result in an increase in staff attention. This data provided support for the hypothesis that diving onto the floor was not maintained by staff attention. Given that there was a high level of demands in this condition, we had reasonable evidence that escape was a maintaining function of diving to the floor. Additionally, when the demand was removed contingent on diving to the floor, Roy stood up and returned to his previous activity.

The data from condition two require a bit more explanation. In condition two, Keven and his staff reduced the value of escape from demands by ensuring that no task demands were made. They increased the value of attention by not interacting with him. In this condition, they were reasonably sure that the diving to the floor and diving into people were not maintained by escape. In each of the trials, diving into people appeared to be the second part of a chain of behavior that started with diving on the floor and then, after some short period of time, he would get up and dive into a staff person. In this condition, diving to the floor was never followed by simply getting up and returning to his usual activity. A return to his usual activity occurred only after some level of staff interaction. In the third condition, the zero rates of both of the targeted behaviors provide evidence that attention and escape accounted for all of the occurrences of both topographies of the diving behaviors and that diving was not maintained by direct reinforcement.



Keven and the staff made the following conclusions based on this test data.

1.When Roy was asked to do tasks (even when already receiving attention) he would dive to the floor and remain there until the staff presenting the demand left the area. He did not dive into other people. The result was the demand was removed. (Escape function)

2.When Roy was not asked to do a task and staff did not interact with him, he would dive to the floor, and when staff continued to ignore him he would eventually get up and dive into the nearest staff person. The result was an increase in attention. (Attention function)

3.When they made no demands and talked with him he did not dive to the floor and did not dive into people. The diving behaviors did not occur when attention was already available and no demands were presented. (Ruled out direct reinforcement as maintaining contingency—possible sensory effects produced by grinding body against floor.)

Probably of greatest practical significance, the data from the third condition generated an effective temporary intervention to immediately reduce both topographies of the diving behavior while Keven and the staff developed a plan to teach Roy more acceptable behaviors to escape demands and access staff attention.


1.Why would a behavior that has both SMA and SME functions be difficult for staff to treat?

2.How would you make it easier for the staff to discriminate which motivating condition (EO) was at sufficient strength at any given moment?


There are four major function-based diagnostic classifications of problem behavior in the Cipani BCS. DA behaviors are maintained at high levels because they directly access the positive reinforcer. SMA problem behaviors produce a positive reinforcer indirectly, through the behavior of another person. DE behaviors remove or avoid aversive events directly. SME problem behaviors remove or avoid aversive events or conditions indirectly, through the mediation of another person’s behavior.

Within each major category, there are subcategories reflecting the same functional relationships as the major category. The subcategories delineate specific reinforcers, such as access to adult attention under SMA functions or escape from difficult tasks under either DE or SME functions. The Cipani BCS can be of great utility to determine the specific functional treatment strategy for the problem behavior.


✵Generate problem behaviors found in school, home, or community settings that reflect both SME and SMA functions. Delineate what the EO condition is, and how the problem behavior is efficient in producing an abolishing condition, that is, relevant reinforce.

✵Generate problem behaviors found in school, home, or community settings that reflect both DE and DA functions. Delineate what the EO condition is, and how the problem behavior is efficient in producing an abolishing condition, that is, relevant reinforce.

✵A classic study by Solnick et al. (1977) examined the effectiveness of timeout. The researchers found that time-out did not decrease the target behavior. Explain why the 6-year-old girl with autism exacerbated her level of tantrum behavior, to produce the programmed contingency of time-out.

✵See the hypothetical example of a DA 1.2 function regarding food scavenging in this chapter (see Food Scavenging!). Explain why such behavior is a DA function. Can food scavenging also be maintained by a socially mediated function? Explain how such a behavior might serve an attention function. How is the EO different in the DA case versus the attention function?

✵Discuss the case of a hypothetical student, Dolly, and how some appropriate behaviors do not result in attention while an inappropriate behavior does.

✵The real-life case (see Elopement) of how eloping from the facility resulted in this client getting soda is quite unique. Explain how such a function, especially in light of the establishing operation, is an example of an SMA 2.3 function.

He Likes to Be Restrained provides the case of an individual in which all staff said that the function of his problem behavior was getting people to restrain him. What was the maintaining contingency? Explain why the restraint was a requisite for getting this reinforcer.

✵How does Running Away From the Facility meet a DE 3.1 function-escape from unpleasant social situations?

✵What is the “wacky contingency”?

✵Explain how an instructional mismatch creates a contextual condition under which escape behaviors become likely.

✵Explain the case of Shut Up and Leave Me Alone! Found on page 170

✵The hypothetical case of a student, “I’m Bored,” is presented. Why does this child engage in property disruption?


1.While there are studies that show various behavioral treatments work for children with certain disorders, such does not constitute proof that those treatments work only for that particular disorder, for example, response cost contingencies are effective only for clients with ADHD. In point of fact, the effectiveness of specific behavioral contingencies rests on the behavior-environment arrangement that is affected by the treatment as being of relevance to the current function of the target problem behavior.

2.This should not be conducted unless written consent is obtained from guardians as well as administrative personnel at the site. Additionally, and just as important, obtain medical clearance for the potential welfare of the client if she or he ingests a few cigarette butts in one session. Also, if you conduct this test, designate a limit in which the assessment will be halted (e.g., after the second ingestion).

3.If you have special needs students who have significant problem behaviors during standardized testing, consider that such a context involves no breaks between sections and encountering difficult material as possible EOs for such behavior.

4.For an explanation of why some children with ASD have difficulty with instructional tasks, the reader is enjoined to read Appendix B: Why Artie Can’t Learn!


Caudery, G. E., Iwata, B. A., & Pace, G. M. (1990). Effects and side effects of DRO as treatment for self-injurious behavior. Journal of Applied Behavior Analysis, 23, 497—506.

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. (2014). Co-morbidity in DSM childhood mental disorders: A functional perspective. Research in Social Work Practice, 24, 78—85.

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. (2007). Functional behavioral assessment, diagnosis, and treatment: A complete system for education and mental health settings. 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.

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.

Mason, S. A., & Iwata, B. A. (1990). Artifactual effects of sensory-integrative therapy on self-injurious behaviors. Journal of Applied Behavior Analysis, 23, 361—370.

Schock, K., Clay, C., & Cipani, E. (1998). Making sense of schizophrenic symptoms: Delusional statements may be functional in purpose. Journal of Behavior Therapy and Experimental Psychiatry, 29, 131—141.

Skinner, B. F. (1957). Verbal behavior. New York, NY: Appleton-Century-Crofts.

Solnick, J. V., Rincover, A., & Peterson, C. R. (1977). Some determinants of the reinforcing and punishing effects of time out. Journal of Applied Behavior Analysis, 10, 415—424.

Switzer, E. B., Deal, T. E., & Bailey, J. S. (1977). The reduction of stealing in second graders using a group contingency. Journal of Applied Behavior Analysis, 10, 267—272.

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

Vaughn, M. E., & Michael, J. (1982). Automatic reinforcement: An important but ignored concept. Behaviorism, 10, 217—227.