Behavior Analysis and Social Work

  • Bruce A. ThyerEmail author
Living reference work entry
Part of the Social Work book series (SOWO)


This chapter provides an overview of the social work applications of the practice approach known as behavior analysis to the field of mental health. Behavior analysis has been an empirically supported approach to mental health practice for over 50 years, and its general principles have been accepted within social work since the 1930s. The chapter distinguishes between the philosophical foundations of this approach, known as behaviorism, from the practice applications derived from learning theory, and its hallmark approach to evaluation of clinical practice known as single-system research designs. Behavior analysis makes use of the theoretical principles of respondent, operant, and observational learning and through these concepts has derived a very wide array of empirically supported methods of mental health assessment and treatment. The focus on behavior analysis is on a client’s behavior, but behavior includes not only overt, publicly observable actions but also one’s emotions and thoughts, since these too are functions of the human body. Behavior analysis has always been concerned with promoting functional behavior of clients, but also the enhancement of positive feelings and thoughts, with parallel efforts to reduce dysfunctional behavior, maladaptive thinking, and dysphoric affect. The widespread belief that behavior analysis is solely focused on overt behavior is a harmful myth. Some clinical social workers are complete, or radical, behavior analysts, accepting the philosophy of science called behaviorism, the conceptual framework of social learning theory, and the research approach of single-subject designs. Others borrow selectively from among these principles to improve their practice in a more eclectic manner.


Social work Behavior analysis Evidence-based practice Clinical practice Empirical Social learning theory 


Approaches to social work practice in the field of mental health derived from learning theory, and given the labels of behavior analysis or behavior therapy, have been present in our professional literature for a very long time. One quote illustrating this perspective can be found in Robinson (1930, pp. 83–84): “Two dominant schools of thought may be recognized as differentiating case work approach and treatment at the present time: behaviorist psychology and psychiatric interpretation. The former emphasizes habit training, conditioning and reconditioning in treatment…illustrations of a partial use of this psychology in treatment are abundant in any case work area.” Other early social work writers commenting on the value of behavioral methods for social work include Berman (1927) and Bruno (1936). A general definition of behavioral social work was provided by Thyer and Hudson (1987, p. 1): “…the informed use by professional social workers of assessments and interventions based on empirically derived learning theories. These theories include, but are not limited to, respondent learning, operant learning, and observational learning. Behavioral social workers may or may not subscribe to the philosophy of science known as behaviorism.”

Let’s be clear from the outset of what is meant by the word “behavior” when we talk about behavior analysis and therapy. In turning to The Social Work Dictionary (Barker 2014, p. 38), we read that behavior is defined as “Any action or response by an individual, including observable activity, measurable physiological changes, cognitive images, fantasies, and emotions.” In other words, the term behavior subsumes everything a person does, internal or external, volitional or not. A fist pump is as much behavior of the person as is the pumping of their heart. Squeezing the hand in a friendly handshake is behavior, as is the peristaltic squeezing of our intestines. Reading aloud is as much behavior as is whispering or reading to oneself without making a sound. The toothache pain one experiences is as real as one’s teeth chewing food, and love for one’s family is as real as a physical kiss. All are seen as behavior and can be the focus of behavioral intervention. Indeed, from the earliest years of the development of behavioral therapies, many have been focused on non-observable behaviors, dysfunctional feelings such as depression and anxiety, reducing nightmares or hallucinations, changing irrational thoughts, correcting delusions, and ameliorating obsessional thinking. This point is stressed because behavioral approaches are often (incorrectly) claimed to only focus on overt, public behavior and that this is a serious limitation of this approach. Such incorrect claims are simply ignorant or malicious.

The complete or radical perspective of behavior analysis includes four discrete elements: a philosophy of science (properly called behaviorism), an empirical theoretical foundation, a repertoire of research-supported methods of assessment and intervention, and an approach to assessment and evaluating clinical practice outcomes known as single-system research designs. Each of these elements will be reviewed.

Behaviorism as a Philosophy of Science

We shall start with behaviorism’s philosophy of science. Critical social work reminds the profession that often the assumptions of social work practice go unacknowledged or at least unchallenged. Behaviorists cannot be reasonably accused of these sins, as there is an extensive literature outlining, elaborating, and sometimes critiquing its philosophical assumptions. The journal Behavior and Philosophy has been published since 1972 and is explicitly devoted to the philosophical foundations of the study of behavior (see, articulating and defending the behavioral position. A number of books are also available that do the same thing (e.g., Hayes and Ghezzi 1997; Lattal and Chase 2003; O’Donnell 1985; Rachlin 1991; Thyer 1999). The core assumptions of behaviorism appear in Table 1 and generally reflect the perspective of mainstream contemporary scientific inquiry. It is of course recognized that these assumptions may not be themselves “provable” in a scientific sense. Like the axioms devised by Euclid, from which he constructed an extremely useful system of plane geometry, behaviorally oriented clinicians and scientists find the array of axiomatic positions listed in Table 1 very useful in their approach to social work practice and scholarly research.
Table 1

Some of the components of the philosophy of science called behaviorisma

Acceptance of

Realism – The point of view that the world has an independent or objective existence apart from the perception of the observer

Determinism – The assumption that all phenomena have physical causes that are potentially amenable to scientific investigation

Positivism – The belief that valid knowledge about the objective world can be arrived at through scientific research

Empiricism – A preference to rely on evidence gathered systematically through observation or experiment, and capable of being replicated (i.e., verified) by others, using satisfactory standards of evidence

Rationalism – The belief that reason and logic are useful tools for scientific inquiry and that,

ultimately, truthful explanations of human behavioral will be rational

Operationalism – The assertion that it is important to develop measures of phenomena (e.g., client problems, interventive procedures) that can be reliably replicated by others

Parsimony – A preference of the simpler available adequate explanations for behavioral phenomena

Pragmatism – The view that the meaning or truth of anything resides in its consequences in action

Scientific skepticism – The point of view that all claims should be considered of doubtful validity, until substantiated by credible scientific data

aFrom Royse et al. (2016, p. 30)

Respondent Learning Theory Foundations

Behavioral social work is strongly grounded in several interconnected theories of human behavioral development and change. The broad outlines of each of the distinct approaches will be familiar to most social workers. Most fundamentally, the principles comprising respondent learning theory attempt to partially explain the development, maintenance, and elimination of relatively simple reflexive types of behavior and of human feelings. Also known as Pavlovian conditioning, after its foremost investigator the Russian physiologist Ivan Pavlov, respondent learning is a person-in-environment approach that examines the simple forms of reflective reflexive, not reflective behavior all humans demonstrate from birth (and sometimes in utero) and can acquire afterward. We blink in response to a puff of air on our eye. Humans do not learn this behavior; it exists from birth. We flinch when we hear an unexpected loud noise. This too is unlearned, and similar responses have been observed with the human fetus. These natural phenomena have been given some labels, with the event that triggers a naturally occurring reflexive reaction called an unconditioned stimulus (UCS) and the unlearned reflex called a unconditioned response (UCR). Pavlov (and many others prior to and after him) observed that events which immediately precede a UCS can themselves come to evoke a URC. When this happens the formerly neutral prior event is called a conditioned stimulus, or CS, and the response evoked by the CS is now called a conditioned response, or CR.

If I stand behind you and murmur the word “bang,” you will likely not react very much. But if I murmur the word bang and a moment later I crash a pair of cymbals behind you, you will flinch and winch to the loud sound. If I do this a number of times, shortly, when I murmur the word “bang,” you will flinch and winch, even if I do not clash the cymbals behind you. When I was young, I had my first bowl of Chinese hot and sour soup. Initially, when the bowl was set before me, nothing in particular happened, but during later meals over the years, when the hot and sour arrives, I begin to salivate, just like Pavlov’s famous dogs. In these examples of soup and cymbals, can you label the UCS, UCR, CS, and CR?

The phenomena of respondent learning have been demonstrated not only among human beings all around the world but across all animal species examined. It seems to be a universal property of living animals, down as far as single-celled organisms. While human beings may possess certain unique characteristics that distinguish us from other animals (and that itself is a debatable topic), we are undeniably members of the animal kingdom and are subject to its fundamental principles of learning.

To bring this topic closer to the field of social work and mental health, there is strong evidence that respondent learning principles are implicated in the etiology, as least partially, of a wide variety of behavioral disorders. Specific and social phobias, agoraphobia, post-traumatic stress disorder, acute stress reactions, and racism (see Arhin and Thyer 2004) are some examples. Many specific phobias, for example, begin when someone has a frightening experience with a previously neutral stimulus (a dog, a scary movie depicting something horrible). Coelho and Purkis (2009) provide a review of the evidence relating the onset of phobias to conditioning experiences. Sometimes the CS is experienced directly (being bitten by a dog), or sometimes vicariously (seeing a horror movie involving a mad dog), via information (reading about the diseases dogs can spread to humans) or seeing someone else (e.g., a parent) act extremely frightened around a dog.

One may be able to eliminate a dysfunctional CR through the process called respondent extinction. The clinical social worker presents the CS at a low level which does not evoke a strong CR. This is repeated, and gradually the intensity of the CS is increased, slowly, at a level tolerable to the client. I recently used this approach to help someone with a severe fear of balloons. She lived several hundred miles away, so we used the audiovisual program called Skype to conduct a respondent extinction-based treatment called exposure therapy to help her.

At a prearranged time, she called me via Skype and both of us could see and hear each other clearly. She could not tolerate watching and hearing me blow up a balloon to its full size and then pop it (her worst fear), so I had her minimize her screen so my image was very small and turn off the volume. She could tolerate watching me partially blowing up balloons then, which I did and proceeded to pop them. Soon she stopped flinching to the soundless image of small balloons being popped, and I asked her to maximize the image so she had greater exposure to the sight of popping balloons. Then, I asked her to increase the volume slightly, so she could, barely, hear the popping sounds. At her own pace, she increased the volume. After an hour or so, she was flinching much less and intermittently yelling in fear and laughing while watching me pop balloons with the volume and image at their maximum. At that point she could try and blow balloons up herself, initially only part way, and then pop them and then blow them up bigger and bigger. After 2 h she was blowing them up all the way and popping them with a pin, albeit with flinching and muted screams. After this session, when she could pop balloons on her own, I had her continue daily balloon inflating and popping exercises. After several months her severe fear of balloons had all but disappeared (see Elmhurst and Thyer 2019, for details).

The details of respondent learning principles are quite involved, and it is not at all such a simple process of learning new behavior as is commonly viewed. One social work resource which reviews this approach to learning can be found in Thyer (2012).

Operant Learning

More complex or volitional behavior in people can often be explained through a different form of learning theory, called operant conditioning. This too is a person-in-environment perspective which focused on how the consequences which have followed a given behavior in the past influence its future occurrence. Behavior which has been followed by a rewarding consequence usually becomes strengthened or more probable in the future, a process called reinforcement. In operant learning there are two types of reinforcement, positive reinforcement, wherein something is presented or experienced by a the person, after they display a given behavior, and negative reinforcement, wherein something is removed after a particular behavior, and the behavior is also subsequently strengthened or becomes more probable. If you put coins in a vending machine and candy is delivered, you are more likely to put coins in that machine when you are hungry. If hard work on term paper is followed by a good grade, you are more likely to work hard on similar assignments in the future. The word positive in positive reinforcement refers to something being presented, and the word negative to something being removed, after a behavior.

If turning the volume down on the television reduces obnoxious noise, this is a form of relief, and you are more likely to use the volume control in the future when obnoxious noise is being played. Because the consequence consists something being removed and the behavior is strengthened, this process is called negative reinforcement. Negative reinforcement is a good thing and examples abound – putting on sunglasses in the bright sun, opening an umbrella in the cold rain, escaping the boring house when chores are completed, and scratching an itch, all examples of behavior maintained by negative reinforcement.

Punishment involves consequences that are either presented (positive punishment) or removed (negative punishment), which reduces the likelihood of performing a given behavior in the future, at least under similar circumstances. If you are jabbed with a fork by your mother when you rudely reach across the dining table, and you reach rudely across the table less often in the future (at least when your mom is there), your mom is using positive punishment, positive in that she presented you with something and punishment because behavior became less likely. We do not like experiencing positive punishment (usually). If your mom deducts some money from your usual allowance when you reach across the table, and you decrease your rudeness, your mom is using negative punishment, negative because something was removed and punishment because the behavior became less likely. Examples of both types of punishment are also all around us. If we get a fine for speeding and speed less in the future, the ticket and fine served as effective negative punishment (because your money was taken away). Fines as a form of social control in general are designed, intentionally or not, as negative punishment. See if you can identify what form of operant learning is present in each of the following four examples:
  • You tend to leave your dirty clothes lying around the house. Your partner begins a program of picking them up and throwing them away. After a while you leave your clothes lying around the house less frequently (even though you have not run out of clothes).

  • You are late returning a library book and are charged a fine. You become less tardy in the future in terms of returning library books.

  • You put some money in a slot machine and WIN! Hundreds of dollars come spilling out into your bucket. You play slot machines even more frequently in the future.

  • You have a headache. You take aspirin, and within an hour, your headache has disappeared. In the future you are more likely to take aspirin when you have a headache.

How did you do? Get them all right? Good job!

When reinforcement that has been maintaining a given behavior is halted, the behavior tends to temporarily strengthen and then to decrease and eventually cease. In the past, a child has found that when she yells for her mom, mom tends to approach her and give her attention. If one yell does not do the trick, then two, three, or more yells have usually induced mom to attend to the child, perhaps give her something tangible she desires. Such reinforcement processes, often extending out during months of learning history, can produce a child who has severe tantrums regularly. If mom is instructed by the social worker, as a part of a behavior management plan, to only attend to the child when she makes a request in a normal voice, and to steadfastly ignore loudly yelling demands for mom’s presence, the first few times this approach is followed, the “tantrums” will likely increase, before diminishing. This eventual decrease in problem behavior is called extinction, and the temporary increase is called an extinction burst.

Another operant learning process is called shaping and involves the differential reinforcement of approximations to a desired terminal behavior that may ultimately be quite complex. When I am teaching my students to write papers in APA style, initially their prose and citations are hopelessly incorrect. I provide some corrections to their draft and praise them for what they got correct, and they turn in a more accurate assignment. This is repeated, and after a few drafts, with corrective feedback and praise along the way as they improve, soon most are completing APA-compliant papers. When teaching my kids to set the table, I showed them a complete service setting and asked them to duplicate it. They got some things right and other things wrong (e.g., the fork should be on the left). I praised them for what they got right and the next time asked them to try again. They usually got a few more steps correct. Again, praise, repeat, praise, and soon they were setting the table correctly (and I could comfortably pass along this one onerous chore to them for years!). Professors use shaping with their doctoral students. The ultimate task for most Ph.D. students is to produce an independent research project called the dissertation, a long and complex affair. However, during the 4 or more years of doctoral study, the students engage in tasks that successively approximate steps toward a real dissertation. They learn APA style, a statistical software package, and how to search the literature and perform a systematic review, collect empirical data and assess for reliability and validity, come up with one or more research questions and hypotheses, assist on my projects or those of senior doctoral students, and eventually begin their own research exercise, shaping.

There are two additional ways to categorize operant learning and that is viewing some types as contingency-shaped, wherein there is actual contact with reinforcing, punishing, and shaping consequences, and of personal experience with the removal of reinforcement and the decrease in behavior (Do you recognize operant extinction?). However, with humans, there is a second, perhaps broader, category of operant learning called rule-governed behavior (RGB). In RGB, behavior is changed by a verbal or written description of contingencies of reinforcement and punishment, rather than by direct contact. Some of us wear sunscreen because we have gotten burned, and later, when we wore sunscreen, we did not get sunburned and had a much nicer day at the beach. Others of us heard the advice, “Wear sunscreen at the beach so you do not get burned.” We wore sunscreen and have avoided most sunburns. Following directions, verbally or manually, and reading and adhering to what signs say are efficient ways of conveying the presence of the contingencies present in a given situation, without having to directly experience them. “Do not touch the stove element. It is hot and you will get burned” is an example most of us have followed. Using the manual containing instructions on setting up your computer is another.

Take learning to drive a car. We could give someone a key, let them get in the car, and fumble around for days figuring out what to do. Learning would take very long, be terribly inefficient, and involve many mistakes. But having an instructor tell us what to do, each step of the way, is a much more rapid way of learning to drive than the trial and error method of self-shaping. Following a recipe for baking a cake is more likely to produce an edible cake than giving someone the ingredients and letting them figure it out on their own. RGB only works so long as it usually leads to reinforcement. If the directions are incorrect or omit a crucial step, people may abandon attempting to acquire the new behavior. Usually RGB involves the verbal/written specification of the behavior and the outcome. Think of assembly instructions for a piece of furniture, or install a new software on your computer. But sometimes RGB is a more generalizable method of learning. Learning to comply when mother says “Come here” is usually a good idea, and most children learn to obey such rules reasonably consistently, but only because compliance has been followed by reinforcement in the past. If in the past when mother said “Come here” and compliance was followed by punishment, then adherence to the rule will be followed weakly, if at all. To be most effective, rules describing contingencies should yield fairly immediate consequences. Having a deadline is also a way to strengthen the efficacy of rules. Telling my children “If you make an A on this week’s test I will take you to the movies” is more likely to strengthen studying behavior this week than telling them “If you make an A for the semester I will take you to the movies then.” Hayes (1989) is a good source to learn more about the distinctions between contingency-shaped and rule-governed behavior. Both are seen as forms of operant learning.

Like respondent learning, operant learning has been demonstrated in all animal species tested – other primates, mammals like dogs, cats, pigs, elephants, reptiles, fish, amphibian, insects, etc. Again this suggests this is a universal and powerful mechanism of learning which is crucially important in human development across the life-span and in the emergence of both functional and dysfunctional behavior. Wong (2012) is an excellent overview of operant theory, written from a social work perspective.

Observational Learning

A third form of learning which is foundational to behavioral social work is learning via imitation, also known as modeling. In short “…a good deal of learning occurs through vicarious rather than a personal experience. We observe the behavior of others, observe the consequences, and later we may imitate their behavior” (Mazur 1998, p. 289). Learning by imitation is not limited to humans but has been observed through the animal kingdom (other primates, dogs, cats, rodents, bird, octopi, and even fish!), including learning fairly complex repertoires. One-month-old infants have demonstrated the capacity to imitate adults, suggesting a strong evolutionary component to this form of learning. It is likely that the ability to learn from others (especially their mistakes and successes) enabled our distant ancestors to survive and reproduce more successfully than those who lacked such a capacity. Of course the same can be said for respondent and operant learning.

Overlapping with the principles of rule-governed behavior, one can readily see the value of being able to learn not from what others tell us to do but what they demonstrate doing. Being shown a new skill is often a more effective method of skill acquisition than being told what to do. Modeling is facilitated by observing people similar to oneself, by observing models being rewarded for their success, and models who initially display less than perfect skill but who gradually improve and then get rewarded, are more effective exemplars than demonstrating perfect performance the very first time. Many behavior therapies combine all three elements of learning theory principles. When I do real-life exposure therapy, I am using principles of respondent extinction. I almost always model what I am ask fearful clients to do before asking them to do something difficult (e.g., touch a snake), and I am liberal in my use of reinforcement contingent upon their undertaking more fearful behavior, with praise and the judicious use of humor being foremost tools in this regard. Very few behavioral methods are purely one form or another.

Practice Methods in Mental Health

Social workers making use of behavioral methods fall along a continuum with respect to adherence to the complete behavioral model. At one end are practitioners who respect the empirical research foundations enjoyed by many behavioral interventions, and they apply them with their clients, who they may have diagnosed using the criteria found in the Diagnostic and Statistical Manual of Mental Disorders (DSM-5, American Psychiatric Association 2013). They may not subscribe to the philosophy of science called behaviorism, and they may make no use of single-system research designs to evaluate their practice. They simply pragmatically adopt methods that work. Many practitioners fall into this camp. Others, such as the author of this chapter, fully embrace the perspective known as behavior analysis, make little to no use of the DSM, use learning theory to guide their assessment and interventive techniques, and attempt to use single-system designs to evaluate the outcomes of their practice.

The DSM is viewed by behavior analysts as a scientifically flawed document which is both empirically and conceptually of little value in terms of being able to effectively help clients. As reviewed in a number of critical articles (e.g., Thyer 2006, 2014, 2015), the very definition of mental disorder found in the DSM-5 is tautological. This definition states “A mental disorder is a syndrome characterized by clinically significant disturbance in an individual’s cognition, emotional regulation, or behavior that reflects a dysfunction in the psychological, biological or developmental processes underlying mental functions” (APA 2013, p. 20). Claiming that a mental disorder is a disturbance in the processes underlying mental functioning is simply repeating words, like saying you can recognize sleeping pills because they put you to sleep. Imagine cardiology claiming that we can recognize heart disease because heart functioning is disturbed. Maybe the cause is not heart disease but diet, endocrine problems, medication side effective, or overuse of caffeine. Yet virtually any condition that is distressful can be labeled a mental disorder, according to the DSM-5. Behavior analysts suggest that when conditions adversely effecting thoughts, emotions, or behavior can be etiologically attributed to factors such as genetics, physical diseases such as Alzheimer’s, the effects of drugs or other ingested substances, or one’s environmental history, it is scientifically inappropriate to label the resultant changes in behavior as a mental illness.

The disciplinary overreach contained within the DSM-5 can be correctly viewed as a self-serving effort on the part of psychiatry to reclaim and even expand upon their professional purview. If all disturbances in cognitive, emotional, and behavioral functioning are said to be mental illnesses, irrespective of the etiology of that disturbance, and we are further told that the “DSM is a medical classification of disorders…” (APA 2013, p. 10), then the supremacy of psychiatrists as being the diagnosticians and treatment providers of choice (by virtue of their medical background) is unblinkingly asserted. This means that nonmedical professionals such as clinical social workers, psychologists, behavior analysts, etc. are seen as sub-optimally trained practitioners. Situational contexts, upbringing, parents, peers, impoverishment, and lack of opportunities are largely ignored in the DSM, even though these factors are frequently the etiological precursors to disturbances in behavior (and recall that in behavior analysis, the term behavior also includes emotions and thoughts).

One response by clinical social workers is to hold our nose and use the DSM criteria and terminology in our practice, in effect turning a blind eye to the deficits of the DSM. This is perhaps the most common practice. Another approach is to not use the DSM at all. Some very distinguished practitioners have advocated this approach, people such as Thomas Szasz and Carl Rogers. The whole approach known as client-centered therapy largely eschews the DSM categorical system, as do practitioners who use family systems therapy, who tend to view individual disturbance as a reflection of familial dysfunction. The approaches to clinical social work developed by de Shazer and Berg, solution-focused brief therapy, Michael White’s narrative therapy, and Saleebey’s strengths-based perspective all avoid psychiatric categorization, focusing instead on client’s strengths and possible solutions, not problems.

Behavior analysts too make no use of the DSM (except perhaps for insurance billing purposes). They are aware of the conceptual and reliability problems contained within the pages of the DSM-5 and instead focus on measuring behaviors and addressing them without stigmatizing labels. For example, a client who may meet the DSM-5’s criteria for PTSD may not be approached with an omnibus therapeutic approach (e.g., a medication, psychoanalysis, etc.) which is intended to treat the “disease” of PTSD. Instead, the behavioral analyst would work with the clients to assess the discrete behavioral issues that are disturbing them. These may include features such as nightmares, intrusive thoughts during the day, avoidance of situations resembling their trauma, an exaggerated startle response, hypervigilance, etc., all of which are amenable to empirical assessment. Each such difficulty could be individually addressed using different therapies derived from learning theory, and by reducing the severity of each aspect of the problem, which collectively gave rise to the presumptive diagnosis called PTSD. At some point the client is significantly improved, ideally demonstrated by a reduction in the presenting problems which lead the client to seek treatment. If the nightmares are now gone, avoidance is eliminated, intrusive thoughts absent, etc., the client is “cured,” all without resorting to using the label of PTSD. These various behavioral aspects troubling the client are not seen as symptoms of some underlying illness or disease; there is no disease present, not any other medical condition behind the problems. The nature of PTSD is environmental, not mental. The etiology resides in the client’s past environmental experience, the traumatic event. Now, of course the trauma may interact with some pre-existing aspect of the client. Someone with prior vulnerabilities may develop PTSD-like problems, whereas someone with a background that had presented less averse experiences may emerge relatively unscathed from the same “trauma.” The young soldier with a history of being severely battered as a child, or parental substance abuse and spousal fighting, may well be more likely to develop PTSD to a horrible combat-related situation, but the heightened vulnerability to develop PTSD is itself related to the soldier’s prior environmental experiences, not to their character, intrapsychic functioning, or attachment patterns. These three patterns, if they exist at all, are behaviors caused by one’s past environment; they do not cause or contribute to PTSD directly.

Assessment and Measurement

In keeping with good social work practice, after the conventional introductory session(s) spent establishing a relationship and a verbal or written treatment contract with the client, the behavioral practitioner devotes some time to selecting a method to empirically measure client functioning, with a focus on their presenting problem(s). As defined by The Social Work Dictionary, behavioral assessment is “…the attempt to establish the antecedent stimuli and consequences that shaped and/or maintain a given discrete behavior or more complex repertoire. The psychosocial origins of behavior are sought in terms of environment events, not mental ones. Behavior is defined in quantifiable dimensions such as how much, how often, and how long” (Barker 2014, p. 38). Unlike psychodynamic therapy, behavior is not seen as symbolic of underlying intrapsychic issues. Little Hans’ fear of horses would be taken at face value, and not interpreted as a fear that his father would castrate him (Freud 1909).

Assessment may involve the direct measurement of overt behavior; obtaining client self-reports of private events, either overt behaviors which are not amenable to direct observation by others (e.g., bedroom or bathroom-related behaviors) or of internal experiences such as feelings, thoughts, obsessions, hallucinations, etc.; or of some relevant physiological measurement (e.g., biofeedback-related assessments, blood sugar, blood pressure, drug testing, etc.). When clinically feasible such measures are repeatedly assessed so as to establish a baseline or data-based sense of client functioning. In a very real way, taking a baseline is congruent with the social work dictum of “Beginning where the client is at.” It is also fulfilling Mary Richmond’s advice from over 100 years ago when she told us “…special efforts should be made to ascertain whether abnormal manifestations are increasing or decreasing in number and intensity, as this often has a practical bearing on the management of the case” (Richmond 1917).

In some clinical situations, such as a crisis, the delays involved in obtaining a baseline may not be advisable, and this step is omitted. Ideally the baseline shows that the client’s problem is stable, or perhaps growing worse. Thus when intervention begins and continuing measurements are taken, the social worker is in a better position to determine if the client is improving, stable, or perhaps getting worse, on the selected outcome measures (there are be more than one). In behavioral social work, results are usually displayed in the form of a line graph, in addition to the traditional narrative case report. One can augment clinical judgment based on talking with the client by referencing the graphed data and thus be able to more compellingly say if the client is improving or not. There is a substantial literature on the use of this methodology, usually called single-system research designs, in the evaluation of social work practice (e.g., Thyer and Myers 2007). In some situations, wherein the intervention is removed and reinstated, the level of causal inference can be substantially enhanced, to the point one can say not just that the client got better, but they improved because of your social work services. Here the adoption of single-system designs segues into what the behavior analysts call the experimental analysis of behavior, and in the contemporary practice of behavior analysis, this is almost expected of each case. But this level of methodological rigor has not penetrated the field of social work to any great extent. Still, the use of regular, systematic, reliable, and valid empirical measurement of client functions, ideally before, during, and after social work intervention, is a hallmark method which has long distinguished this behavioral mode of practice of others.

Specific Disorders


Anxiety is usually defined as the client self-report of fear and active avoidance or escape from, feared situations, behaviors which are disproportionate to the actual danger of the situation. There are many ways in which such clinically anxious behavior has been categorized, and the use of this common nomenclature may be used for convenience sake, so long as one keep in mind there are not actual diseases in a medical sense, and the boundaries between the supposed discrete conditions offer considerable blurriness.

One behavioral therapy which has been around for many years, and enjoys a considerable amount of supportive clinical research, is called real-life exposure therapy and is based on respondent and operant learning theory. In this approach the social worker attempts to determine the nature of the anxiety-evoking stimulus (AES) for the client and to arrange to recreate it in carefully titrated amounts or dosages of exposure. For example, someone with a fear of heights may be introduced to an open stairwell, wherein during treatment they may be induced to gradually ascend at their own pace. The client with a severe fear of dogs enters a room containing a small puppy and only encounters larger dogs in later sessions as fears to the small dogs have been removed. The client is introduced to the AES with their full and informed consent, knowing they can have the situation terminated at any time. The social worker models coming into contact with the AES and asks the client to reflect on what they are feeling and thinking as they watch. With time the client usually calms down, and the therapist asks them to come closer, to briefly touch, and then to hold the AES. Homologous steps are arranged for different types of AES – heights, enclosed spaces, public speaking, insects, etc. With prolonged contact, the client almost inevitably calms down, permitting upping the intensity of the exposure exercise. With time, the clinician usually sees a reduction in avoidance, with subjective fear and physiological arousal remaining elevated. With more time, variables such as heart rate and perspiration subside, leaving the self-report of anxiety to be the last variable comprising “anxiety” to extinguish. Sessions usually last at least 1 h and should end on a “high note,” when the client is reporting feeling calm. Subsequent sessions pick up where the previous one left off or perhaps a bit behind the end of the previous session. Significant improvements are usually seen within the first few hours, and most clients obtain significant improvements so that all functional limitations are eliminated, as is excessive physiological arousal and the experience of subject anxiety. A search of YouTube using phrases such as “exposure therapy for phobias” will bring up many clips depicting this approach (e.g.,

Clearing houses and websites describing research-supported treatments for anxiety disorders indicate the high value of this approach (e.g., and in some cases provide clinical materials to assist with the application of the method, such as therapist treatment manuals, client manuals, assessment forms, and training opportunities. The applications of real-life exposure therapy are strongly supported for a wide array of conditions described in the DSM, such as specific and social phobias, post-traumatic stress disorder, panic disorder, agoraphobia, obsessions, and compulsions. Adaptions of this approach are used to effectively help reduce the craving of substance abusers and people with anorexia, bulimia, morbid grief, and pathological jealousy.


The varied conditions that have been labeled schizophrenia have been known for millennia. While it is possible that the condition will ultimately be shown to have a primarily biological etiology, at present insufficient evidence exists to support such a claim. In particular the hypothesis that this condition is related to underlying disturbances in neurotransmitter function has been known to be false for decades, even though this idea remains widely prevalent, reinforced by the pharmaceutical industry which profits from promoting this point of view (Whitaker 2002). While it is ideal that social workers have a clear and valid etiological formulation crafted before undertaking the treatment of clients, it is not always possible and for many conditions such as schizophrenia and other chronic mental illnesses is unlikely to be obtained. Nevertheless we must proceed with offering clinical services, and fortunately the absence of a clear etiological understanding of a condition does not prevent us from being to effectively care for someone. Just as lime juice was known to prevent scurvy before the discovery of vitamins, and proper hygienic practices beneficial in surgery and wound care before the discovery of bacteria, there are many effective interventions that can be used to help people with chronic mental illnesses, absent a scientifically compelling etiological theory, or causal mechanism for the treatment’s effectiveness.

One such approach to provide effective care for persons with schizophrenia is called assertive community treatment (ACT), a multidisciplinary form of therapy which has been around for over three decades. Free treatment resources to learn about the research-based foundations of ACT can be found on the Evidence-Based Practices Resource Center ( with information on how to build such a program, how to train frontline staff, treatment manuals, and how to evaluate your local ACT program. ACT has many behavioral methods built within it, relating to enhancing job skills; remaining employed; taking medication reliably, anger control, social skills training; and performing the functional daily living activities needed for independent, community-based living. A 15-min video describing ACT can be found here – Social workers have a central role on the ACT team.

In inpatient settings reinforcement-based programs called token economies have long been used to promote adoptive behavior and reduce psychotic-like actions of clients (Paul and Lentz 1977; Hackenberg 2018). In these systems patients are regularly awarded tokens or points throughout the day for displays of socially adaptive behavior and other actions associated with normality. The earned points are exchanged daily (usually) for consumables and extra privileges such as outings, movies, etc. above and beyond those normally used in the hospital. The general approach called social skills training makes extensive use of reinforcement and shaping techniques to help persons with psychosis (see Turner et al. 2018, for a recent meta-analysis of this research) and has the potential to not simply promote functional skills but also to reduce the so-called negative symptoms of schizophrenia (Grantholm and Harvey 2018). In this method, discrete aspects of positive social interaction are broken down into small behaviors, actions like gaining and maintaining appropriate eye contact, initiating appropriate greetings, engaging in conversation, saying farewell, enquiring about the well-being of others, etc. Each is taught, sometimes alone, using praise and shaping, the use of mirrors and video feedback, and repeatedly practiced under conditions that increasing resemble real-life settings. Social skills training is a remarkable effective psychosocial intervention which social workers are quite naturally able to provide.

There is also considerable literature on the use of behavioral methods to reduce or eliminate hallucinations, without the concomitant use of antipsychotic medications (Haddock et al. 1998). Very simple put, hallucinatory or delusional behavior is put on extinction, and normal conversation is selectively attended to and reinforced. Distraction methods (listening to music via earphones) can be useful and are teaching the client to simply ignore the experiences. Data can be taken surrounding the circumstances and times of day when hallucinations and delusions are more frequent and also time when they are less likely to occur. The social worker then works with the client to seek to expand the environmental situations where hallucinations and delusions are less likely (productively engaged in interesting activities) and minimize the times when they are more likely (e.g., alone with nothing to do). This latter approach is similar to the exception-seeking questioning undertaken in solution-focused brief treatment (SFBT), except behaviorally oriented therapists had been doing it years before SFBT was invented, and they augment simply asking questions about exceptional times by taking empirical data to see if their client’s reports are accurate. And then data continues to be taken to see if the intervention is having any effects. The use of methods such as these has even been shown to impact the display of so-called multiple personalities (e.g., see Kohlenberg 1973).


The experience of depression can take many forms and occur under different circumstances. There is postpartum depression; post-myocardial infarction depression; depression following the death of a loved one, the loss of a job, or loving relationship; and depression which is otherwise apparently uncaused (an assumption the behaviorist rejects, since all phenomena are seen as have potentially identifiable causes). In traditional psychodynamic theory, depression is seen as anger-turned-inward. Behaviorally oriented clinicians take a more parsimonious approach and investigate environmental aspects of the client’s life which may contribute to dysphoric mood. The model of Peter Lewinsohn is a particularly well-developed theory of depression in this regard.

“In the mid 1970s, Peter Lewinsohn argued that depression is caused by a combination of stressors in a person’s environment and a lack of personal skills. More specifically, the environmental stressors cause a person to receive a low rate of positive reinforcement….According to Lewinsohn, depressed people are precisely those people who do not know how to cope with the fact that they are no longer receiving positive reinforcements like they were before. For example, a child who has newly moved to a new home and has consequently lost touch with prior friends might not have the social skills necessary to easily make new friends and could become depressed. Similarly, a man who has been fired from his job and encounters difficulty finding a new job might become depressed. In addition, depressed people typically have a heightened state of self-awareness about their lack of coping skills that often leads them to self-criticize and withdraw from other people (e.g., depressed people may avoid social functions and get even less positive reinforcement than before). To make matters worse, some depressed people become positively reinforced for acting depressed when family members and social networks take pity on them and provide them with special support because they are ‘sick.’ For example, some spouses may take pity on their depressed partners and start to do their chores for them, while the depressed person lays in bed. If the depressed person was not thrilled to be doing those chores in the first place, remaining depressed so as to avoid having to do those chores might start to seem rewarding.” (c.f.

Lewinsohn has developed an array of measures which can be used in the assessment and response to treatment of depressed clients, such as the Unpleasant Events Scale, the measurement of the experience of pleasant events, the measurement of coping skills, and the like. Treatment involves attempting to minimize the client’s further exposure to unpleasant, depressogenic experiences and to schedule and promote the client’s exposure to pleasant events that they find rewarding. Lewinsohn developed a manualized “Coping with Depression” (CWD) course. “Treatment” can be undertaken more or less independently by the client (e.g., self-help) as well as in a group format or via individual therapy, and the program is now available in hard copy book form and also via computer, CD-ROM, and DVD. Clients are taught an array of skills to cope with depression – social skills, how to seek out and proactively schedule their engagement in rewarding activities, identifying negative thoughts, and promoting ways of thinking that elevate one’s mood. Baseline data is usually gathered for each client and the results used to plan an individualized program of treatment. The CWD program has been used with an array of diverse client populations and in many countries and has been tested in over 25 randomized controlled trials. It has also been shown to be an effective method to prevent depression (Cuijpers et al. 2009).


Social work practice guided by the principles of modern social learning theory, the philosophy of science known as behaviorism, the assessment and interventive methods of behavior analysis and therapy, and the evaluation of outcomes using single-system research designs is a comprehensive approach to mental health practice. Illustrations were given of the effective use of this model for clients with anxiety, persons with schizophrenia, and clinically depressed persons. This approach is widely used to help people with an array of so-called mental disorders, as well as to help clients with substance abuse problems, intellectual disabilities, and associated problems such as self-abusive behavior and aggression.

The professional infrastructure of behavior analysis and therapy is solidly established. The Association for Behavior Analysis International (ABAI) has over 7500 members worldwide and an affiliated chapter membership of more than 28,000. The ABAI presidency is now held by a social worker, Dr. Mark Mattaini, and the organization supports numerous journals and annual national and international conferences. The practice of behavior analysis is now licensed as an independent discipline in over 25 states, and insurance reimbursement for behavior analytic services is widespread. The primary independent practice qualification is the masters-level Board Certified Behavior Analyst (BCBA) credential, awarded by the Behavior Analyst Certification Board (, after passing the accredited courses, passing a rigorous national-level examination, acceptance of the BACB Code of Ethics, and obtaining suitable continuing education biannually. There is a doctoral qualification to the BCBA, a bachelor’s level qualification, and a lower-level Registered Behavior Technician qualification. There are over 29,000 BCBAs and over 300 academic programs providing approved courses and degrees leading to certification.

The related discipline of behavior therapy has its own professional association, the Association for Behavioral and Cognitive Therapies ( and is dominated by psychologists although social workers are welcome to join. The past president of ABCT is social worker Dr. Gail Steketee, one of our field’s most widely cited scholars, notably for her outcomes research in the behavioral treatment of persons with obsessive compulsive disorder, PTSD, and hoarding. Previously, social worker Dr. Richard B. Stuart served as the President of this organization (1974–1975). He was a pioneer in the fields of behavioral marital and family therapy, the treatment of obesity, and violent behavior. ABCT publishes a number of high-quality journals and sponsors annual conferences and continuing education programs.

ABCT is clearly more of a cognitive-behavior therapy organization, while ABAI largely eschews any notions of the role of cognitive factors as etiological agents of mental disorders. For the behavior analysts, cognitions, while real, are simply more behavior to be explained. Where do the negative thoughts of the depressed come from? Most likely from their upbringing and current environment experiences. Thus thoughts and beliefs, like overt actions, stem primarily from the environment, and the focus of treatment is on what is felt, not the feeling. Behavior analysis omits the middleman. A traumatic experience causes behavioral avoidance, negative feelings, and upsetting thoughts. All three aspects of behavior are the sequelae to the experience. The cognitive-behavior therapist is more likely to contend that the trauma occurs; it causes thoughts and feelings, which lead to changes in overt behavior. Thus the focus would be on changing the client’s ways of thinking.

Most social workers will perhaps not be prepared to make the jump to identifying more as a behavior analyst or as a behavior therapist by becoming involved with these membership organizations. But all social workers should be thoroughly familiar with social learning theory and of the principles and practices of behavior methods. To the extent we believe that it is important to follow the latest research demonstrating that some treatments are effective and some are less, so (and some are harmful) behavior analysis and therapy should be an array of methods we are prepared to provide our clients with so-called mental disorders. Indeed, it has been said that our clients have a right to effective treatment (Myers and Thyer 1997). If we really believe this, becoming familiar and competent with these methods is an ethical imperative (Thyer 1995).



  1. American Psychiatric Association (2013) Diagnostic and statistical manual of mental disorders, 5th edn. Author, Washington, DCCrossRefGoogle Scholar
  2. Arhin A, Thyer BA (2004) The causes of racial prejudice: a behavior analytic perspective. In: Chin JL (ed) The psychology of prejudice and discrimination: volume I, Racism in America. Praeger, Westport, pp 1–19Google Scholar
  3. Barker R (ed) (2014) The social work dictionary, 6th edn. National Association of Social Workers, Washington, DCGoogle Scholar
  4. Berman L (1927) The religion called behaviorism. Boni and Liveright, New YorkGoogle Scholar
  5. Bruno F (1936) The theory of social work. DC Health, New YorkGoogle Scholar
  6. Coelho CM, Purkis H (2009) The origins of specific phobias: influential theories and current perspectives. Rev Gen Psychol 13:335–348CrossRefGoogle Scholar
  7. Cuijpers P, Munoz RF, Clarke GN, Lewinsohn PM (2009) Psychoeducational treatment and prevention of depression: the “Coping with Depression” course thirty years later. Clin Psychol Rev 29:449–458CrossRefGoogle Scholar
  8. Elmhurst K, Thyer BA (2019) Self-conducted and Skype-mediated exposure therapy of a woman with a severe balloon phobia: a single-case study. Manuscript submitted for publicationGoogle Scholar
  9. Freud S (1909) The analysis of a phobia in a five year old boy. In: Strachey L (ed) Standard edition of the complete psychological works of Sigmund Freud. Volume 10. Hogarth Press, LondonGoogle Scholar
  10. Grantholm E, Harvey PD (2018) Social skills training for negative symptoms of schizophrenia. Schizophr Bull 44:472–474CrossRefGoogle Scholar
  11. Hackenberg TD (2018) Token reinforcement: translational research and application. J Appl Behav Anal 51:393–435CrossRefGoogle Scholar
  12. Haddock G, Tarrier N, Spaulding W, Yusupoff L, Kinney C, McCarthy E (1998) Individual cognitive–behavior therapy in the treatment of hallucinations and delusions: a review. Clin Psychol Rev 1:821–838CrossRefGoogle Scholar
  13. Hayes SC (ed) (1989) Rule-governed behavior: cognition, contingencies, and instructional control. Plenum Press, New York, pp 191–220CrossRefGoogle Scholar
  14. Hayes LJ, Ghezzi PM (1997) Investigations in behavioral epistemology. Context Press, RenoGoogle Scholar
  15. Kohlenberg RJ (1973) Behavioristic approach to multiple personality: a case study. Behav Ther 4:137–140CrossRefGoogle Scholar
  16. Lattal KA, Chase PN (eds) (2003) Behavior theory and philosophy. Plenum, New YorkGoogle Scholar
  17. Mazur JE (1998) Learning and behavior, 4th edn. Prentice-Hall, Upper Saddle RiverGoogle Scholar
  18. Myers LL, Thyer BA (1997) Should social work clients have the right to effective treatment? Soc Work 42:288–298CrossRefGoogle Scholar
  19. O’Donnell JM (1985) The origins of behaviorism. New York University Press, New YorkGoogle Scholar
  20. Paul GL, Lentz RJ (1977) Psychosocial treatment of chronic mental patients: milieu versus social-learning programs. Harvard University Press, Cambridge, MAGoogle Scholar
  21. Rachlin H (1991) Introduction to modern behaviorism, 3rd edn. W.H. Freeman, New YorkGoogle Scholar
  22. Richmond M (1917) Social diagnosis. Russell Sage Foundation, PhiladelphiaCrossRefGoogle Scholar
  23. Robinson V (1930) A changing psychology for social casework. University of North Carolina Press, DurhamCrossRefGoogle Scholar
  24. Royse D, Thyer BA, Padgett DK (2016) Program evaluation: an introduction to an evidence-based approach, 6th edn. Cengage, BostonGoogle Scholar
  25. Thyer BA (1995) Promoting an empiricist agenda within the human services: an ethical and humanistic imperative. J Behav Ther Exp Psychiatry 26:93–98CrossRefGoogle Scholar
  26. Thyer BA (ed) (1999) The philosophical legacy of behaviorism. Kluwer, DordrechtGoogle Scholar
  27. Thyer BA (2006) It is time to rename the DSM. Ethical Hum Psychol Psychiatry 8:61–67CrossRefGoogle Scholar
  28. Thyer BA (2012) Respondent learning theory. In: Thyer BA, Dulmus CN, Sowers KM (eds) Human behavior in the social environment: theories for social work practice. Wiley, New York, pp 47–81Google Scholar
  29. Thyer BA (2014) A review of Essentials of psychiatric diagnosis: responding to the challenge of DSM-5 by Allen Francis. Res Soc Work Pract 24:165–169CrossRefGoogle Scholar
  30. Thyer BA (2015) The DSM-5 definition of mental disorder: critique and alternatives. In: Probst B (ed) Critical thinking in clinical assessment and diagnosis. Springer International, Cham, pp 45–68Google Scholar
  31. Thyer BA, Hudson W (1987) Progress in behavioral social work: an introduction. J Soc Serv Res 10(2/3/4):1–6CrossRefGoogle Scholar
  32. Thyer BA, Myers LM (2007) A social worker’s guide to evaluating practice outcomes. Council on Social Work Education, AlexandriaGoogle Scholar
  33. Turner DT, McGlanaghy E, Cuijpers P, van der Gaag M, Karyotaki E, MacBeth A (2018) A meta-analysis of social skills training and related interventions for psychosis. Schizophr Bull 44:475–491CrossRefGoogle Scholar
  34. Whitaker R (2002) Mad in America: bad science, bad medicine, and the enduring mistreatment of the mentally ill. Basic Books, New YorkGoogle Scholar
  35. Wong SE (2012) Operant learning theory. In: Thyer BA, Dulmus CN, Sowers KM (eds) Human behavior in the social environment: theories for social work practice. Wiley, New York, pp 83–123Google Scholar

Copyright information

© Springer Nature Singapore Pte Ltd. 2019

Authors and Affiliations

  1. 1.College of Social WorkFlorida State UniversityTallahasseeUSA

Personalised recommendations