Reader Note: This is the second in a series of articles written by a practicing speech-language pathologist who was previously employed as an ABA behavior technician. The author prefers to remain anonymous due to the significant and sometimes career-ending stigma SLP professionals experience when advocating for an end to the use of ABA (applied behavioral analysis) with autistic and other disabled children.

Content warning for ABA-related trauma.

Part 2: The Problem with Pairing

The first time I realized that I didn’t play “correctly” (i.e. neurotypical play skills) was one of my first days as an ABA tech. My first client was playing with a car set at a table, and, as I sat beside him, I was silently mesmerized by the way he was manipulating each item in front of him. It reminded me so much of the magnetic balls I used to arrange in different patterns during free play in grade school. He examined each tree, road sign, and vehicle on the track and methodically placed them into a sequence. While he did this, I began to take the remaining items on the table and examine them in much the same way as I arranged my toys in a pattern alongside his. We didn’t speak much, but we were interacting. He made subtle adjustments to my pattern and giggled with joy as I started to make my pattern match his. It felt so natural to play this way.

I soon became aware that my supervisor was watching me and, as she caught my gaze, she motioned for me to come over to her. 

“Let me show you how to play in a more engaging way,” she suggested.

And I followed her back to the table. 

She grabbed some of the toys my client and I had arranged so carefully and began to drive the cars and trains on the track, heeding the different road signs, making vehicle noises, and commenting on what the cars were doing. I don’t recall how my client reacted because, for a moment, I had withdrawn into myself. 

“How did I mess this up?” I thought to myself as I watched my supervisor play with the cars. “How do I not know how to play?”

My supervisor handed the cars back to me, satisfied with her demonstration. As she watched from across the room, I began to rub my feet together hard and fast under the table (an activity that I have since learned is one of my self-regulating stims), and awkwardly pushed the cars along the track. I tried to bump my cars into my client’s to “crash,” but he just looked down at the table, perplexed. While I was busy making train sounds, he abruptly walked away and began to play with the bubbles instead. 

“Phew.” I thought to myself. “At least I know how to play with bubbles the right way.”

The rest of this first session continued like this. And my supervisor told me that that’s all I really needed to do for the first few sessions.

The beginning of an ABA program is all about something called “pairing,” and it’s something that continues throughout the program as well. She said that pairing was especially important because it would make it more likely that my client would comply when I placed demands while we were “working.”

At the time, I thought “pairing” was the same as “playing” but, as I’ve reflected on my experiences and become familiar with the research, I’ve realized that pairing is not play. It’s something else entirely. 

What is pairing? 

According to the ABA literature, “pairing” is a procedure used to establish rapport with clients “through the delivery of preferred tangible and/or edible stimuli, attention, and activities in the absence of demands (Lugo et. al., 2018). In addition to using pairing to develop a relationship with a client, ABA providers also implement pairing procedures because pairing, particularly pairing that takes place prior to session activities, “has shown promise in reducing the aversive nature” of many therapy activities (Lugo et. al., 2017). 

Pairing isn’t just free play, though. Pairing is a procedure that involves multiple steps, and its purpose is not enjoyment. The purpose of pairing is for the ABA provider to associate (i.e. pair) themselves with activities and objects that the child enjoys developing a relationship with them. The crucial difference between therapeutic rapport building and pairing is this: During pairing, the ABA provider uses their relationship with the child to later increase the child’s willingness to comply with demands that they find aversive. 

In this post, I will be exploring the literature surrounding ABA pairing procedures and highlighting the areas that are particularly problematic within a neurodiversity-affirming framework. 

Pairing in ABA Literature

Like many articles within ABA literature, the article I will highlight is a single case study. Because so much of the ABA evidence base comprises single case studies, it is not best practice to generalize the findings of these studies, particularly studies involving an exceptionally heterogeneous group such as autistic individuals, unless an extensive meta-analysis or systematic review of the literature has been conducted. To date, no meta-analyses or systematic reviews have been conducted to determine the effectiveness of pairing procedures alone or the long-term consequences of repeated pairing. At the time of this writing, the largest study investigating the effectiveness of pairing procedures that this author could identify included 3 individuals (McLaughlin & Carr, 2005). Despite this, pairing procedures are a routine part of ABA programming. 

You may access the study discussed in this post here.

In this study, researchers evaluated a 4-year-old autistic child’s response to three conditions: pairing before discrete trial training; free play before discrete trial training; and immediate initiation of discrete trial training with no pairing or free play prior to beginning session activities. (To see discrete trial training in action, see the first video in Neuroclastic’s article “Invisible Abuse”).

At the onset of the study, the child routinely demonstrated negative vocalizations and other negative “behaviors” when asked to participate in discrete trial training. Based on their results, the authors conclude that, even when the free play and pairing conditions were simultaneously available to the child, the child consistently preferred pairing with a preferred therapist prior to engaging in discrete trials. Additionally, the researchers report that the child no longer produced negative vocalizations in response to discrete trial training by the conclusion of the study.

This means that pairing works, right? 

Well, it depends on how you define success.

If success is defined as the child engaging in an activity that they previously found aversive, then, yes, pairing often works. But it leaves many vital questions unanswered, such as:

• Why did the child find the activity aversive to begin with? 
• Could the activity have been modified to accommodate the child and be an activity the child enjoyed? 
• Why, exactly, does pairing increase a child’s willingness to do something they find aversive? 
• Does the child truly find the activity less aversive, or do they engage in aversive activities to please the adult they have developed a relationship with? 

The researchers admit that a failure to conduct a functional behavior assessment as part of the study may have impacted their ability to attribute reduced negative vocalizations to the pairing procedure. Because they did not identify the “function” of the negative vocalizations, they suggest that it is possible that “the decrease [in negative vocalizations] was the result of prolonged exposure to [discrete trial training] sessions in which negative vocalizations, which may have served a socially mediated function (e.g., escape or attention) in the past were placed on extinction” (Lugo et. al., 2018). Ultimately, the authors are unable to provide a definitive conclusion for why the child demonstrated decreased distress during previously aversive activities after engaging in pairing. 

What are the steps for “successful” pairing? 

While there are a variety of “pairing skills” that providers and parents can implement (Barbera, 2007; Lugo et. al., 2017), a review of the literature regarding pairing reveals that neurotypical play skills are particularly valued within this procedure. 

(For additional insights on the philosophy of ABA and the ways in which it dehumanizes neurodivergent people, please see one of the above author’s blog posts here in which she applies her experience clicker training chickens to providing ABA to young children. 

“The better you get at training animals, the better you get at training and teaching people (especially young children with little to no language)”
– Mary Barbera BCBA-D and “Autism Mom”, 2018

Let’s examine the pairing procedure that Lugo et. al. (2018) used in their study in depth. 

The Problems with Pairing:

In this study, ABA providers followed these presession pairing steps:

1. Maintaining proximity to the child
2. Providing behavior specific praise contingent upon appropriate play
3. Repeating or commenting on participant vocalizations
4. Imitating appropriate play
5. Describing appropriate play
6. Offering toys
7. Modeling novel ways to engage with toys

Obviously, some of these activities are things I do every day in my sessions with my clients. For instance, I frequently offer my clients toys, as most therapists do, and occasionally play with toys in ways that are unfamiliar to my clients when we are engaged in parallel play. Sometimes I engage in neurotypical play schemes but, frequently, I also play in neurodivergent ways because that is the play that I, personally, most enjoy. No type of play is criticized. No type of play is praised. Play is not used as a “reward” for “good behavior.”

Play is just play – we play because it’s fun and we enjoy it.

In contrast, the steps that authors outline in this study, as well as those that she trains ABA providers to employ in an earlier study (Lugo et. al., 2017), clearly indicate that a central component of the pairing procedure is encouraging “appropriate” (i.e. neurotypical) play. 

Problem 1: Pairing procedures prioritize neurotypical play while devaluing neurodivergent play.

Through pairing procedures, many ABA providers establish a hierarchy in which neurotypical play is prioritized and valued above neurodivergent styles of play. Throughout the pairing literature, providers are instructed to praise, imitate, and describe “appropriate” play to encourage the development of neurotypical play.  Being ignored is devastating to people of all ages and neurotypes, which is why neurodiversity-affirming therapists, teachers, and parents refrain from this practice. 

Of course, autistic play is not the only thing that providers in this study ignored.

To encourage the child in their study to favor presession pairing over free play, Lugo et al. (2018), forbid therapists from interacting with the child during free play: “Any bids for attention (e.g. questions, physical contact), were ignored.” While the authors state that they did this to “control for the access to toys” it is difficult for the reader to imagine that this “planned ignoring” would not be distressing to the child, particularly when considering the fact that the authors intentionally chose therapists for the study who the child preferred interacting with. Providers were selected by asking the child to select pictures of preferred therapists to identify which therapists would administer discrete trial training. 

Ultimately, it makes sense that the child repeatedly chose the condition (i.e. pairing) that resulted in a familiar adult acknowledging them.

Researchers have sufficiently debunked the myth that autistic individuals have no interest in social connections (see a summary of studies regarding autistic friendships). Being ignored is devastating to people of all ages and neurotypes, which is why neurodiversity-affirming therapists, teachers, and parents refrain from this practice.

Problem 2: Pairing can set children up for future manipulation, exploitation, and abuse

Although the topic is woefully under-researched, the small body of literature that exists investigating abuse among disabled individuals indicates that disabled children and adults, particularly those who require higher levels of support, are significantly more likely to experience abuse. Some studies estimate that disabled individuals experience a lifetime prevalence rate of interpersonal violence ranging from approximately 25%-90% (Hughes et.al., 2019), while other studies suggest that 39%-68% of developmentally disabled female children and 16%-30% of developmentally disabled male children will experience sexual abuse (Mahoney & Poling, 2011). Additionally, these statistics don’t include the ways in which disabled people may experience less visible forms of maltreatment including financial exploitation, emotional abuse, and medical neglect. 

What place do pairing procedures have in a discussion about the horrifying rates of abuse and mistreatment that disabled children and adults experience? 

Disabled individuals experience abuse, particularly sexual abuse, most commonly at the hands of people with whom they have a relationship. Perpetrators are often family members or health care providers who their victims trust and rely upon (Mahoney & Poling, 2011). Abusers leverage their relationship with their victims to ensure compliance, and the grooming process abusers employ bears a striking resemblance to the pairing procedures included in most ABA-based programming. Like pairing, the grooming process starts by discovering the victims’ interests and using those interests to develop a relationship with them. Once trust is established, the abuser first asks their victim to do small things that they may feel uncomfortable with, but they comply with in order to please their abuser (Winters & Jeglic, 2017). At this point, you could say that the abuser has effectively “paired” with their victim. As the abuse continues and increased compliance is established, the abuser is then able to ask their victim to engage in more and more aversive activities, which the victim may comply with in an effort to please their abuser. 

Alarmingly, children who undergo ABA interventions are likely to experience pairing repeatedly with many adults over the course of their “treatment,” contributing to increased and unquestioned generalization of compliance and making them more likely to experience abuse and mistreatment in the future.

ABA has a well-documented high turnover rate, with nearly 40% of ABA technicians in 2018 reporting that they intend to quit their jobs. The average “turnover intent” of workers in the general population is only approximately 18% (Kazemi,Shapiro, & Kavner, 2015). Even more concerning is the fact that ABA technicians commonly employ programming for toileting, dressing, bathing, and other highly personal activities of daily life, so children often become accustomed to the expectation that they will not question or refuse physical contact from adults. If they are sufficiently “paired” with their ABA professional, they’ll comply with the demand, no matter how uncomfortable it makes them.

While this is not to suggest that ABA providers are sexually abusing individuals, the pairing procedures used in ABA can set individuals up for a variety of abuses due to its focus on eventual compliance. 

Therefore, grooming and pairing are, essentially, one and the same: Both processes are intended to develop a relationship that an adult then leverages to encourage a formerly unwilling participant to do something that they may not have originally felt comfortable consenting to. 

Problem 3: Pairing procedures are intended to make children do things they find aversive, often without identifying why something is aversive in the first place. 

While planned ignoring and devaluation of autistic play are highly problematic, they are not the most nefarious aspects of the pairing procedure.

Instead, the most striking and dangerous aspect of pairing is its intended result: to increase the child’s likelihood of doing something they originally found aversive. 

The central problem with pairing, in both the literature and in practice, is that it does not consider why children are avoiding and protesting particular activities. Instead, the purpose of pairing is to, through developing a relationship with the child, increase the likelihood that the child will do something that they fundamentally do not wish to do.

Through the pairing procedure, children begin to learn that doing things that feel good to them (e.g. autistic play, stimming, etc.) are less valuable than the things that make the therapist feel good (i.e. neurotypical play, compliance).

They learn to ignore their feelings of distress when confronted with a task they find aversive and mask their distress to please their caregivers. They become motivated by the praise they receive when they “do the thing” and become reliant on external reinforcement rather than intrinsic motivation, regardless of how much they are suffering.  

Proponents of ABA will argue that pairing encourages children to engage in vital activities that they would otherwise refuse to participate in. They argue that the purpose of their ABA, the “modern,” “good,” and “play-based” ABA, is not to make an autistic child indistinguishable from their neurotypical peers as Lovaas originally intended. Instead, they claim that the goal of modern ABA is to increase socially significant behaviors. On the surface, that sounds benign, even useful. However, when we consider what “socially significant” actually means, we begin to see how ABA interventions are inherently at odds with the neurodiversity paradigm. 

Socially Significant Behaviors are Neurotypical Behaviors 

Disabled people do not currently have any voice in determining what is “socially significant.” Social significance is determined by the neuro-majority, which is made up of nondisabled people.

Mounting research indicates that autistic individuals communicate successfully with autistic peers, experience sensory input differently, and, when they act “too autistic,” continue to face incredible discrimination from a neurotypical society.

Therefore, socially significant behaviors are behaviors that neurotypical people have determined are valuable, appropriate, and desirable. Inherently, socially significant behaviors are neurotypical behaviors. This is why many seemingly well-intentioned ABA goals are still fundamentally ablest: the goal is often masking. Perhaps the goal is not to mask as significantly as Lovaas originally imagined, but it is masking nonetheless. 

Every time we tell an autistic child to “use a respectful tone,” we are altering their inherent way of communicating. When we misidentify their stimming as a “barrier” to their learning, we are disregarding the ways in which they move their bodies to regulate and learn. When we write goals for them to play with toys “appropriately and as they are designed,” we are telling them that their way of playing is invalid. When we target reduced protesting and increased compliance, we diminish autonomy and self-determination. In all of this, we are telling them to mask. 

The pairing procedures used in ABA create a relationship that the adult then leverages to increase the likelihood that a child will do something they don’t want to do. Reduced protests and fewer “negative” vocalizations do not necessarily indicate an increased desire to perform an activity that one previously found aversive. Instead, it is far more likely that the activity continues to be aversive but the child has learned to mask their distress to please an adult they care about. 

Ultimately, pairing is just another procedure ABA providers use to encourage masking and compliance. It’s not play; it’s disrespectful, dehumanizing, and manipulative. 

Does this characterization of something so seemingly innocent seem extreme? It’s not. Here is one of the many “how-to” guides to pairing for parents provided by an ABA agency that is now owned by the large company Proud Moments ABA.

ABA Pairing “Rules” for Parents

Respectful and affirming therapy isn’t about manipulating children to do what we want them to do by offering them toys, candy, or praise. If a child expresses distress when asked to participate in an activity, our reaction should not be to leverage our relationship with that child to get them to do what we want them to do. Instead, we should be asking ourselves why this child is distressed, and that is often something that the “four functions” of behavior (i.e. sensory stimulation, escape, attention, tangible) utilized by ABA is not equipped to identify because the fundamental reasons underlying many behaviors are not measurable or directly observable.  

The neurodivergent experience is one that is marked by inherent differences in sensory processing, communication, and socialization. As a result, the experience of navigating a neurotypical world is often marked by trauma, anxiety, isolation, and depression. To provide truly ethical and meaningful therapy and education, we must recognize the complexity of neurodivergent people’s experiences by listening to their voices and validating their experiences. True rapport between professionals and the people they serve is based on this understanding, which requires a relationship far more complex and meaningful than any pairing procedure can establish.

Article References

Barbera M.L. (2007) The verbal behavior approach: How to teach children with autism and related disorders. London, UK: Jessica Kingsley Publishers.

Kazemi, E., Shapiro, M. & Kavner, A. (2015). Predictors of intention to turnover in behavior technicians working with individuals with autism spectrum disorder. Research in Autism Spectrum Disorders. 17. 106-115. 10.1016/j.rasd.2015.06.012. 

Lugo, A. M., King, M. L., Lamphere, J. C., & McArdle, P. E. (2017). Developing Procedures to Improve Therapist-Child Rapport in Early Intervention. Behavior analysis in practice, 10(4), 395–401. https://doi.org/10.1007/s40617-016-0165-5

Lugo, A. M., McArdle, P. E., King, M. L., Lamphere, J. C., Peck, J. A., & Beck, H. J. (2018). Effects of Presession Pairing on Preference for Therapeutic Conditions and Challenging Behavior. Behavior analysis in practice, 12(1), 188–193. https://doi.org/10.1007/s40617-018-0268-2

Mahoney, A., Poling, A. Sexual Abuse Prevention for People with Severe Developmental Disabilities. Journal of Developmental and Physical Disabilities, 23, 369–376 (2011). https://doi.org/10.1007/s10882-011-9244-2

McLaughlin, D. M., & Carr, E. G. (2005). Quality of Rapport as a Setting Event for Problem Behavior: Assessment and Intervention. Journal of Positive Behavior Interventions, 7(2), 68–91. https://doi.org/10.1177/10983007050070020401

Hughes, R. B, Robinson-Whelen, S., Raymaker, D., Lund, E. M., Oschwald, M., Katz, M., Starr, A., Ashkenazy, E., Powers, L. E., Nicolaidis, C. (2019) The relation of abuse to physical and psychological health in adults with developmental disabilities. Disability Health Journal. 12(2):227-234. doi: 10.1016/j.dhjo.2018.09.007. 

Schillingsburg, M. A., Bowen, C. N., & Shapiro, S.K. (2014). Increasing social approach and decreasing social avoidance in children with autism spectrum disorder during discrete trial training. Research in Autism Spectrum Disorders, 8(2), 1443-1453. https://doi.org/10.1016/j.rasd.2014.07.013.

Winters, G. M. & Jeglic, E. L. (2017) Stages of Sexual Grooming: Recognizing Potentially Predatory Behaviors of Child Molesters, Deviant Behavior, 38:6, 724-733, DOI: 10.1080/01639625.2016.1197656

• April 2023
• March 2023
• February 2023
• April 2022
• March 2022
• January 2022
• December 2021
• November 2021
• October 2021
• September 2021
• August 2021
• July 2021
• June 2021
• May 2021
• March 2021
• January 2021
• September 2020
• August 2020
• July 2020
• May 2020
• April 2020
• March 2020
• February 2020
• January 2020
• November 2019