Guest Post by Kimberly Marshall, University of Oregon
Author’s Note
I am the program director and an assistant teaching professor in the Applied Behavior Analysis master’s program at the University of Oregon. My research interests are in the effective training of ethical, compassionate, and culturally responsive behavior analysts and the successful dissemination of the science of behavior analysis. I am particularly interested in evaluating the impact of jargon on successful practice and dissemination. In this article, I will share about my work as a volunteer at the Association for Science in Autism Treatment (ASAT), helping to educate consumers about what does and doesn’t work in autism treatment. I’d like to express my appreciation to the brilliant members of the scientific review committee at ASAT and ASAT’s tireless leader, Dr. David Celiberti, who have all been integral to the work discussed in this post.
If you are or have been a practitioner in ABA, you have likely experienced a situation similar to the following:
A family comes into your clinic excited to finally come off the wait list and start ABA services. During the intake, they share about the interventions they have started while they were waiting for ABA services, which they plan to continue using simultaneous with ABA. You learn that one of the client’s other providers is using Rapid Prompting Method (RPM), and the family intends to continue with these services. (For any readers who are unfamiliar with RPM, see Daly & Celiberti (2021) and the American-Speech-Language-Hearing Association’s Position Statement.) You know that RPM is similar to Facilitated Communication and, consequently, lacks evidence and can lead to harm.
As the behavior analyst on this case, what do you do?
An Ethical Imperative To Inform
The Ethics Code (3.12 Advocating for Appropriate Services; Behavior Analyst Certification Board, 2020) requires you to “advocate for and educate clients and stakeholders about evidence-based assessment and behavior change intervention procedures.” However, limited guidance is provided on how to go about this. For example, Schreck & Miller (2010) described a comprehensive process that behavior analysts can use to determine if a client- or stakeholder-suggested treatment has research evidence. However, when the answer is no, how should this be shared with the family? In general, what is the best way of dissuading dissuading the client or stakeholder from embracing a non-empirically supported treatment?
As a clinician, I received the advice to throw a lot of scientific evidence at the family, share the harms and dangers of the treatment, and tell them why they need to stop the treatment (e.g., RPM) right away. But we seem to have little evidence that this approach works and for all we know it may, in fact, have a detrimental impact on our relationship with the family.
ASAT’s Treatment Summaries
I volunteer as an Action Editor for treatment summaries for the Association for Science in Autism Treatment (ASAT), a non-profit organization that promotes safe and effective science-based treatments for people with autism.
To accomplish its mission, ASAT hosts a website with comprehensive information about empirically supported and non-empirically supported autism treatments and publishes a monthly newsletter with new and updated information about autism treatments.
The purpose of ASAT’s treatment summaries is to provide clients, stakeholders, and professionals with information about common autism treatments, both empirically supported and non-empirically supported. This includes:
- explaining what the intervention entails;
- presenting an overview of the research that does or does not support its use;
- making recommendations related to the use of the treatment; and
- in the case of non-empirically supported interventions, sharing potential alternative, empirically supported interventions.
The intended audience includes autistic people, families, community members, educators, and other professionals outside of ABA (moving forward, I will refer to this broad category of audiences as consumers). Given this consumer audience, treatment summaries are written to be user friendly and accessible (to the greatest extent possible). On its face, this may seem fairly straightforward; however, as the vignette above indicates, effective dissemination to consumers is complex, particularly when non-empirically supported interventions are often easier to access, well marketed, and presented in a manner that raises (false) hope.
Applied Verbal Behavior: Information as Persuasion
Over the last few years, along with colleagues at ASAT, I have worked on trying to figure out how to effectively disseminate scientific information to a wide audience who probably, much of the time, does not have a great interest or excitement for the messages we are sharing. The goal of our treatment summaries also goes beyond simply sharing information. Treatment summaries are intended to “change someone’s mind.” We want to not only explain to clients and consumers that RPM lacks empirical evidence but also persuade them to change their beliefs about RPM and take action according to these new beliefs. Persuading consumers to be wary of and stop using non-empirically supported treatments will likely require that behavior analysts use consumers’ preferred methods of communication, rather than the communication methods that they are accustomed to using within their scientific community (Detrich & Critchfield, 2025).
Changing someone’s perspective is a multi-dimensional process impacted by a variety of variables including the person’s learning history, the value they place on different outcomes, the value they place on different types of information, and the way in which that information is presented. In the psychological literature, Sharot et al. (2022) suggested that belief changing is essentially an economic process with change driven by the relative value of potential outcomes of the new belief versus historical outcomes of the old belief. This aligns well with a behavior analytic view. A person’s history of reinforcement in the same or similar situations drives their beliefs about what is most likely to contact reinforcement in the future.
Changing someone’s mind about something like RPM becomes complicated because a consumer’s experience in the past with RPM and other pseudo-scientific interventions may have been good. They may have experienced their child “communicating” with them in ways that they did not in the past. In such cases, ongoing use of RPM may appear to be the step that is most likely to contact reinforcement in the future. So, how do we show that reinforcement can be contacted another way?
This involves a nuanced approach. If we too forcefully state that RPM does not work and it is harmful, then this misalignment with their previous experience may lead them to believe that we cannot be trusted. This could lead them to assume that anything else we say may be a potential falsehood or misrepresentation since our predictions about reinforcement are different from their experience. In this case, they are likely to seek reinforcement in the form of approval for their current choices from their social community, including current providers of these pseudo-scientific treatments, who hold beliefs similar to theirs (and are likely influenced by other contingencies (e.g., financial incentives) as well). However, if we are not clear about potential harm, then there would be no reason for them to change their behavior. As such, a balanced strategy must involve presenting motivating stimuli (e.g., highlighting the potential reinforcement available) for stopping current interventions and selecting new interventions, while not presenting stimuli that will motivate consumers to avoid our communications.
At ASAT we have had many discussions about the best strategies to accomplish these goals and what I’ll share with you are several of the practices we have landed on. While some of these practices have empirical support, many do not and I acknowledge the inherent limitation of that. However, I am hopeful that sharing our steps and process can lead to some interesting discussion in the field and hopefully to some much-needed evaluation (let me know if you are interested in working on this together, I’m in!). As I discuss these strategies, I will provide examples of our authors’ use of these strategies in treatment summaries of non-empirically supported interventions. Many of these strategies are also used in our discussions of empirically supported interventions. If you are interested, you can see examples of those by looking at the “What works” interventions: https://asatonline.org/for-parents/learn-more-about-specific-treatments/
Six Persuasive-Informative Strategies
Strategy 1: Previewing Level of Evidence
Research suggests that previewing a treatment’s level of evidence may be an important strategy when presenting information about non-empirically supported treatments. Paynter et al. (2019) evaluated an “optimized-debunking” strategy as part of a professional development training for individuals in the field of autism, to decrease teacher and staff misperceptions about autism treatments. One of their optimized strategies was “warning.” They suggested that warning someone before exposing them to misinformation has a few potential benefits: (1) it decreases the likelihood that exposure to the information leads to belief; (2) it also decreases the likelihood that after learning that it was misinformation, a person must retrospectively go back and re-read the information from a new lens; and (3) it improves recall, making it more likely that the person will rely on accurate information when stimulus cues are present at a later time that could lead to responding with the misinformation.
There are two primary ways that consumers are likely to access our treatment summaries, and both include warnings about level of evidence:
(1) They receive the Science in Autism Treatment (SIAT) newsletter and after reading the blurb introducing the article, they click the link to access the full article. These blurbs consistently highlight the level of evidence for the treatment summary, for example:
In this treatment summary, Scott Myers, MD, and I discuss the research on craniosacral therapy. Craniosacral therapy relies on the premise that there is fluid around our brains and spinal cords that pulses or flows and can sometimes be blocked or disrupted. Therapists purport that they can manipulate the cranium to restore normal pulses or flows and thereby alleviate symptoms. Unfortunately, research on the use of craniosacral therapy across medical conditions and populations, including individuals with autism, has not provided strong evidence for its effectiveness. Learn more about craniosacral therapy and the caution that should be exercised in considering the use of this treatment. — Dr. Kimberly Marshall, Treatment Summary Coordinator
(2) They find the treatment on the treatment summaries main page (https://asatonline.org/for-parents/learn-more-about-specific-treatments/) where treatments are broken down into three categories (psychological, educational, and therapeutic interventions; behavioral interventions; and biomedical interventions) according to their level of evidence (what works, what needs more research, what doesn’t work or is untested). For example, Craniosacral Therapy is presented under the category Biomedical Interventions within What doesn’t work or is untested.
As shared above, all our treatment summaries start with a description of the treatment. Within this description, we share what the treatment looks like and what proponents assert it accomplishes. Hence, warning readers prior to providing this information is essential to avoid them reading these descriptions (including the purported outcomes of the intervention) as facts that back the value of the non-empirically supported treatment.
Strategy 2: Highlighting Goals that Consumers are Trying to Achieve
Another of the optimized debunking strategies used by Paynter et al. (2019) was asking participants to affirm their values. They suggested that affirming one’s values made it less difficult to accept information that was misaligned with one’s worldview and current beliefs. Other researchers have also highlighted the effectiveness of focusing on societal values when disseminating scientific information (Detrich & Critchfield, 2025). While we cannot ask our readers to self-affirm their goals (or know exactly what is important to each reader), we can write about things that consumers supporting people with autism and, consequently, using these non-empirically supported treatments, often care about. Thus, in our treatment summaries, we highlight the goals of the non-empirically supported treatment. This can motivate a reader to keep reading as it is quickly clear that the article will be focused on the goals that they are hoping to achieve. For example:
From the treatment summary of DIR/Floortime:
These exchanges are intended to help children regulate their bodies and attention, form secure relationships, and gradually develop more complex thinking and communication skills (Singh & Pattieshaw, 2025).
As you will see below, we also focus on values when we recommend evaluating interventions and suggest alternative intervention options.
Strategy 3: Simplifying Research Evidence and Incorporating Credible Sources
People listen to those they believe are credible – this may be related to perceived expertise or because they share common roles and experiences. Paynter et al. (2019) and Detrich & Critchfield (2025) have highlighted the importance of source credibility, in particular noting that people may be more likely to adopt interventions when they perceive the source of the information to be credible. A similar case may likely be made for not adopting or stopping the use of non-empirically supported treatments. Receiving information from a credible source may make the issues and harms more believable.
Our summaries are typically written by professionals and not individuals who parents may relate to from the perspective of shared roles (although ASAT does use this strategy quite a bit in other areas; e.g., Three Mothers Share Their Journeys Navigating Profound Autism: Part 1 of a Two-Part Interview with Lorri Unumb, Judith Ursitti, and Eilem Lamb). But we do include references to sources that we believe hold broad credibility. Most summaries include findings of large task force reports and, when available, positions statements from medical associations and discipline-specific organizations and agencies. These sources likely hold substantial credibility and familiarity in our broader society. ASAT’s editors, authors, and volunteers are primarily behavior analysts, which might lead consumers to perceive information they share as biased. Use of task force reports and position statements may decrease the likelihood of consumers dismissing summaries due to perceived bias. For example, from the treatment summary of Facilitated Communication:
It is also important to note that numerous professional organizations including, but not limited to, the American Academy of Pediatrics (AAP) and the American Psychological Association (APA), have issued position statements advising against the use of FC in treating autism (please see the extensive list below with links to the original statements) (Celiberti et al., 2024).
Strategy 4: Promoting Group and Individual Level Evaluation Tied to Values
A related strategy is encouraging evaluation of treatments (excluding harmful treatments and/or those that have been found to be ineffective) at both the group and individual level. Recommendations for evaluation are focused on practitioners, researchers, and caregivers.
Though it may seem controversial to many behavior analysts, we have been working to find ways to temper the message around non-empirically supported treatments that lack evidence but are not directly linked to significant harm. Our goal here is twofold; (1) to choose our battles (we can save our strongest messaging for treatments that are dangerous) and (2) to provide people with ideas for how they might evaluate the use of these treatments. In other words, if we cannot decrease your likelihood of using a non-empirically supported treatment, can we get you to evaluate its use and allow the outcomes to speak for themselves? If so, the result might be that we have a greater chance of preventing further harm to the client. For example, from the treatment summary of Relationship Development Intervention (RDI®):
In the meantime, if families decide to use RDI® (as many already do), they are encouraged to work with providers to carefully define the treatment targets, implement RDI® carefully in a consistent and thorough manner, collect objective data to measure benefits, and to use those data with their providers to make decisions about the path forward (Celiberti et al., 2025).
This approach is largely instructed by what we know about countercontrol (Skinner, 1953). Pushing hard for readers to change their use of non-empirically supported treatments that they have likely bought into and potentially even contacted reinforcement for (albeit likely superstitious reinforcement or indirect social reinforcement from providers of pseudo-scientific treatments), can lead to emotional responses including vocal or nonvocal verbal behavior about our intelligence, or statements opposing our beliefs, and most importantly, closing the tab or window on their screen.
To avoid this, we need to balance between providing good information that could serve as a motivating operation for them to question their held beliefs on the treatment (engage in some new verbal behavior), while not pushing so hard that we motivate behavior that ceases their contact with our message.
This approach has the added benefit of focusing the reader’s attention on what they are hoping to get out of the treatment. By encouraging readers to reflect on their values (calling back to Strategy 2) and evaluating if the intervention is producing outcomes aligned with those values, we are essentially asking them to operationally define the behavior and conduct a treatment analysis. Given that they may be better convinced by seeing it for themselves than any research we can provide, we believe this approach has benefit (I add a caveat here, that this is only true if it is leading to consumers actually engaging in evaluation of these non-empirically supported treatments, which we do not currently have a way of evaluating).
For researchers who are proponents of these non-empirically supported interventions, we recommend research methods and questions that would need to be asked to provide empirical support for the intervention. For example: from the treatment summary of Natural Language Acquisition Protocol for Gestalt Language Development:
We recommend NLA proponents support future research investigations by (1) defining clear features of GLP so that learner profiles can be reliably classified, (2) detailing objective methods for interpreting the “meanings” of gestalts, (3) operationally defining desired outcomes at each stage, (4) standardizing data analysis methods in a way that can be reliably replicated, (5) specifying what constitutes progress (versus no progress or regression), (6) clearly detailing how to implement recommended strategies, and (7) developing fidelity assessments/checklists for ensuring accurate and consistent implementation. It is imperative that independent researchers conduct studies on the NLA protocol using strong research methodologies to determine whether the stages and recommended practices are valid and effective” (Forbes et al., 2024).
An exception: When a treatment summary is focused on an intervention that has immediate detrimental impacts on health and safety, we include clear statements about the risks and do not suggest evaluation. For example, from the treatment summary of Bleach Therapy:
According to the FDA, people who have ingested chlorine dioxide solutions have experienced severe vomiting, severe diarrhea, life threatening low blood pressure, dehydration, and acute liver failure – all signs of bleach poisoning (Cicero, 2025).
Strategy 5: Offering Alternatives
After discussing steps that a consumer might take related to the non-empirically supported intervention, we offer suggestions for empirically supported interventions that focus on the same goals. This aligns with many of our other strategies in terms of focusing on values. Identifying what can lead someone toward their goals may be more persuasive for consumers than focusing on what they should stop doing. For example, from the treatment summary of Rapid Prompting Method:
Interventions which are recommended for treating autism, specifically targeting communication skills, include the Picture Exchange Communication System, Augmentative and Alternative Communication (AAC), Verbal Behavior, Functional Communication Training, Incidental Teaching, Pivotal Response Treatment/Natural Language Paradigm, Discrete Trial Instruction, and other behavior analytic teaching procedures (Daly & Celiberti, 2021).
Strategy 6: Using Everyday Language and Relatable Parallels
If you have read my work or know me, then you know this is my soapbox and I won’t belabor it because many expert behavior analysts have spoken about this extensively (Bailey, 1991; Critchfield et al., 2017; Jarmolowicz et al., 2008). Jargon is a problem – the technical terminology of behavior analysis, for example, is not understood by people outside of the field (Marshall et al., 2025; Jarmolowicz et al., 2008) and elicits negative emotional responses in those people (Critchfield et al., 2017; Marshall et al., 2023). Therefore, since our primary audience is consumers rather than professionals, we do our best to use everyday terms in treatment summaries.
Admittedly, this is not always easy, as explaining the basis of medical treatments or even some complex therapeutic or behavioral treatments in the absence of scientific terms can be a challenge. For example, from the treatment summary of The Wilbarger Brushing Protocol:
In particular, this approach asserts that symptoms of avoidance are due to an individual’s oversensitivity to sensory input becoming over-reactive, overwhelmed, and hyper-reactive to touch (i.e. tactile defensiveness). In other words, an individual may be trying to get away from the sensation (Keleher et al., 2024).
One strategy that we are increasingly employing in some of our summaries is to use relatable parallels to help consumers understand some of the more complex content. For example, from the treatment summary of Relationship Development Intervention (RDI®):
Furthermore, consumers should be aware that some providers may boast an evidence-base by mere association with other interventions that possess scientific support. This argument is flawed as illustrated by the examples below:
A bobcat has characteristics similar to cats.
• Cats can be safe pets.
• Bobcats can be safe pets.
Now we are applying this to autism intervention.
Intervention A overlaps a bit with Interventions B and C.
• Interventions B and C are evidence-based.
• Intervention A is therefore evidence-based (Celiberti et al., 2025)
Conclusion
I share these strategies not as a definitive list but as a starting point to help others who are struggling with the dissemination of scientific information to consumers, or potential consumers, of ABA. Despite the comprehensive discussion around and empirical grounding of these strategies, it is evident that we have much more work to do to solidify recommendations for effective dissemination. There are more strategies that have been shown to support “debunking” misinformation (e.g., highlighting the correct message and using graphical representation; Paynter et al., 2019) that should be considered and trialed. Most importantly, we need to evaluate the impact of our current strategies on supporting families and other stakeholders with the daunting task of finding effective, empirically supported interventions for the people with autism who are important to them.
The consumers coming to our website (and exemplified in the vignette at the start of this post) are often in a position of vulnerability, desperately trying to find interventions that will support those people with autism that they care about. They may be willing to try anything that promises success in decreasing challenging behavior, improving communication, or increasing adaptive skills. This is why it is essential that as behavior analysts, in a position to support these consumers, we continue to work towards and prioritize successful dissemination of scientific information.
References
Bailey, J.S. (1991). Marketing behavior analysis requires different talk. Journal of Applied Behavior Analysis, 24(3), 445-448. PDF
Behavior Analyst Certification Board. (2020). Ethics code for behavior analysts. https://bacb.com/wp-content/ethics-code-for-behavior-analysts/
Jarmolowicz, D.P., Kahng, S., Ingvarsson, E. T., Goysovich, R., Heggemeyer, R., & Gregory, M.K. (2008). Effects of conversational versus technical language on treatment preference and integrity. Intellectual and Developmental Disabilities, 46(3), 190–199. PDF
Sharot, T., Rollwage, M., Sunstein, C.R., & Fleming, S.M. (2023). Why and when beliefs change. Perspectives on Psychological Science, 18(1), 142-151. PDF
Skinner, B.F. (1953). Science and human behavior. The Free Press.