The Promises and Possibilities of Artificial Intelligence in the Delivery of Behavior Analytic Services.

Artificial intelligence (AI) has begun to affect nearly every aspect of our daily lives and nearly every industry and profession. Many readers of this journal likely work in one or more areas of behavioral health. For readers who work in behavioral health and who are interested in AI, the purpose of this article is to highlight the pervasiveness of AI research being conducted around many facets of behavioral health service delivery. To do this, we first provide a brief overview of some of the areas within AI and the types of problems each area of AI attempts to solve. We then outline the prototypical client journey in behavioral healthcare beginning with diagnosis/assessment and ending with intervention withdrawal or ongoing monitoring. Next, for each stage in the client journey, we highlight several areas that parallel existing behavior analytic practice where researchers have begun to use AI, often to improve the efficiency of service delivery or to learn new things that improve the effectiveness of behavioral health services. Finally, for those whose appetite has been whet for getting involved with AI, we close by describing three roles they might consider trying out and that parallel the three main domains of behavior analysis. These three roles are an AI tool designer (akin to EAB), AI tool implementer (akin to ABA), or AI tool supporter (akin to practice).

Starting the Conversation Around the Ethical Use of Artificial Intelligence in Applied Behavior Analysis.

Artificial intelligence (AI) is increasingly a part of our everyday lives. Though much AI work in healthcare has been outside of applied behavior analysis (ABA), researchers within ABA have begun to demonstrate many different ways that AI might improve the delivery of ABA services. Though AI offers many exciting advances, absent from the behavior analytic literature thus far is conversation around ethical considerations when developing, building, and deploying AI technologies. Further, though AI is already in the process of coming to ABA, it is unknown the extent to which behavior analytic practitioners are familiar (and comfortable) with the use of AI in ABA. The purpose of this article is twofold. First, to describe how existing ethical publications (e.g., BACB Code of Ethics) do and do not speak to the unique ethical concerns with deploying AI in everyday, ABA service delivery settings. Second, to raise questions for consideration that might inform future ethical guidelines when developing and using AI in ABA service delivery. In total, we hope this article sparks proactive dialog around the ethical use of AI in ABA before the field is required to have a reactionary conversation.

Efficacy of Edible and Leisure Reinforcers with Domestic Dogs.

Preference assessments are often used to identify stimuli that function as potential reinforcers for training or intervention purposes. Specifically, various preference assessment formats have been used to identify preferred stimuli for humans, cockroaches, cotton-top tamarins, tortoises, and wolves, to name a few. However, to date, no study has evaluated the differential efficacy between food and leisure stimuli within domestic dogs. The current study aimed to compare the reinforcing value and efficacy between food and leisure stimuli for domestic dogs by comparing rates of behavior when receiving access to either their top-preferred food or leisure items. Overall results suggest (1) domestic dogs prefer food over leisure items, and (2) food is more likely to function as a reinforcer than leisure items for domestic dog's behavior. These results suggest that dog owners and trainers should consider using food reinforcers over leisure items as reinforcers when attempting to train dogs.

Identifying Trends in the Open-Access Behavior Analytic Literature via Computational Analyses (I): Simple Descriptions of Text.

Published research in scientific journals are critical resources for researchers as primary sources about: what is important in the field, the direction the field is headed, how the field relates to other sciences, and as a historical record for each of these. In this exploratory study, we analyzed the articles of five behavior analytic journals to identify trends in these areas. To do this, we downloaded all available articles (N = 10,405) since the inception of five behavior analytic journals and one control journal. We then used computational techniques to turn the collection of raw text into a structured dataset for descriptive, exploratory analyses. We found consistent differences in the length and variability of published research across behavior analytic journals compared to a control journal. We also found increasing article lengths over time which, combined with the previous finding, may highlight changing editorial contingencies that influence the writing behavior of researchers. Further, we found evidence suggesting distinct (though still connected) verbal communities between the experimental analysis of behavior and applied behavior analysis. Lastly, keyword trends suggest that increased focus on "functional analyses," "problem behavior," and "autism spectrum disorder" currently dominates the research being published in these journals similar to the practitioner arm of behavior analysis. Researchers interested in studying published behavior analytic textual stimuli will find the corresponding open dataset useful. And, for those interested in computational analyses of these data, this first pass at simple descriptions provides a launching point for much fruitful future research. Supplementary Information: The online version contains supplementary material available at 10.1007/s40616-022-00179-4.

Effects of economic context and reward amount on delay and probability discounting.

Steep delay and shallow probability discounting are associated with myriad problem behaviors; thus, it is important to understand factors that influence the degree of discounting. The present study evaluated the effects of economic context and reward amount on delay and probability discounting. Two hundred thirteen undergraduate psychology students completed four delay- or probability-discounting tasks. Participants were exposed to hypothetical narratives involving four bank amounts ($750, $12,000, $125,000, and $2,000,000). The delayed/probabilistic amount was $3,000 for the two smaller bank amounts and $500,000 for the two larger bank amounts. The discounting tasks included five delays to, or probabilities of, receipt of the larger amount. The area under the empirical discounting function was calculated for each participant. Participants discounted delayed and uncertain outcomes more when the bank amount was smaller than the outcome (i.e., the economic context was low). Participants discounted the delayed larger amounts less than delayed smaller amounts, even when the relative economic context was the same. In contrast, probability discounting did not differ across magnitudes, which suggests that economic context may attenuate the magnitude effect in probability discounting. The results further highlight the importance of considering the economic context in delay and probability discounting.

What's the Big IDEA? A Preliminary Analysis of Behavior Analysts' Self-Reported Training in and Knowledge of Federal Special Education Law.

Many behavior analysts currently work in school settings or with individuals who may qualify for educational services through federal special education law. However, it remains unclear what training, if any, behavior analysts receive in this law. Behavior analysts have an ethical responsibility to practice within their scope of competency and in compliance with legal regulations. Thus, it is important to determine whether behavior analysts practicing in the United States are adequately prepared and familiar with federal special education law. The current study consisted of a survey wherein respondents answered questions pertaining to the relevance of federal special education law, their familiarity with core terminology, and the alignment between the law and the Professional and Ethical Compliance Code for Behavior Analysts (Behavior Analyst Certification Board, 2016). Respondents' self-report indicates that behavior analysts hold conflicting views on how federal special education law aligns with and influences their role as service providers. As such, practitioners and agencies alike may benefit from explicit clarification of the responsibility Board Certified Behavior Analysts have to seek training in and adhere to federal special education law.

Training Staff to Create Equivalence-Based Instruction Materials in Qualtrics.

Staff training is an important line of research to ensure that clinicians in the field of applied behavior analysis provide quality services. One approach to providing training involves the use of asynchronous training materials in which the trainer and trainee do not need to be physically present at the same time. This allows for training despite limited numbers of trainers or geographic restrictions. The purpose of this study was to train participants to create equivalence-based instruction (EBI) materials in Qualtrics, a commercial survey software package. In the first phase of the study, participants experienced a training package consisting of a video model and task analysis that described how to create EBI materials. Phase 2, evaluated whether the EBI materials created in Phase 1 led to equivalence-based responding. Results indicated that the training was effective for training staff and that staff found the training socially acceptable. Results also demonstrated that the EBI materials resulted in the learning of the trained and derived relations. Supplementary Information: The online version contains supplementary material available at 10.1007/s40732-021-00497-4.

Verbal behavior related to drug reinforcement in polysubstance cannabis users: Comparison across drugs.

Verbal reports of drug effects are often used in behavioral pharmacology. Two reports related to reinforcement are drug use (Harford, 1978; Liu et al., 2018) frequency and drug preference. Anecdotally, some individuals may specify a favorite/preferred drug (e.g., psilocybin) despite using another drug more frequently (e.g., tobacco). Research comparing these two measures has led to contradictory findings and included ratings from participants who may not have experience with the rated drugs. No comparisons have been made between use frequency and preference across multiple drugs in polysubstance users. To compare use frequency and preference for drug classes, and examine relations across drug classes, individuals reporting polysubstance use (N = 428) provided frequency and preference ratings for nine drug classes. Mean ratings showed smoked tobacco, alcohol, and cannabis were the most frequently used and most preferred drugs. Mean ratings showed 3,4-Methylenedioxymethamphetamine (MDMA) and classic hallucinogens were the least frequently used and least preferred drugs. However, more divergence between use frequency and preference was observed when these metrics were examined among individuals. Correlation coefficients between use frequency and preference were lower than previously published literature. The majority of polydrug comparisons were nonsignificant, and correlations between different drug classes differed depending on whether use frequency or preference was examined. Verbal reports about use frequency are likely not strongly predictive of verbal reports about the same drug preference. Clinicians and researchers should recognize that different verbal reports related to drug reinforcement might be proxies for distinct aspects of reinforcement and should consider these implications for assessment and research findings. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

The Influence of Episodic Future Thinking and Graphic Warning Labels on Delay Discounting and Cigarette Demand.

Delay discounting and operant demand are two behavioral economic constructs that tend to covary, by degree, with cigarette smoking status. Given historically robust associations between adverse health outcomes of smoking, a strong preference for immediate reinforcement (measured with delay discounting), and excessive motivation to smoke cigarettes (measured with operant demand), researchers have made numerous attempts to attenuate the extent to which behaviors corresponding to these constructs acutely appear in smokers. One approach is episodic future thinking, which can reportedly increase the impact of future events on present decision making as well as reduce the reinforcing value of cigarettes. Graphic cigarette pack warning labels may also reduce smoking by increased future orientation. Experiment 1 evaluated the combined effects of episodic future thinking and graphic warning labels on delay discounting; Experiment 2 evaluated solely the effects of episodic future thinking on delay discounting and operant demand. We observed no statistically significant effects of episodic future thinking when combined with graphic warning labels or when assessed on its own. These results serve as a call for further research on the boundary conditions of experimental techniques reported to alter behaviors associated with cigarette smoking.

Predicting changes in substance use following psychedelic experiences: natural language processing of psychedelic session narratives.

Background: Experiences with psychedelic drugs, such as psilocybin or lysergic acid diethylamide (LSD), are sometimes followed by changes in patterns of tobacco, opioid, and alcohol consumption. But, the specific characteristics of psychedelic experiences that lead to changes in drug consumption are unknown.Objective: Determine whether quantitative descriptions of psychedelic experiences derived using Natural Language Processing (NLP) would allow us to predict who would quit or reduce using drugs following a psychedelic experience.Methods: We recruited 1141 individuals (247 female, 894 male) from online social media platforms who reported quitting or reducing using alcohol, cannabis, opioids, or stimulants following a psychedelic experience to provide a verbal narrative of the psychedelic experience they attributed as leading to their reduction in drug use. We used NLP to derive topic models that quantitatively described each participant's psychedelic experience narrative. We then used the vector descriptions of each participant's psychedelic experience narrative as input into three different supervised machine learning algorithms to predict long-term drug reduction outcomes.Results: We found that the topic models derived through NLP led to quantitative descriptions of participant narratives that differed across participants when grouped by the drug class quit as well as the long-term quit/reduction outcomes. Additionally, all three machine learning algorithms led to similar prediction accuracy (~65%, CI = ±0.21%) for long-term quit/reduction outcomes.Conclusions: Using machine learning to analyze written reports of psychedelic experiences may allow for accurate prediction of quit outcomes and what drug is quit or reduced within psychedelic therapy.

Sexual Arousal Discounting: Devaluing Condom-Protected Sex as a Function of Reduced Arousal.

Sexual discounting is a growing area of research aimed at identifying factors that reduce people's reported willingness to have safe sex. One commonly reported reason for condom non-use is that a condom reduces sexual arousal. However, researchers have yet to isolate the specific role of sexual arousal using a sexual discounting framework. We extended past research by measuring how sexual arousal reduced people's willingness to have condom-protected sex ("sexual arousal discounting": SAD). College students (n = 379; 67.5% females) selected partners they most wanted and least wanted to have sex with and were randomized to one of two groups. In one group, participants rated their willingness to have sex with a condom if their own arousal decreased (from 100 to 10%) from condom use. The other group completed the same task except their partner's arousal decreased from condom use instead of their own. We observed a three-way interaction between arousal levels, most versus least desirable partners, and self versus partner groups. Participants' willingness to have condom-protected sex systematically reduced as a function of sexual arousal. This was observed more with the most (vs. least) desirable partner and in the self-arousal (vs. partner-arousal) group but only when the partner was their least desirable. Men (vs. women) displayed more arousal discounting but only with the most desirable partner. Finally, higher arousal discounting was associated with lower safe-sex self-efficacy and higher reported frequency of unprotected sex in the past 3 months. This study demonstrates how reduced sexual arousal from condom use can be measured as a factor influencing sexual risk-taking.

Scaling N from 1 to 1,000,000: Application of the Generalized Matching Law to Big Data Contexts.

The generalized matching law (GML) has been used to describe the behavior of individual organisms in operant chambers, artificial environments, and nonlaboratory human settings. Most of these analyses have used a handful of participants to determine how well the GML describes choice in the experimental arrangement or how some experimental manipulation influences estimated matching parameters. Though the GML accounts very well for choice in a variety of contexts, the generality of the GML to all individuals in a population is unknown. That is, no known studies have used the GML to describe the individual behavior of all individuals in a population. This is likely because the data from every individual in the population has not historically been available or because time and computational constraints made population-level analyses prohibitive. In this study, we use open data on baseball pitches to provide an example of how big data methods can be combined with the GML to: (1) scale within-subjects designs to the population level; (2) track individual members of a population over time; (3) easily segment the population into subgroups for further analyses within and between groups; and (4) compare GML fits and estimated parameters to performance. These were accomplished for each of 2,374 individuals in a population using 8,467,473 observations of behavior-environment relationships spanning 11 years. In total, this study is a proof of concept for how behavior analysts can use data-science techniques to extend individual-level quantitative analyses of behavior to the population-level focused on domains of social relevance.

Sexual discounting: A systematic review of discounting processes and sexual behavior.

Behavioral processes underlying sexual behavior are important for understanding normal human functioning and risk behavior leading to sexually transmitted infections (STIs). This systematic review examines delay and probability discounting in human sexual behavior through synthesis of 50 peer-reviewed, original research articles. Sixteen studies focusing exclusively on monetary delay discounting found small effect size positive correlations with sexual risk behaviors. Eleven studies examined delay or probability discounting of sexual behavior itself using tasks that varied duration, frequency, or quality of sex to determine value. Results show delay and uncertainty of sex causes systematic decreases in value. These studies also show consistent medium effect size relationships between sexual discounting measures and sexual health and substance use, supporting utility above and beyond monetary discounting. Twenty-three studies have modeled clinically relevant decision-making, examining effects of delay until condom availability and STI contraction probability on condom use. Observational and experimental designs found condom-use discounting is elevated in high-risk substance use populations, is sensitive to context (e.g., partner desirability), and is more robustly related to sexual risk compared with monetary discounting or condom use decisions when no delay/uncertainty was involved. Administering cocaine, alcohol, and, for some participants, methamphetamine increased condom-use discounting with minimal effect on monetary discounting or condom use when no delay/uncertainty was involved. Reviewed studies robustly support that sexual behavior is highly dependent on delay and probability discounting, and that these processes strongly contribute to sexual risk. Future research should exploit these systematic relationships to design behavioral and pharmacological approaches to decrease sexual risk behavior. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Descriptive and normative ethical behavior appear to be functionally distinct.

Many philosophers and scientists have described a scientific approach to ethical behavior. Historically, ethical behavior has been categorized as descriptive (i.e., what is right) or normative (i.e., why it is right). Whether this topographical distinction is functionally relevant is unknown. In 2 experiments, participants chose what behavior was correct and why. In Experiment 1, participants did not agree on either of these measures. Normative ethical behaviors were also well described by common Western theories of bioethics (i.e., consequentialism, deontology, and virtue theory). In Experiment 2, manipulating the ethical context led to within-subject changes in responses to what, why, or both. Importantly, change in what rarely coincided with change in why, suggesting descriptive and normative ethical behaviors are functionally distinct. A visual-descriptive model describing a functional approach to descriptive and normative ethical behavior is provided. Behavior analysts interested in observing, measuring, and changing ethical decision-making should consider collecting data on descriptive and normative ethical behaviors.

A Guide to Establishing Ethics Committees in Behavioral Health Settings.

Ethical statements typically involve rules. All rules can vary in accuracy and specificity depending on the context to which they are applied. Codes of ethics often involve ethical rules that are written generally to cover the wide-ranging set of possible situations that any one member of the profession may encounter. But, despite being written generally, codes of ethics are applied to specific situations that professional members encounter. The application of general rules to specific contexts can sometimes be challenging and complex. Health care organizations have several options to help their employees behave ethically. One approach is to appoint a single ethics coordinator. In contrast, the dominant approach in most health care organizations is to develop an organizational ethics committee (Moon Pediatrics, 143(5), e20190659, 2019). Despite the popularity of the ethics committee in other professions, the extent to which organizations that provide applied behavior analysis services have established and operate ethics committees is unknown. Ethics coordinator roles and ethics committees both have benefits and drawbacks. This article reviews the benefits and drawbacks of appointing an ethics coordinator and establishing an ethics committee. And, for interested organizations, this article outlines the steps and considerations that organizations can use to guide the creation of an ethics committee.

Comparison of multiplicative and additive hyperbolic and hyperboloid discounting models in delayed lotteries involving gains and losses.

Many day-to-day decisions may involve risky outcomes that occur at some delay after a decision has been made. We refer to such scenarios as delayed lotteries. Despite human choice often involves delayed lotteries, past research has primarily focused on decisions with delayed or risky outcomes. Comparatively, less research has explored how delay and probability interact to influence decisions. Within research on delayed lotteries, rigorous comparisons of models that describe choice from the discounting framework have not been conducted. We performed two experiments to determine how gain or loss outcomes are devalued when delayed and risky. Experiment 1 used delay and probability ranges similar to past research on delayed lotteries. Experiment 2 used individually calibrated delay and probability ranges. Ten discounting models were fit to the data using a genetic algorithm. Candidate models were derived from past research on discounting delayed or probabilistic outcomes. We found that participants' behavior was best described primarily by a three-parameter multiplicative model. Measures based on information criteria pointed to a solution in which only delay and probability were psychophysically scaled. Absolute measures based on residuals pointed to a solution in which amount, delay, and probability are simultaneously scaled. Our research suggests that separate scaling parameters for different discounting factors may not be necessary with delayed lotteries.

A Proposed Process for Risk Mitigation During the COVID-19 Pandemic.

Recent executive orders have led some applied behavior analysis (ABA) providers to interpret themselves as "essential personnel" during the COVID-19 pandemic. In this article, we argue against a blanket interpretation that being labeled "essential personnel" means that all in-person ABA services for all clients should continue during the COVID-19 pandemic. We believe this argument holds even if ABA providers are not in a jurisdiction currently under an active shelter-at-home or related order. First, we provide a brief description of risks associated with continued in-person ABA service delivery, as well as risks associated with the temporary suspension of services or the transition to remote ABA service delivery. For many clients, continued in-person service delivery carries a significant risk of severe harm to the client, family and caregivers, staff, and a currently overburdened health care system. In these situations, ABA providers should temporarily suspend services or transition to telehealth or other forms of remote service delivery until information from federal, state, and local health care experts deems in-person contact safe. In rare cases, temporary suspension of services or a transition to remote service delivery may place the client or others at risk of significant harm. In these situations, in-person services should likely continue, and ongoing assessment and risk mitigation are essential.

Kratom (Mitragyna speciosa): User demographics, use patterns, and implications for the opioid epidemic.

BACKGROUND: Kratom, a Southeast Asian plant with opioid-receptor mediated effects, has emerged as a potential substance of abuse, with limited data on its use and effects. This study characterized kratom user demographics, use patterns, and perceived drug effects. METHODS: A cross-sectional, anonymous online survey was conducted between January and December 2017. RESULTS: 2,798 kratom users - mean age 40 (SD = 12); predominantly White (90 %), female (61 %), and located in the US (97 %) - completed the survey. Kratom was primarily taken orally in doses of 1-3 g (49 %), with daily use (59 %) being most common. Kratom was used for pain (91 %), anxiety (67 %), and depression (65 %), with high ratings of effectiveness. 1,144 (41 %) used kratom to stop or reduce prescription or illicit opioid use, citing decreased opioid withdrawal and craving related to kratom use, with 411 reporting >1-year continuous abstinence from opioids attributed to kratom use. Roughly one-third of respondents reported adverse effects of kratom, largely rated as mild in severity and lasting ≤24 h. Seventeen participants (0.6 %) sought treatment for adverse effects. Fifty-six individuals (2 %) met DSM-5 criteria for a past-year moderate or severe kratom-related substance use disorder (SUD). When asked how troubled they felt regarding their kratom use, the mean (SD) rating was 3.2 (9.8) on a scale from 0 to 100. CONCLUSION: Kratom is used among White, middle-aged Americans for symptoms of pain, anxiety, depression, and opioid withdrawal. Although regular use was typical, kratom-related SUD and serious adverse effects were uncommon. Additional research on kratom epidemiology and pharmacology is imperative in light of the present opioid epidemic.