FAQs: Ethical Issues in Standardized Decision-Making Algorithms for Behavior Analysts

1. What is a decision-making algorithm (DMA) in behavior analysis?

A DMA is a structured tool—typically a flowchart or decision tree—that guides practitioners through a sequence of yes/no questions toward a recommended clinical action. In ABA, DMAs have been developed for functional assessment selection, treatment planning, and restrictive procedure authorization.

They are designed to reduce variability and increase consistency, but their value depends heavily on the quality of evidence embedded in each decision node and the care with which they are applied to individual clients.

2. Does using a DMA mean I am no longer professionally responsible for the outcome?

No. The BACB Ethics Code (2022) is explicit that behavior analysts retain full professional and ethical responsibility for every clinical decision, regardless of the procedural tools used.

A DMA can inform and structure your reasoning, but it cannot substitute for your professional judgment. If an adverse outcome results from following an algorithm, the practitioner—not the tool—bears accountability.

3. How can I tell if a DMA is empirically supported?

Examine whether each major decision node cites peer-reviewed research from populations and settings comparable to your own. Check whether the studies cited are methodologically sound, recent, and relevant to the specific clinical question the node addresses.

An algorithm that cites only theoretical or procedural literature without empirical support for its specific recommendations should be treated with significant caution before clinical adoption.

4. What are the main ethical risks of DMA use under the BACB Ethics Code?

Key risks include undermining individualization required by section 2.01, obscuring accountability for decisions, embedding outdated or poorly validated evidence, and generating outputs that fail for populations underrepresented in the development research. Practitioners must actively evaluate whether an algorithm's recommendations fit each individual client rather than applying recommendations uncritically.

5. Can DMAs be ethically used with clients from diverse cultural backgrounds?

Only with careful scrutiny. If the DMA was developed using data from culturally homogeneous populations, its decision nodes may not generalize appropriately.

The BACB Ethics Code section 1.07 requires cultural responsiveness, which means examining how an algorithm's inputs and outputs may interact with cultural variables the tool was not designed to capture. Pichardo et al.

(2026) note that caregiver-report data—a common DMA input—can vary systematically across populations, which can distort algorithmic outputs.

6. How should I document my reasoning when I follow or deviate from a DMA?

Documentation should capture both the algorithm's recommendation and the individual client variables that informed your final decision. If you deviated from the DMA, describe the specific clinical rationale.

If you followed it, note which client characteristics supported that recommendation. Transparent documentation of individualized reasoning protects the client, demonstrates ethical practice, and creates a record that supports supervision and review.

7. What obligations do organizations have when they adopt a DMA system-wide?

Organizations must ensure staff have adequate training not just in following the algorithm but in recognizing when to deviate from it. They should establish a formal review schedule to update the DMA as new research emerges, conduct local validation to monitor whether DMA-guided decisions are producing good outcomes, and provide supervision structures that reinforce individualized reasoning rather than rote algorithmic compliance.

8. Are there types of clinical decisions where DMAs are particularly risky?

High-risk applications include decisions about restrictive procedures, discharge criteria, and treatment for severe challenging behavior where individual variation is high and adverse outcomes are serious. Research such as Kok et al.

(2026) on externalizing behavior interventions shows wide variability in individual treatment response—precisely the conditions under which standardized pathways are most likely to miss important client-specific information.

9. What role should functional assessment data play in a DMA?

Functional assessment data should serve as the primary input driving DMA pathways related to treatment selection. However, the quality of that data matters enormously.

Kaur et al. (2026) demonstrated that protective procedures can mask the true behavioral function—meaning DMA inputs derived from masked data will generate systematically flawed outputs.

Any DMA relying on functional assessment data must be paired with procedures that ensure assessment validity.

10. How should CEU training on DMAs inform my professional development plan?

This training establishes a conceptual foundation for critically evaluating algorithmic tools rather than passively adopting them. As a professional development priority, pair this content with review of the current evidence base in your practice area so you can evaluate DMA decision nodes against the literature yourself.

Consider developing a personal checklist for DMA evaluation that addresses empirical support, population fit, measurement requirements, and organizational capacity—then apply it whenever a new algorithm is proposed for adoption in your setting.