Remote Supervision Done Right

Remote supervision can be just as rigorous as in-person supervision, but only when it holds to the same standards: direct observation, objective measurement, and actionable performance feedback (Ferguson, Craig, & Dounavi, 2019). A strong session starts with a one-sentence target, such as observing prompting and error correction during natural environment teaching, and uses a short fidelity checklist of no more than ten items. At least one data point is collected live — percentage of steps correct, rate of prompts, or a treatment-integrity score — and the session ends with a single skill assignment for next time. The catch unique to telehealth is that poor camera angles and audio create false confidence: the session looks fine until outcomes drop. Requiring multiple camera views when possible, running a quick pre-session tech check, and placing the camera at a useful height protect the observation, and when visibility is poor the right move is to pause clinical decisions and switch to training-only mode.

Telehealth also raises confidentiality risks that can be violated quickly without anyone intending to. Overheard sessions, unapproved recording, unsecured devices, and casually texting protected health information are all real hazards, so a supervisor confirms privacy at the start, uses headphones when possible, sticks to secure platforms, and is explicit that no recording occurs without written consent (Behavior Analyst Certification Board, 2020). A second quiet risk is that caregivers can drift from support into implementation without training, consent, or a defined role. The fix is to define the caregiver's role out loud at the start — for example, focusing on safety, materials, and minimal prompting — give them a simple script for what to do and when to step in, and treat any actual implementation as behavioral skills training rather than a normal session. It also helps to track caregiver participation briefly as independent, coached, or directed, because prompts from a caregiver change the data.

Remote feedback fails when it is vague, because "good job" does not build clinical skill. A reliable formula is behavior, then impact, then fix — for instance, noting that reinforcement was delayed several seconds, that motivation dropped as a result, and that reinforcement should come within a second or two — delivered as one priority correction at a time. Pairing that with an immediate model and rehearsal, tracking integrity from baseline to post-coaching within the same session, and ending on a concrete micro-goal turns feedback into measurable improvement. All of this works best inside a repeatable structure: a couple of minutes to set the target and confirm tech and privacy, a block of direct observation with the checklist, a few minutes of immediate data-based feedback, a few minutes of modeling and rehearsal, and a final couple of minutes to document and assign the next step. Done this way, remote supervision stays structured, observable, and defensible, which is exactly what keeps it clinically valid (Sellers, Valentino, & LeBlanc, 2016).

References

Behavior Analyst Certification Board. (2020). Ethics code for behavior analysts.

Ferguson, J. L., Craig, E. A., & Dounavi, K. (2019). Telehealth as a model for providing behaviour analytic interventions to individuals with autism spectrum disorder: A systematic review. Journal of Autism and Developmental Disorders, 49(2), 582–616. https://doi.org/10.1007/s10803-018-3724-5

Sellers, T. P., Valentino, A. L., & LeBlanc, L. A. (2016). Recommended practices for individual supervision of aspiring behavior analysts. Behavior Analysis in Practice, 9(4), 274–286. https://doi.org/10.1007/s40617-016-0110-7