An Evaluation of a Staff Management Strategy to Minimize Reactivity in Procedural Fidelity of Intervention Implementers

Reyes et al. (2025) tested a simple twist on feedback. They watched staff run social-skills sessions, but the staff never knew when the check happened. After the silent visit, the supervisor gave feedback. The study used a multiple-baseline design across three staff members.

Every staff member hit higher fidelity for positive social engagements after the surprise feedback. The gains showed up only after the feedback started, not before. The team kept the new level even when no one was watching.

Akers et al. (2024) ran a similar test with BCBA trainees, but the feedback came through Zoom after remote sessions. Both studies show the same core idea: feedback works best when staff think no one is watching.

Ruppel et al. (2023) added a quick Zoom feedback call after standard preference-assessment training. Their results match Reyes et al. (2025): a short feedback session fixes small errors and keeps fidelity high.

Shapiro et al. (2016) used a self-instruction manual plus brief feedback for preference assessments. The older paper and the new one both find that feedback, not more training, is the key final step.

You can stop doing daily in-room coaching. Slip in unannounced, record a five-minute probe, and give feedback later. Staff learn you might check at any time, so they stay sharp even when the door is closed. One quiet visit per week can protect treatment integrity without extra pay or prep.