Enhancing frequency recording by developmental disabilities treatment staff.

Shabani et al. (2006) tested a short staff-training package in a residential facility. Eight direct-care staff first got a one-hour class on how to count client behaviors. Then a supervisor showed up, watched, and gave quick verbal feedback after each shift.

The researchers tracked how close the staff counts were to gold-standard observations. They ran the study across three cottages using a multiple-baseline design.

Accuracy jumped for every staff member as soon as supervisor feedback started. The gains held even after feedback stopped and carried over to times when no supervisor was in the room.

In plain words, ten minutes of feedback turned shaky tallies into trustworthy data.

The result lines up with older residential studies. Alsop et al. (1995) showed that supervisors themselves need training to give good feedback, and Bickel et al. (1984) proved that monthly feedback can keep skills alive. The 2006 paper adds: even lighter feedback keeps data accurate.

Brown et al. (2021) later moved the same idea into ABA clinics. They swapped frequency counts for session notes and still saw big gains, showing the package travels across settings.

Fuesy et al. (2025) sounds like a contradiction at first. Their staff only behaved well when an observer was visible, while B et al. found generalization to unsupervised periods. The gap is in the procedure: Fuesy used self-monitoring alone, but B et al. paired supervisor feedback with clear written criteria. The extra clarity plus brief post-shift reviews let staff keep scoring accurately even when no one was watching.

You can copy this package tomorrow. After any in-service, stay five extra minutes, check two recordings, and tell staff what they got right. The study shows this micro-feedback is enough to lock in accurate data for months. Reliable counts mean better clinical decisions and fewer wasted sessions chasing phantom trends.