Three male children with behavior-disorders were treated by tactile, vestibular, and proprioceptive sensory stimulation in 20-minute sessions 3 times per week for a period ranging from 4-to-6 weeks. Attention to a mathematical computation task was measured for several sessions before the introduction of treatment and immediately after each treatment session. Attention was operationally defined as the total time spent attending to the task during a ten minute test period.

Of the three subjects, one showed negligible change, one showed some performance improvement and a reduction in response variability, and the third showed improvement followed by performance deterioration, though not to original low baseline levels.

Although improvements were coincident with the onset of treatment, they may not be attributed to it, because no response stability was achieved during the baseline condition.