A voice-detecting sensor and a scanning keyboard emulator to support word writing by two boys with extensive motor disabilities.

Two boys with severe motor impairments tested two ways to type words on a computer. One way used a voice-detecting sensor. The other used a pressure sensor. The boys switched between the two tools every session so the team could see which was faster and which they liked.

The voice sensor picked up any small sound the boy made. Each sound click moved a highlight across an on-screen keyboard. When the highlight reached the wanted letter, the boy made another sound to select it. The pressure sensor worked the same way, but the boy had to press a switch instead of make a sound.

One boy typed more letters per minute with the voice sensor. The other boy typed faster with the pressure sensor. So speed was a tie.

Both boys, and also student raters who watched them, strongly preferred the voice sensor. They said it felt easier and less tiring than pressing a switch.

van der Meer et al. (2012) ran a similar horse-race: speech-generating device versus manual signing. Like E et al., they saw kids learned requesting faster with the tool they liked best. The message across both studies: preference predicts success.

Meier et al. (2012) took the voice-sensor idea and moved it to adults with ALS. They swapped the voice click for an eye-tracking click. Text writing still worked, showing the sensor type can change while the concept holds.

Shih et al. (2009) hacked a cheap mouse into a motion sensor the same year. Both studies turned everyday hardware into body-friendly switches for people with almost no movement. The mouse team aimed for play; the voice team aimed for writing.

If you work with clients who can’t use their hands, try a voice sensor first. It costs little, sets up fast, and kids usually like it. Keep the old switch handy, though—some will still do better with it. Run a quick alternating-treatment probe for two or three days, let the client vote, and you have your answer.