Discrimination training for persons with developmental disabilities: a comparison of the task demonstration model and the standard prompting hierarchy.
Use task demonstration—full model plus fading and mixed examples—to teach discriminations; it halves errors and doubles correct responses.
01Research in Context
What this study did
The team compared two ways to teach new discriminations to adults with developmental disabilities.
One group got the usual least-to-most prompts. The other group got the task demonstration model.
That model shows the full task first, then slowly fades cues while mixing many examples.
What they found
Task demonstration doubled unprompted correct answers and cut errors by forty percent.
The edge held in twenty-nine of thirty-two checks, including tests with brand-new items.
How this fits with other research
Peterson et al. (2019) looked at varied versus rote modeling during DTI. They saw brief sparks of flexibility but slower learning—supporting the 1990 message that one clear model beats many wobbly ones.
Waddington et al. (2021) later used prompt-delay fading with an adult with aphasia. Their positive results show the fading logic works beyond kids and across diagnoses.
Neuringer (1973) warned that presence-absence training gives flat stimulus control. The 1990 study answers that warning with a fix: teach along one dimension and fade within it, just as the pigeons learned better with fast-versus-slow clicks.
Why it matters
If you run DTI or teach daily-living tasks, swap your least-to-most prompt stack for task demonstration. Start with a full model, fade cues bit by bit, and mix many examples early. You should see fewer errors and stronger use of the skill with new materials.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Film yourself doing the full task, show the clip first, then fade one cue each trial while rotating three exemplars.
02At a glance
03Original abstract
A comparison was made between two procedures for teaching persons with severe handicaps: (a) the task demonstration model, which is based upon a fading procedure and general case programming, and (b) the standard prompting hierarchy, a least-to-most intrusive prompting procedure commonly used to teach these individuals. Five phases were used in comparing the procedures: pretesting, training, two generalization tests, and a 6-month maintenance test. Eight students learned two discrimination tasks by each procedure, with each task involving two- or three-digit numbers. Results showed that under the task demonstration model all 8 subjects had more unprompted correct responses (about 1.5 times as many) in training, all 8 subjects had fewer errors (about 0.6 times as many) in training, all 8 subjects had more correct responding in the generalization test with untrained stimuli in the training room, 6 of 8 subjects had more correct responding with untrained stimuli in another room, a 7th had equivalent amounts, and 7 of 8 subjects had more correct responding on a 6-month maintenance test. Thus, the task demonstration model proved superior to the standard prompting hierarchy in 29 of 32 tests of correct responding. Results are discussed in terms of implications for stimulus control training strategies.
Journal of applied behavior analysis, 1990 · doi:10.1901/jaba.1990.23-43