Validating predicted activity preferences of individuals with severe disabilities.
Staff can guess client preferences pretty well, but only when likes are far apart.
01Research in Context
What this study did
Staff guessed which activities adults with severe disabilities would pick.
Researchers then let clients choose for real and scored hits and misses.
The goal was to see how well everyday caregivers know client likes without formal tests.
What they found
Staff got it right about three times out of four.
Accuracy jumped when the liked item was much stronger than the other choice.
When preferences were close, predictions were little better than chance.
How this fits with other research
Lancioni et al. (2006) extends this work by showing that just asking clients what they like is not enough.
Adding real items and letting clients touch them boosts prediction power.
Rodriguez et al. (2024) refines the idea further: they removed bias in machine-style choice tests by giving kids short forced trials first.
Together the three studies say staff hunches help, but you still need to put the item in the client’s hand before you trust the result.
Why it matters
You can save session time by using staff guesses to pick the first toy, then run a quick tangible test to be sure.
When choices look equally liked, skip the guess and go straight to a brief paired-stimulus assessment.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Pick two toys staff say the client loves and two they are unsure about, then run a 2-min tangible paired test to check.
02At a glance
03Original abstract
We examined the accuracy of 24 staff members' predictions of activities preferred by 14 individuals with severe disabilities. For each of 144 activities, staff members assigned a client preference rating of "likes a lot," "likes," or "dislikes." Two activities from each category were randomly selected for each individual with disabilities. Pairs of selected activities were presented to the individuals, who were prompted to choose an activity. Staff members' activity preference ratings correctly predicted the choices made by the individuals with disabilities for 78% of the trials. The more divergent the preference ratings of the paired activities, the more likely staff members were to predict correctly the activity selected by a participant.
Journal of applied behavior analysis, 1993 · doi:10.1901/jaba.1993.26-239