Predicting the relative efficacy of three presentation methods for assessing preferences of persons with developmental disabilities.

Conyers et al. (2002) asked which way works best to find what people with developmental disabilities like. They tried three ways: showing real items, showing pictures, or saying the item names.

First they gave a quick test to see if the person could match pictures or follow spoken words. Then they ran a two-choice preference check with each method. They flipped the methods back and forth to be sure.

The quick test predicted success with 94% accuracy. People who passed conditional tasks picked the same top items no matter how you asked. People who failed those tasks only picked the same top items when you showed the real stuff.

If the learner could not match pictures or words, pictures and spoken names gave wrong answers. Real items gave the right answers every time.

Morris et al. (2020) extends this idea to autism. They found short looping videos work better than still pictures for social reinforcers. Both studies say: test the format first.

Madden et al. (2003) used the same two-picture setup and showed that naming the pictures out loud helps people pass delayed matching. This backs the idea that verbal skill drives picture success.

van der Meer et al. (2012) also tested modality preference, but in AAC. They showed kids learn faster when you let them pick speech devices or signs. Same rule: assess first, then teach.

You can save session time and avoid false negatives. Run a five-minute probe: present two pictures and ask "Give me the ___," then try two spoken names. If the learner fails both, skip picture or spoken preference checks and place real items in front of them. You will get truer high-preference reinforcers and spend less time on useless trials.