A preliminary examination of social preference across assessors

The team ran paired-stimulus social preference assessments with one adult who has autism. They compared what he chose when his mom ran the test versus two unfamiliar adults.

Each adult showed the client pairs of social options: high-five, brief chat, selfie, or nothing. The client picked one. They repeated this until each adult had a clear winner.

The client picked different social reinforcers for each assessor. With mom he wanted selfies. With strangers he wanted high-fives. All top choices beat the no-interaction control.

This shows the same person can value different social contact depending on who is giving it.

Duker et al. (1996) proved that high-preference items from choice tests work as reinforcers. Huntington et al. extend that rule to social reinforcers and add a twist: the assessor changes what is "high-preference."

Matson et al. (2013) found unfamiliar toys can become top picks for preschoolers with autism. The new study flips the unfamiliarity angle: unfamiliar adults changed the adult client’s social picks, showing the paired-stimulus format is sensitive to who is in the chair.

Schulingkamp et al. (2023) saw no effect of partner familiarity on rats’ social demand. The human data look opposite, but the methods differ. Rats worked for any rat; the human chose different actions for different people. Species and task differences likely explain the gap.

If you run a social preference assessment, rotate staff and family members through the trials. The reinforcer that wins with you might bomb with mom. Test each key person, then use that person’s top social pick during their own sessions. This small step can save you from "reinforcer failure" and speed up skill acquisition.