A brief computer-based assessment of reinforcer dimensions affecting choice.

Cullinan et al. (2001) built a 10-minute computer game. Students clicked between two boxes that gave points.

The team changed four things: how fast points came, how good the prize was, how long the wait, and how hard the click. They watched which box each student chose most.

Students spread their time differently when any dimension changed. Some chased fast points, others chased big prizes, others avoided hard clicks.

The quick task showed which dimension controlled each student’s choice.

Peck et al. (2024) later showed that if you let delay sneak in, effort looks more powerful than it is. Their new task keeps delay constant, so the numbers are cleaner. Cullinan et al. (2001) did not control delay, so their effort data may have been inflated.

Sievers et al. (2020) added another layer: longer access to toys makes kids pick them over candy. This extends the 2001 idea that magnitude, not just rate or quality, can flip the hierarchy.

Horrocks et al. (2008) used the same choice logic with sounds for teens with disabilities. Both studies say the same thing: a brief paired or concurrent choice predicts what will work as reinforcement.

You can copy the 10-minute screen task during intake. Watch which dimension the student chases, then set up your token board or break schedule to match. If Peck et al. (2024) is right, also check that you are not adding hidden delays that could fake the results. Quick, kid-friendly, and data you can use Monday.