Within‐ and across‐session increases in work requirement do not produce similar response output

Leon et al. (2021) asked a simple question: if you keep raising the work price inside one visit, do you get the same demand curve as when you raise the price across many visits?

They used progressive-ratio (PR) schedules. In one setup the ratio climbed every few minutes within the same session. In the other setup the ratio climbed only at the next daily session.

They watched how many responses each setup produced and whether problem behavior popped up.

The within-session curve looked nothing like the across-session curve. It dipped sooner and shot problem behavior up.

In plain words, cramming all the price hikes into one meeting made the task look harder and the learner look more frustrated.

Thompson et al. (1986) once showed that within-session reversals can give clean preference data without extra days of baseline. Leon’s team found the opposite: within-session PR warped the data. The difference is the task. Quick choice reversals are brief; climbing PR schedules keep adding work.

Richling et al. (2019) warned that the old "80 % for three sessions" mastery rule can fail. Leon adds another warning: the way we test motivation can also mislead if we squeeze it into one sitting.

Benvenuti et al. (2024) review how schedules shape social give-and-take. Their PR logic assumes the curve means the same no matter how you build it. Leon shows that assumption is shaky when the whole climb happens fast.

If you run a demand assessment in one long burst, you may think a learner is done after 20 responses when they would really work for 200 if the price rose across days. Switch to across-session PR steps, or at least split the hike over two visits, before you write "low motivation" in the report. Your next session will start with clearer data and fewer problem behaviors.