Preference reversal between impulsive and self-control choice.
Choice between small-soon and large-late rewards can reverse inside one session as delays grow, so watch for the flip and be ready to adjust.
01Research in Context
What this study did
Emma and team worked with pigeons in a lab.
The birds chose between two keys. One key gave a small food reward right away. The other key gave a bigger reward after a longer wait.
During each daily session the scientists slowly made the wait times longer. They watched to see if the pigeons switched their choice.
What they found
Early in the session most birds picked the small-soon reward.
As the delays grew longer inside the same session the same birds flipped. They started picking the large-late reward.
This within-session flip shows that bigger rewards become more attractive when all delays get long.
How this fits with other research
Green et al. (2004) tested pigeons the same way but saw no flip. The difference: they kept each delay fixed across days. Emma moved the delay inside one sitting, revealing the hidden change.
Schmitt (1984) first showed that longer terminal-link delays shift choice. Emma’s work extends that idea by proving the shift can happen in minutes, not days.
Lloyd (2002) found that preference weakens as absolute delays grow even when the gap between choices stays the same. Emma’s reversals give a live picture of that weakening inside one session.
Why it matters
Your client may pick a tiny immediate reinforcer during short waits. When wait times stretch, the same client may suddenly work for a bigger later payoff. Watch for this flip during long therapy sessions or when schedules drift. You can use the moment of reversal to probe true preference and adjust token boards, break lengths, or reinforcement size on the fly.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →During your next long session gradually stretch the wait for the big reinforcer and note the minute the client switches preference—then reset the schedule to match that point.
02At a glance
03Original abstract
In a concurrent-chains procedure, pigeons chose between reinforcers varying in delay and amount. Reinforcer amount was determined by duration of access to grain, and delay was determined by fixed-interval schedules in the terminal links. Preference was measured by the ratio of responses in initial links. Dependent scheduling of variable-interval schedules in initial links ensured that delay and amount were not confounded with frequency of outcomes, which remained equal for the two choices. In Experiment 1, in components signaled by red keys in the initial links, small and large reinforcers were delivered after delays of 1 s and 10 s respectively. In components signaled by green, additional time was added to both delays. Smaller-sooner reinforcers were preferred in red components. In green components, smaller-sooner reinforcers were preferred at short delays, and choices for the larger-later reinforcer generally increased with increasing duration of the added delay. At longer delays, up to 15 s, the larger-later reinforcer was preferred. That is, the pigeons showed within-session preference reversal, with impulsive choice at short delays in red components and self-control choice at long delays in green components. In Experiment 2, added delay to both reinforcement and reinforcer amount were varied. Sensitivity of initial-link response ratios to ratios of amount increased with increasing duration of the added delay. This interaction between delay and amount was predicted if the temporal discounting functions assumed the magnitude effect in which discounting rate was inversely proportional to amount. It was also predicted by the contextual choice model of performance in concurrent-chains procedures.
Journal of the experimental analysis of behavior, 2013 · doi:10.1002/jeab.23