Flash rate discrimination in rats: rate bisection and generalization peak shift.
Flash rate is a reliable, scalable visual dimension for teaching conditional discrimination and studying peak shift.
01Research in Context
What this study did
Researchers taught rats to tell two flashing-light speeds apart. The rats pressed one lever for a slow flash and another lever for a fast flash.
After the rats learned the rule, the team tested new flash speeds the animals had never seen. They wanted to see how the rats would sort these in-between speeds.
What they found
The rats split new speeds near the geometric middle of the trained pair. This 'bisection' shows they treated flash rate like a ruler with numbers.
The rats also showed peak shift. After training, they avoided speeds close to the non-reinforced cue and peaked their responding away from it, just like pigeons do with color or tone.
How this fits with other research
Derenne (2010) got the same peak shift with human students judging face symmetry. The leap from rat flash rate to human face preference shows the rule works across species and stimuli.
Pear et al. (1971) mapped peaked curves for light brightness. Both studies prove you need a wide test range to see the true hill-shaped gradient; narrow testing hides it.
Brown et al. (2025) found low-prevalence flashes hurt rat accuracy. Flapper et al. (2013) kept all test speeds frequent, so their rats looked 'smarter'—a reminder that how often you show a target can swamp what the learner actually knows.
Why it matters
If you need a fine-grained visual cue for conditional discrimination, flash rate is easy to program and gives clean data. Use many test values and keep each one common enough to avoid prevalence effects. The same rules that shape rat lever presses can shape human visual choices—handy when you design generalization probes for clients who rely on visual cues.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Program your visual discrimination task with at least five flash speeds and equal exposure to each to get a clean gradient.
02At a glance
03Original abstract
Two experiments were conducted to determine whether responding by albino rats can be brought under the stimulus control of different flash rates. In the first experiment, a conditional discrimination procedure was employed whereby two different flash rates (fast or slow) signaled the availability of reinforcement on one of two levers (left or right). Stimulus control emerged rapidly and improved with continued training. When intermediate flash rates were presented during probe sessions, the bisection point of the fast and slow flash rates was near their geometric mean, consistent with research employing other stimulus types. In the second experiment, a successive discrimination procedure was employed whereby responding in the presence of one flash rate (S(+) ) was reinforced while responding in the presence of another flash rate (S(-) ) was not reinforced. Again, stimulus control emerged quickly and improved with continued training. Test sessions in which many different flash rates were presented for brief periods in extinction revealed the peak shift phenomenon, in which peak response rates are shifted from the S(+) in a direction away from the S(-) . Flash rate is endorsed as a continuous stimulus dimension that is useful for differentially signaling schedule components.
Journal of the experimental analysis of behavior, 2013 · doi:10.1002/jeab.36