The long-term effect of high- and low-rate responding histories on fixed-interval responding in rats.
After many sessions on the same fixed-interval schedule, past high- or low-rate histories no longer change how rats respond.
01Research in Context
What this study did
Ley (2001) asked a simple question. Do old reinforcement schedules still sway rats after months on a new schedule?
First, rats learned either a high-rate schedule or a low-rate schedule. Then all rats switched to the same 30-second fixed-interval schedule for 80-100 sessions. The team watched whether the old history still shaped response speed or pause patterns.
What they found
After the long run on FI 30-s, the two groups looked identical. High-rate history rats no longer fired off responses faster. Low-rate history rats no longer waited longer to respond.
Old learning had washed out. Current contingencies, not past ones, now controlled the lever.
How this fits with other research
Okouchi (2003) seems to disagree. In that study, rats kept their old pace when the prior inter-reinforcement interval matched the new FI. The difference: Hiroto stopped after only a few FI sessions, while Ley (2001) tracked the change across many more. Short run, history wins. Long run, it fades.
Hirai et al. (2011) extends the idea to humans. College students with FR histories responded faster on FI than students with DRL histories. The effect faded with long exposure but bounced back when the schedule flipped. The rat data in Ley (2001) warn us that, with enough trials, even humans may lose the history cue.
Davison et al. (1991) is the earlier rat study that first showed history matters. Ley (2001) acts as a successor, adding the timeline: history matters early, then disappears.
Why it matters
If you keep the same schedule for weeks, past reinforcement rates probably won't skew your data. For brief probes or quick reversals, though, watch for carry-over. Start a new client on a fresh baseline, run enough sessions, and let the current contingencies do the talking.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Run at least ten stable sessions before you trust that 'history is washed out' in your baseline.
02At a glance
03Original abstract
Tell rats were given extended lever-press training on a fixed-interval (FI) 30-s food reinforcement schedule from the outset or following exposure to one or two previous reinforcement schedules. For 4 rats the previots schedule was either fixed-ratio 20, which generated high response rates, or differential-reinforcement-of-low-rate 20 s, which produced low response rates. For 4 additional rats the extended training on FI 30 s was preceded by experience with two schedules: fixed-ratio 20 followed by differential-reinforcement-of-low-rate 20 s; or the same two schedules in the reverse order. Fixed-interval response rates were initially affected by the immediately preceding schedule, but after 80 to 100 sessions, all traces of prior schedule history had disappeared. The results also showed no long-term effect of schedule history on the interfood-interval patterns of responding on the FI 30-s schedule. These results support one of the most central tenets of the experimental analysis of behavior: control by the immediate consequences of behavior.
Journal of the experimental analysis of behavior, 2001 · doi:10.1901/jeab.2001.75-43