A comparison of measures of responding under fixed-interval schedules.
Stick to post-reinforcement pause and running rate—they cleanly capture every fixed-interval change without extra math.
01Research in Context
What this study did
The team ran fixed-interval schedules with pigeons. They wanted to know which numbers best show how behavior changes across the interval.
They compared five common measures. The birds pecked a key for food on a 3-minute FI. Data came from steady sessions after training.
What they found
Post-reinforcement pause and running rate won. They were steady, easy to count, and tracked every shift in responding.
Other numbers either jumped around or took too long to figure out. Pause and run gave a clear story with simple math.
How this fits with other research
Schwarz et al. (1970) had already shown that the sixth response after food sets the new pace. Wilson et al. (1973) now say you only need two scores to see that shift: pause and run.
Mahoney et al. (1971) proved pause grows like a ruler when the next food time changes. The 1973 paper keeps pause as a top tool but inside plain FI schedules, not cycling ones.
Austin et al. (2015) later found that letting the bird start its own interval muddies timing. Their mixed results still used pause and run, showing the 1973 pair travels well even when the procedure changes.
Why it matters
When you graph FI data, plot pause on one axis and run rate on the other. You will spot trends faster and explain them to staff or parents without fancy stats. These two measures save time and keep everyone looking at the same clear picture.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Start timing the first pause and the run rate after it for any FI program; graph both each session.
02At a glance
03Original abstract
Average response rate, post-reinforcement pause, elapsed time to the fourth response, average quarter-life, and running rate were examined to see how they reflected changes in fixed-interval performance. Rats were exposed to a mixed schedule of water presentation comprising fixed-interval schedules of two durations. Changes in responding were produced by varying the duration of the shorter component. The five measures were derived only from the longer schedule component. Post-reinforcement pause, elapsed time to the fourth response in the interval, and quarter-life all showed high, positive inter-correlations (0.78<r<0.99). Running rate and post-reinforcement pause were not as highly correlated. Quarter-life reliably reflected changes in fixed-interval performance but changes in the quarter-life value did not necessarily result from similar changes in fixed-interval response pattern. The two measures that adequately described changes in response patterning were post-reinforcement pause and running rate. These two measures also had the advantage of being simple both computationally and in terms of the instrumentation involved in their recording.
Journal of the experimental analysis of behavior, 1973 · doi:10.1901/jeab.1973.20-281