Describing behavior with ratios of count and time.

Davison et al. (1989) wrote a how-to guide, not an experiment.

They listed every common way to turn counts and time into one number.

The paper shows when to pick rate, celeration, or bounce to describe behavior.

Each ratio answers a different question.

Rate tells you speed right now.

Celeration tells you if speed is speeding up or slowing down.

Bounce tells you how much the speed jumps around.

Rider (1977) did the same thing for reliability indices twelve years earlier.

Both papers say, "Pick the number that matches your question."

Wilder et al. (2023) later used the same rate idea on procedural-fidelity checks.

They showed percentages alone miss drift; adding rate catches it.

Newland (2024) pushed the idea further by turning the same raw data into risk ratios.

This gives you new stats and graphs while keeping the 1989 logic.

Stop using only one favorite metric.

Ask, "Do I need speed, trend, or variability?" then plug in the matching ratio.

Next time you graph fidelity or self-monitoring, add a rate line under the percentage.

Your data story becomes clearer, and reviewers smile.