Measure the Return, Not the Slip

Most people measure behavioral change with a binary question: Did I do the old thing or not?

That metric is clean. It is also often too crude to show what is actually improving.

Suppose the old pattern once ran for three days before you recognized it. Then it ran for six hours. Then twenty minutes. Then you noticed it in the middle of the conversation. Eventually you recognized the first bodily cue before the behavior fully formed. A binary score records failure, failure, failure, failure, success. A learning model records progressively earlier detection and faster recovery.

Habit decay and behavior change are not uniform processes. Recent longitudinal work found large person-to-person differences in how self-reported habit strength changed over time (Edgren et al., 2025). Habit research also emphasizes context: automatic responses are often strongly tied to the situations in which they were learned (Bouton, 2024). That means progress may appear uneven because the same person can be highly practiced in one context and surprisingly vulnerable in another.

So measure more than occurrence.

Measure detection latency: How long before you knew what was happening? Measure interruption latency: Once you knew, how long before you changed course? Measure duration: How long did the old sequence continue? Measure recovery: How quickly did you return to an aligned response? Measure generalization: In how many different contexts can you access the new behavior? Measure effort: Does the new response still feel like lifting a car, or is it starting to feel available?

These measures do something psychologically important as well. They separate a momentary automated response from the story a person tells about that response. One slip can become “nothing changed,” and that conclusion can create more dysregulation than the original behavior.

Progress is not permission to ignore harm or excuse repeated behavior. It is permission to measure change accurately.

The return tells you what the system has learned. Pay attention to it.

References

Edgren, R., Baretta, D., & Inauen, J. (2025). The temporal trajectories of habit decay in daily life: An intensive longitudinal study on four health-risk behaviors. Applied Psychology: Health and Well-Being, 17(1), e12612. https://doi.org/10.1111/aphw.12612

Bouton, M. E. (2024). Habit and persistence. Journal of the Experimental Analysis of Behavior, 121(1), 88-96. https://doi.org/10.1002/jeab.894

Buabang, E. K., Donegan, K. R., Rafei, P., & Gillan, C. M. (2025). Leveraging cognitive neuroscience for making and breaking real-world habits. Trends in Cognitive Sciences, 29(1), 41-59. https://doi.org/10.1016/j.tics.2024.10.006

Labrecque, J. S., Lee, K. M., & Wood, W. (2024). Measuring context-response associations that drive habits. Journal of the Experimental Analysis of Behavior, 121(1), 62-73. https://doi.org/10.1002/jeab.893