EV-0050
How long it takes, and what a missed day costs
Lally, P., van Jaarsveld, C. H. M., Potts, H. W. W., & Wardle, J. (2010). European Journal of Social Psychology, 40(6), 998-1009 · Source
What this entitles us to say
A 12-week study of habit formation in daily life finding that automaticity builds along an asymptotic curve, that time to near-automaticity ranged from 18 to 254 days across individuals, and - the finding that matters most here - that missing one opportunity to perform the behavior did not materially affect the process.
What it found
96 volunteers enrolled, 82 with sufficient data. Each chose a daily eating, drinking or activity behavior to perform in a consistent context for 12 weeks, recording automaticity daily.
- Automaticity built along an asymptotic curve - large early gains, diminishing with each repetition, then a plateau. - Time to 95% automaticity ranged from 18 to 254 days across individuals. - Missing one opportunity to perform the behavior did not materially affect the habit formation process.
Why it matters here
A16 has to answer the participant's real question, which is not "does change work" but "how long, and what happens when I slip?"
On duration: the honest answer is it varies enormously, and saying so is better than a number. A leader told "66 days" who is still struggling at day 90 concludes they have failed. A leader told "somewhere between three weeks and most of a year, and it gets easier fastest at the start" has been told the truth and has a reason to continue.
On relapse: this is the finding the module needs most, and it was named as a research target before the search began. A single missed day did not materially affect the process. That directly contradicts the all-or-nothing story people tell themselves, and it is the difference between one bad week ending an attempt and merely interrupting it.
On shape: the asymptotic curve is worth teaching literally. The early repetitions do the most work, which means the hardest part is the beginning and the effort required falls. That is an encouraging and accurate thing to tell someone starting a practice, and it fits the Discipline finding that self-control operates through automatic behavior rather than deliberate restraint (EV-0027).
The limits
82 participants, simple behaviors - eating a piece of fruit with lunch, going for a walk after dinner - not character traits, and the transfer is by analogy. Only about half the participants reached automaticity at all, so the widely quoted median is a median among those who succeeded rather than a typical outcome. Self-reported automaticity. The 66-day figure should be handled with care precisely because it is so widely repeated: the honest version is the range, not the average. Verified to abstract level, with the lapse finding confirmed against the published abstract.
What we do not say
"It takes 66 days to form a habit." That number is a median among the roughly half who succeeded, on simple daily behaviors, and it is not what the paper claims. Repeating it would be exactly the failure the Attitude file exists to guard against.