Related: Common failure modes in habit formation

I ran across this bit of pop-sci (a review of Jeremy Dean's Making Habits, Breaking Habits), which claims that habits typically take around 66 days to form, not the 21 days that self-help articles tend to cite. The somewhat surprising thing to me, on reflection, was how readily I'd taken the 21-day statistic as fact. From the article:

When he became interested in how long it takes for us to form or change a habit, psychologist Jeremy Dean found himself bombarded with the same magic answer from popular psychology websites and advice columns: 21 days. And yet, strangely — or perhaps predictably, for the internet — this one-size-fits-all number was being applied to everything from starting a running regimen to keeping a diary, but wasn’t backed by any concrete data.

The original article is here. Abstract:

To investigate the process of habit formation in everyday life, 96 volunteers chose an eating, drinking or activity behaviour to carry out daily in the same context (for example ‘after breakfast’) for 12 weeks. They completed the self-report habit index (SRHI) each day and recorded whether they carried out the behaviour. The majority (82) of participants provided sufficient data for analysis, and increases in automaticity (calculated with a sub-set of SRHI items) were examined over the study period. Nonlinear regressions fitted an asymptotic curve to each individual's automaticity scores over the 84 days. The model fitted for 62 individuals, of whom 39 showed a good fit. Performing the behaviour more consistently was associated with better model fit. The time it took participants to reach 95% of their asymptote of automaticity ranged from 18 to 254 days; indicating considerable variation in how long it takes people to reach their limit of automaticity and highlighting that it can take a very long time. Missing one opportunity to perform the behaviour did not materially affect the habit formation process. With repetition of a behaviour in a consistent context, automaticity increases following an asymptotic curve which can be modelled at the individual level. [My emphasis.]

My comments:

• There is an observed “automaticity plateau.” Can individuals influence the height of the plateau through interventions such as rewards? Would this change the exponential rate constant? Or do we have less control over these things than we think?
• 95% of maximum automaticity doesn't quite seem like the right metric to use to describe habit formation, especially if the maximum is on the low side.
• Presumably you'd need familiarity with the SRHI survey to answer this, but it's not clear to me what an automaticity score of 40 really means. (Examples or a baseline might help: what's my automaticity for toothbrushing? checking email?)
• N=96 seems small. It seems slightly problematic that the 14 participants who dropped out were not included in the analysis, and rather problematic that they used a 3-parameter model and only got a ‘good fit’ for half of the participants. (I'm not an expert in this, so I'd appreciate knowing if my intuitions here are right.)
• It seems that changing habits is harder than I'd previously thought, at least in the absence of CFAR-like techniques. (Which we still don't know if it works, as far as I know. I'm looking forward to their research.)