Habit formation

What this subject is

Habits are context-cued, largely automatic responses acquired through repetition with reinforcement. The modern field spans animal-learning theory (goal-directed versus habitual control), social and health psychology (real-world formation, measurement, interventions), and cognitive neuroscience (corticostriatal loops). Its practical core: behaviour change succeeds by designing contexts, cues and friction — not by willpower — and formation takes weeks to months with large individual variance, not 21 days.

What the package holds

Curated scaffolding your agent loads before it researches, so it starts from vetted ground rather than a cold search.

• Tier 1: 13
• Tier 2: 8
• Tier 3: 2
• Tier 4: 3

The questions and claims below are quoted from the package files.

Where the field disagrees

Each one carries real proponents on more than one side, so your agent cannot quietly pick a winner.

• Is habit automaticity mediated by activated goals, or do contexts cue responses directly?

  2 named positions · direct cuing is the dominant account for strong habits; hybrid and hierarchical models are actively developed, and moderate habits fit neither camp cleanly

• Do laboratory habit paradigms capture what "habit" means in daily life?

  3 named positions · open methods problem; lab and field may be measuring only partially overlapping constructs — a package that cites both owes the learner this caveat

• Does the Self-Report Habit Index measure automaticity, or remembered frequency plus identity?

  2 named positions · live; current studies increasingly use the SRBAI precisely because of this debate

• How large and durable are implementation-intention effects outside the lab?

  0 named positions · robust on average with real boundary conditions; a tool, not a magic bullet

• How much of the bestselling habit literature is supported by the evidence it cites?

  3 named positions · tier 4 material works as on-ramp and drill source; the package teaches the journal evidence and uses the bestsellers as framing, never as authority (protocol/research.md)

Myths the package corrects

Widely held claims with the evidence that settles or bounds them.

• “It takes 21 days to form a habit.”

  debunked

  The direct real-world modelling study found a median of 66 days to reach the automaticity plateau, with fitted individual estimates from 18 to 254 days depending on person and behaviour complexity. A 2024 systematic review replicates the pattern at scale: about two months is typical, and the spread is enormous. No dataset supports a 21-day threshold.

• “Science proved it takes 66 days to form a habit.”

  boundary-correction

  66 days was the median of one 96-person sample, not a constant. Drinking a daily glass of water approached its plateau within weeks, exercise behaviours took months, and some fitted curves imply most of a year. The finding is a wide distribution shaped by behaviour complexity and the person — not a deadline.

• “Miss a day and you are back at square one — the streak is the habit.”

  debunked

  In the Lally data, isolated missed opportunities produced no material loss of modelled automaticity; the curve is asymptotic, not a ladder that collapses. What predicts failure is abandoning context-consistent repetition altogether, not a single gap.

• “Lasting behaviour change is mainly a matter of motivation and willpower.”

  debunked-as-general-account

  Experimentally induced medium-to-large changes in intention (d = .66) produce only small-to-medium changes in behaviour (d = .36), and roughly 43 percent of everyday action repeats in stable contexts while thought wanders elsewhere. Context design and if-then planning outperform resolve because they remove the need for in-the-moment willpower.

• “Every habit runs on a craving for its reward — find the reward and you understand the habit.”

  boundary-correction

  Reward learning builds the association — and immediate rewards build it faster than delayed ones — but once established, strong habits run on context cues even when the outcome is devalued: habitual cinemagoers ate stale week-old popcorn at fresh-popcorn rates, and outcome devaluation has been the lab criterion for habit since the 1980s. Craving explains acquisition and relapse pressure, not everyday execution.