Habit Automaticity

What it is

Two claims sit inside the term and they are usually run together. The first is definitional: that a mature habit is a learned cue-response link which can fire without a decision being renewed. The second is empirical: that people performing habits are not attending to what they are doing. The first is Wood's construct, stated in her own words, and the vault cites it as that. The second is a measurement, and it came out weaker than its reputation.

In Wood, Quinn and Kashy's 2002 diary studies, thought correspondence was a dependent variable rather than an entry requirement. Two coders, agreeing on eighty-four per cent of the first study's behaviours and eighty-nine per cent of the second's, classified whether each person's reported thought matched the behaviour they were performing. For habitual behaviour in the larger study, thoughts failed to correspond fifty-one per cent of the time against forty-four per cent corresponding. The authors' summary is that thoughts wandered from the behaviour about half to sixty per cent of the time, and the sentence nobody quotes is that the finding that participants thought about habits about forty per cent of the time suggests the habitual mode of behaviour regulation is best characterised by minimal or sporadic cognitive monitoring and not by the complete absence of thought.

Effortlessness, in Wood's account, is constitutive rather than incidental: the easy, fluid, automatic feeling of acting on habit is a defining property rather than a side effect. Her working distinction between routine and habit is the clearest statement of what the construct adds over things people do a lot. Routine is repetition in a stable context. Habit is routine plus automaticity. That is also precisely why a measurement based on repetition and location alone, which is what the diary studies used, does not reach the construct it is taken to measure.

The best evidence for the definition is a dissociation from Wood's own laboratory, and it is also the best argument against trusting a questionnaire to find it. David Neal, then at the University of Southern California, with Wood, Jennifer Labrecque and Phillippa Lally, ran a subliminal lexical-decision study on runners in 2012. Context primes sped recognition of running words for strong-habit participants, with an interaction at F(1, 53) = 4.10 and a partial eta-squared of .072, while habit strength was unrelated to recognition after goal primes. In the same participants, those with stronger habits reported that their running was goal-driven at r(51) = .54, and that belief was uncorrelated with the implicit measure at r(51) = .07. Belief about automaticity and measured automaticity came apart in the same people. The authors' own conclusion is that research on the actual mechanisms of habit performance cannot rely on judgments and inferences as tests of underlying mechanisms.

The construct is also disputed in print, and the dispute is specific. Jan De Houwer and colleagues at Ghent University and KU Leuven argued in 2023, in Perspectives on Psychological Science, that goal-directed explanations of habitual behaviour have been dismissed too easily, and that the core question is the extent to which habitual behaviour is mediated by stimulus-response associations or by goal representations. Separately, de Wit and colleagues reported in 2018 five failures to produce inflexible habits in humans by overtraining, using variants of four previously published outcome-devaluation paradigms. The laboratory paradigm for making a habit on demand does not cooperate, which is a different problem from the field measurement one and points the same way.

In effect

The runners who were wrong about themselves

The 2012 priming study is the cleanest research case because it puts the two measurements side by side in one sample. Participants with strong running habits showed the implicit context-cue effect and simultaneously reported that their running was driven by goals, and the two measures did not correlate. Neither measure is obviously the truth. What the pair establishes is that a person's account of why they run is not a reading of the mechanism that gets them out of the door, which is exactly what a construct defined by inattention would predict and exactly what makes it hard to measure by asking.

Driving home and arriving without the journey

The everyday case is the one everybody already has: the route taken without decisions, the arrival with no recollection of the middle. On Wood's definition this is a cue-response sequence running off without renewed intention. On the diary paper's evidence, though, the accompanying mental state is not blankness. About forty per cent of the time, people performing habits were thinking about the thing they were doing. The honest description is sporadic monitoring, a hand on the wheel rather than nobody at the wheel.

An app that counts streaks and calls it automaticity

Habit-tracking products measure completions and call the resulting number habit strength. Completions are repetition, which is one of Wood's two ingredients and not the other. A product could in principle instrument the second ingredient, by measuring reaction time to a cue, or latency to begin after the cue appears, or whether the behaviour survives the reward being withdrawn, which is the test Wood calls reward insensitivity. Almost none do, because repetition is cheap to count and automaticity is not. The asymmetry is the same one that produced the field's measurement problem, reproduced in commerce.

What it does not say

It does not say habits are performed with no thought at all. The paper most often cited for that rejects the reading in its own discussion.

It does not say habit and automaticity are the same thing. Wood names reflex, Pavlovian conditioning, automatic goal activation and snap social judgement as separate members of the family, and notes they entangle without being one mechanism. An argument that runs on automatic behaviour in general is not an argument about habit.

It does not settle the mechanism. Whether habitual behaviour is mediated by stimulus-response associations or by goal representations is an open question actively argued in the primary literature, and the laboratory induction paradigm has five published failures behind it.

It is not measured by the instruments most often used to measure it. A frequency-by-stability product, or a single question about past frequency, measures repetition. Treating it as a measure of automaticity is an inference the field's own meta-analysis gives reason to doubt.

Who developed it

The definition and the taxonomy are Wendy Wood's, stated in Good Habits, Bad Habits in 2019 and in her research programme across thirty years. The dissociation evidence behind it is David Neal's first-authored 2012 paper, which popular retellings credit to Wood. The vault credits the first author in both cases.

---