‘The Habitual Use of the Self-report Habit Index’: A Reply

Benjamin Gardner*, Charles Abraham, Phillippa Lally, Gert Jan De Bruijn

*Corresponding author for this work

Research output: Contribution to journalComment/debatepeer-review

14 Citations (Scopus)


‘Habit’ is an abstract psychological construct, and its conceptualisation and measurement are open to debate. Sniehotta and Presseau [1] posit that habit is in essence a form of automaticity. They argue that the Self-Report Habit Index (SRHI) [2] is problematic because it assesses the central characteristic of habit (i.e. automaticity) together with its antecedent (repetition) and a possible consequence of automatic action (assimilation of the action into the self-concept). We agree that self-identity is not a necessary component of habit. Acquisition through repetition, however, distinguishes habit from other forms of automaticity, such as behaviour prompted by implementation intentions and unconditioned ideomotor reflexes, neither of which require prior performance. Where such a distinction is of interest, performance frequency indicators within the SRHI may be needed to demarcate habit from other automaticity subtypes.

Automaticity may be considered the ‘active ingredient’ of habit; it is because they are automatically activated that habits are performed frequently and override motivational tendencies in associated contexts [3]. Habit formation studies have thus focused on automaticity development [4]. Repetition items within the SRHI may conflate true effects of habit-related automaticity on behaviour because repeated performance is not necessarily indicative of habit [5]. As Ajzen [5] concluded, ‘a more appropriate approach would rely on an operationalization of habit that is independent of the behavior it is supposed to explain and predict’ (p114). An automaticity-specific SRHI subscale may be more appropriate where estimating the impact of habit on behaviour or tracking habit formation, and several authors have attempted to exclude SRHI frequency items for these purposes. To our knowledge, the SRHI has not been subject to content validity assessment, and so it is difficult to reliably identify which items best capture automaticity.

Most SRHI applications to date have been based on correlational data [1, 3]. Such studies are useful for identifying priorities for further investigation using more rigorous research designs, and our synthesis of correlation-based work provides considerable evidence of the relevance of habit to health behaviour [3]. Progress in habit theory and application will require more robust methodological approaches. Recent work has explored idiographic experiences of habit formation [4], and a pilot-controlled trial of a weight loss intervention based on habit formation principles yielded positive results [6]. Habit decay and disruption processes also demand research attention. Progress in habit intervention design will require identification of the behaviour change techniques necessary to make or break habits. Further empirical work is needed into the potential moderating effects of related habits on information-based health promotion campaigns predicted by theory [7]. Such work will require a shift away from observational surveys towards experimental manipulations, using both group-level assessments and N-of-1 designs to capture important within-person variation in experiences of habitual regulation.

The SRHI has proven useful for documenting habitual action [3]. The research agenda we have outlined will, however, require adoption of sophisticated techniques to address pertinent issues for theory and application and, in some settings, a more precise measure of the automaticity that distinguishes habits from other forms of repeated behaviours.
Original languageEnglish
Pages (from-to)141-142
Number of pages2
JournalAnnals of Behavioral Medicine
Issue number1
Publication statusPublished - Feb 2012


Dive into the research topics of '‘The Habitual Use of the Self-report Habit Index’: A Reply'. Together they form a unique fingerprint.

Cite this