Implementation intentions are specific strategies for which of the following?

Implementation intentions (Gollwitzer, 1993) are a self-regulatory strategy that involves an individual planning, in advance, the situation in which he/she will perform a particular behaviour. When applied as a behavioural intervention, participants are often required to decide when (e.g., time of day, day of the week) and where (e.g., bedroom) they will act. These manipulations have been effective in promoting desirable behaviours such as increasing fruit and vegetable intake (e.g., Kellar and Abraham, 2005) and reducing more negative actions such as saturated fat intake (e.g., Prestwich et al., 2008). Reviews have shown that implementation intentions have moderate-to-large effects on health behaviours (d = .52, k = 16, Prestwich et al., 2006; d = .59, k = 23, Gollwitzer and Sheeran, 2006) and a range of other behaviours (d = .65, k = 94, Gollwitzer and Sheeran, 2006). However, some recent, well-controlled studies that were not included in these reviews have reported null effects (e.g., Jackson et al., 2005, Jackson et al., 2006, Rutter et al., 2006) and the meta-analyses had significant variation in effect size, suggesting the presence of moderators. This review reports tests of moderators in studies that have manipulated, rather than measured, implementation intentions. The studies reported in this review fall into two categories. Either the study has hinted at possible moderators of implementation intention effects (but not used an appropriate design or analysis) or they have tested moderator effects. Table 1 summarizes studies in the latter category. These have typically used a two-level between-subjects design (implementation intentions: yes/no) where the moderator has been measured or a 2 (implementation intention: yes/no) × 2 (moderator: yes/no) between-subjects design where the moderator has been manipulated. In this review, all of the studies (unless otherwise stated) have manipulated implementation intention formation using written instructions embedded within questionnaires.

Moderator 1: intention

Intentions, the degree to which an individual is willing to try to perform a particular behaviour, typically explain between 25% and 30% of the variance in behaviour (Sheeran, 2002). According to Webb and Sheeran (2006), a “medium-to-large” change in intention (d = .66) leads to a “small-to-medium” change in behaviour (d = .36) so changing intentions is not always sufficient to drive significant changes in behaviour.

Implementation intentions are seen as a technique that helps people to translate

Moderator 2: self-concordance

One's pursuit of a specific goal (e.g., to perform well at work) can be described as self-concordant reflecting personal interests (e.g., I enjoy work) and values (e.g., it is important for me to always do my best) or non-concordant being pursued for controlled reasons such as social pressure or financial gain. Koestner et al. (2006), Study 1) delivered implementation intentions in either an ‘autonomy-supportive’ or in a more controlling format and compared these groups against each other and a

Moderator 3: collaboration

Implementation intentions might be more effective when two people form plans and subsequently perform the behaviour together. Prestwich et al. (2005) randomly allocated participants to implementation intention groups or no-implementation intention groups. In these groups, female participants could either choose to perform breast self-examination (BSE) in the next month alone or with a partner. Those in the implementation intention group choosing to involve their partner were requested to decide

Moderator 4: plan reminders

Sniehotta et al. (2005) showed that an implementation intention intervention supplemented with six personalized weekly diaries that reminded individuals of their plans was generally superior to an equivalent manipulation without reminders and a control condition across some (but not all) physical activity-related measures. Plan reminders should strengthen the underlying processes through which implementation intentions change behaviour (Prestwich et al., 2009). Two mechanisms have been

Moderator 5: ‘if-then’ plans versus ‘global’ implementation intentions

In a study concerned with promoting fruit and vegetable consumption, Chapman et al. (2008) compared two types of implementation intentions against a control group. In the first type, referred to as ‘global implementation intentions’, participants were free to form a behavioural plan but were instructed to pay particular attention to the situations in which they would implement these plans (Armitage, 2004). In the second type, ‘if-then’ plans (Oettingen et al., 2000), participants were required

Moderator 6: goal type

There have been much fewer tests of whether implementation intentions can reduce unwanted behaviours (avoidance goals) compared to promoting desired behaviours (approach goals, cf. Gollwitzer and Sheeran, 2006). Sullivan and Rothman (2008) described ‘moderation of the implementation intention effect’ (p. 443) in the context of goals implying implementation intentions are better for avoidance goals. However, a significant goal × implementation intention interaction was not specified for either

Moderator 7: conscientiousness

Webb et al. (2007) reported implementation intentions were more effective in promoting class attendance for those of low or moderate conscientiousness compared to those that are highly conscientious. While this interaction emerged in relation to a behaviour in which conscientiousness was also a direct predictor of behaviour, it is not known whether this interaction also emerges for behaviours for which conscientiousness is not directly related. For example, while Walsh et al. (2005) detected a

Other tested moderators

Chapman et al. (2008) demonstrated that manipulating implementation intentions increased fruit and vegetable consumption with or without information highlighting they are an effective means to change behaviour. Consequently, demand characteristics do not seem to moderate implementation intention efficacy. In a recent study testing whether stress moderates the effects of implementation intentions on exercise, planning actually backfired such that those forming implementation intentions exercised

Conclusion

The studies described represent a limited subset of the implementation intentions literature. Elsewhere, authors have called for more interactionalist work to address the question of what can account for differential levels of success in turning delayed intentions into behaviour (McDaniel and Einstein, 2000). Despite the limited number of studies available, it is clear that a range of moderators exist. For some of these moderators (e.g., conscientiousness, goal difficulty) the results are

Cited by (47)

Recommended articles (6)