Cooling the heat of temptation: Mental self-control and the automatic evaluation of tempting stimuli

The present research investigated whether mental self-control strategies can reduce the automatic positivity elicited by tempting stimuli. In two studies employing chocolate as the temptation of interest, we found that participants instructed to imagine a chocolate product in a nonconsummatory manner exhibited significantly less automatic positivity with regard to the product as compared to participants instructed to imagine the hedonic, consummatory aspects of the product and control participants engaged in a neutral task. These findings were replicated in a second study. Additionally, in Study 2 we found that automatic evaluations of chocolate were lowest for participants instructed to form implementation intentions to refrain from consumption. Taken together, these findings demonstrate that mental self-control strategies such as nonconsummatory transformation and implementation intentions extend to the level of automatic processing by reducing the positivity of automatically activated affective responses. Copyright # 2009 John Wiley & Sons, Ltd. In their everyday lives, people often face the challenge of resisting tempting impulses (Baumeister & Heatherton, 1996). In order to improve their chances of resisting temptation, people employ a variety of self-control strategies (e.g., Hoch & Loewenstein, 1991; Metcalfe & Mischel, 1999; Trope & Fishbach, 2000). For instance, individuals may precommit themselves to abstinence or control before entering a tempting situation such that yielding to temptation becomes less probable. A classical epic example is the story of Ulysses and the sirens in which Ulysses ordered his seamen to plug their ears with wax and bind him to the mast, so that, no matter how luring the sirens’song, it would be impossible for him to cast himself into the sea. In modern terminology, the seamen and Ulysses engaged in drastic forms of stimulus control (i.e., avoiding the very perception of the tempting stimulus) and response control (i.e., removing the possibility of a response), respectively. Often, however, delegating self-control to precommitment devices such as ropes or the power of others is not an option. Rather, ordinary people have to confront temptations at hand everyday by means of cognitive self-regulatory strategies targeted at stimulus or response control (Hoch & Loewenstein, 1991; Metcalfe & Mischel, 1999). For instance, the seminal research on delay of gratification suggests that cognitive transformation, a form of cognitive stimulus control, can aid selfcontrol when there is no chance to avoid perceiving the tempting stimulus (Mischel, 1974; Mischel & Baker, 1975). Particularly, if children were instructed to cognitively focus on the nonconsummatory qualities of a delayed reward (e.g., imagining marshmallows as white puffy clouds), waiting time was greatly enhanced. In contrast, if children were tment of Psychology, University of Würzburg, Roentgenring 10, 97070 Würzburg, Germany. .de s, Ltd. Received 13 June 2009 Accepted 25 July 2009 18 Wilhelm Hofmann et al. instructed to imagine the consummatory aspects of the reward (e.g., the sweet soft taste of the marshmallows), waiting for the delayed reward became increasingly difficult (Mischel & Baker, 1975). Another prominent mental strategy pertaining to cognitive response control is forming implementation intentions (Gollwitzer & Brandstätter, 1997). In essence, implementation intentions can be described as cognitive structures that link an appropriate goal-directed response to a suitable situation in which to perform that response. Research by Gollwitzer and colleagues suggests that implementation intentions greatly facilitate the translation of intentions into actual behavior (e.g., Gollwitzer, 1999; Gollwitzer & Brandstätter, 1997; Gollwitzer & Sheeran, 2006). In the present research, we built on this work and investigated whether cognitive stimulus-control (i.e., nonconsummatory transformations) and/or cognitive response-control (i.e., implementation intentions) reduce the degree of automatic positivity triggered by the tempting stimulus at which they are directed. In Studies 1 and 2 we investigated the effects of nonconsummatory transformation of a tempting stimulus on its automatic evaluation. We assumed that nonconsummatory transformation of a tempting stimulus may change the automatic evaluation of the stimulus by means of changing which particular information will be triggered in memory. Related experiments in the realm of stereotypes support this general notion. For example, participants who engaged in counterstereotypic mental imagery (e.g., imagining the features of strong women) exhibited reduced scores on an Implicit Association Test of gender stereotyping (Blair, Ma, & Lenton, 2001). Applied to automatic attitudes toward tempting stimuli, we expected that nonconsummatory mental transformation of a chocolate stimulus to decrease its automatic positivity, relative to a neutral comparison group and a group instructed to imagine its consummatory aspects. In Study 2 we additionally investigated the effects of implementation intentions on automatic evaluations of tempting stimuli. The effectiveness of implementation intentions on self-regulatory goal pursuit is well documented (e.g., Achtziger, Gollwitzer, & Sheeran, 2008; Gollwitzer & Sheeran, 2006). There is also increasing evidence about the automatic nature by which implementation intentions produce their effects (e.g., Bayer, Achtziger, Gollwitzer, & Moskowitz, 2009; Webb & Sheeran, 2007). However, there is still scant research on the relation between implementation intentions and automatic associations with regard to the objects toward which implementation intentions may be directed. Recent work by Stewart and Payne (2008) in the domain of stereotypingshowed that implementation intentions to think ‘‘safe’’ upon appearanceofa Black face in the weapon identification paradigm (Payne, 2001) reduced automatic stereotyping. Similar to the mental transformation work cited above, this research primarily addressed automatic stereotyping but not hedonic evaluations of tempting stimuli. Furthermore, implementation intentions in Stewart and Payne (2008) were directed specifically at improving task performance during the implicit measure itself. In the present research, we investigated in the domain of food temptation (e.g., Achtziger et al., 2008) whether more general implementation intentions to resist a temptation in an imagined everyday scenario (e.g., ‘‘If somebody offers me a piece of chocolate, I will say no thanks!’’) can decrease the automatic evaluation of the tempting stimulus. This prediction can be derived from research suggesting that avoidance orientations are strongly related to affectively negative representations in memory and, therefore, facilitate negatively valenced evaluations and judgments (e.g., Cacioppo, Priester, & Berntson, 1993; Neumann, Förster, & Strack, 2003). Engaging in implementation intentions to resist consumptionmay strengthen the link between the temptationandnegativelyvalenced avoidance behaviors, thereby effectively ‘‘cooling the heat’’ of temptation.