Running Head: DISSOCIABLE EMOTION REGULATION STRATEGIES Divergent Cognitive Costs for Online Forms of Reappraisal and Distraction

The present study was set out to evaluate the cognitive costs of two major emotion regulation strategies under conditions of increased challenge. Previous studies have established that cognitive reappraisal (construing an emotional event in non-emotional terms) has no cognitive costs. However, in all of these studies, reappraisal was initiated at the emotional situation onset, before emotional response tendencies sufficiently evolved. In the present study the challenge of regulation strategies was increased by initiating strategies online, at a late time point in an emotional situation. Applying this procedure revealed for the first time a cognitive cost for reappraisal and also provided double dissociation between reappraisal and another major cognitive emotion regulation strategy distraction (diverting attention from an emotional situation via producing neutral thoughts). Specifically, late reappraisal, relative to distraction, resulted in an expenditure of self control resources. Late distraction but not reappraisal impaired memory encoding of the emotional situation. Imagine yourself trying to decide between two products designed for the same purpose. Critical questions that pop to mind might be whether these products achieve their purpose equally and whether their prices are comparable. Psychologists from various sub-disciplines ask similar questions concerning emotion regulation strategies (see Gross, 1998 for a review). In this work we focus on the cognitive profile and costs of two major emotion regulation strategies; Distraction, which refers to engaging in another neutral thought (e.g. Nolen-Hoeksema, 1991), and Reappraisal, which involves reinterpreting the emotion invoking stimulus as non-emotional (e.g., Gross, 1998). More specifically in the present study we set out to answer the questions: in what ways are distraction and reappraisal different from one another and what are their cognitive costs? Various studies, using different procedures and measures, have established that reappraisal has no cognitive costs (e.g., Gross, 2002; Richards, 2004, for reviews). Most of them have focused on memory performance and have shown that initiating reappraisal leaves memory of the emotional situation intact or in some cases even improves the recall of the emotion-related event (e.g., Dillon, Ritchey, Johnson, & LaBar, 2007; Richards, Butler, & Gross, 2003; Richards & Gross, 2000). In addition, a related study found that reappraisal participants were not distracted during an emotional conversation (Butler et al., 2003). Of particular interest in the present context is a single study that has shown indirectly that reappraisal does not consume self control resources (Vohs & Schmeichel, 2003). This study was inspired by the ego depletion theory, which views self control as a limited resource which gets depleted when one tries to inhibit competing behaviors, urges or desires (see Muraven & Baumeister, 2000, for review). According to this theory, the exertion of self-control appears to depend on a limited resource. Just as a muscle gets tired after performing an effortful action, an initial act of a self-control task causes impairments (ego depletion) in the performance of a subsequent self-control task. The main argument of Vohs and Schmeichel (2003) was demonstrated via a mediation model where initial self control affects subjective time estimation which in turn predicts subsequent self regulation performance (see also Wen Wan & Sternthal, 2008). With respect to reappraisal, the authors showed that it did not result in subjective over-estimation of the duration of the film clip used to induce the emotion (Vohs & Schmeichel, 2003, Experiment 2). This indirect evidence that reappraisal does not deplete self control resources, was explained by Gross' (1998) process model of emotion regulation. This model views reappraisal as an antecedent focused emotion regulation strategy which starts operating early in the emotion generative process, before response tendencies are fully activated. Accordingly, it was proposed that reappraisal does not cause ego depletion, because it diverts the emotional trajectory off track quite early, making continuous self monitoring demand negligible (Baumeister, Schmeichel, & Vohs, 2007; Vohs and Schmeichel, 2003). Note that in all of the aforementioned studies, reappraisal was indeed initiated at a time point in which the emotional response has probably not sufficiently evolved (at the mood induction onset), minimizing the self control challenge. We argue that to seriously challenge self control resources and to cause ego depletion, one has to initiate regulation strategies after the emotion has sufficiently evolved. We defined this phenomenon “online regulation” the attempt to change an emotion which starts and continuously operates during an emotional situation (Sheppes & Meiran, 2007). In that study we tested distraction and reappraisal in two ways. First, we replicated previous results in showing that when initiated early in the emotion generative process, both strategies were equally effective in reducing sad mood. However, when both strategies were initiated at a late time point in sadness inducing films, reappraisal resulted in less effective down-regulation of negative mood relative to distraction. To explain this result we relied on the notion that reappraisal involves attending to the emotional situation but changing its emotional meaning into a neutral one (Gross, 1998). Therefore, the reappraised neutral contents are, by definition, associatively linked to, and depend on, the contents which have caused the emotion to rise. Accordingly, in the present study we attempted to show that late reappraisal may deplete self control resources, because it requires overcoming a previously well established tendency of identifying with the emotional content (formed during the long unregulated duration prior to the strategy initiation). But what about the price tag of distraction? Distraction (as we operationally define it) is an attentional deployment strategy that narrows the emotional situation by producing neutral thoughts. Cognitively speaking, distraction involves diluting the proportion of emotional contents in working memory (WM) by loading it with neutral contents retrieved from long term memory (see especially Van Dillen & Koole, 2007). The cognitive cost associated with loading WM with neutral contents is an impairment of the emotional-situation encoding process, demonstrated in later impoverished recall of the emotional situation (e.g., Richards & Gross, 2006; Sheppes & Meiran, 2007). Back to online regulation whereas late reappraisal involves attending to the emotional situation while struggling to transform a well established emotional interpretation of the contents into a neutral interpretation, late distraction involves diverting attention away from the emotional situation and its contents by producing independent neutral contents. The present work was designed to test the predictions concerning divergent cognitive costs of online forms of Distraction and Reappraisal, by manipulating both strategies at a late time point in a sadness inducing film. Based on our previous findings (Sheppes & Meiran, 2007), we predicted that distraction but not reappraisal would impair memory encoding once initiated. To that end, we administered a surprise memory task following the film that checked the recognition for film facts prior and subsequent to the strategy initiation. In contrast, the increased self control challenge associated with online late reappraisal (relative to distraction) was predicted to lead to ego depletion. As in several previous studies, we used the Stroop task to assess ego depletion (e.g. Inzlicht, & Gutsell, 2007; Richeson & Shelton, 2003; Richeson & Trawalter, 2005). We therefore predicted that initiating reappraisal late would result in an increased Stroop effect relative to distraction. Method Participants and Procedure Forty six undergraduate students (34 women, mean age 23.7) participated in the experiment for course credit or monetary compensation (30 NIS; approximately US$7). Since a (Hebrew) Stroop task was used all participants were native Hebrew speakers. The experiment was administered individually. After signing consent forms participants were given short instructions regarding the verbal Stroop task followed by performing a practice phase. Immediately afterwards, participants were randomly assigned to reappraisal (n = 23) or distraction (n = 22) conditions and received verbal instructions. To prevent them from using a strategy immediately after the film began (and prior to the late manipulation), all participants were given two types of verbal instructions: instructions of one of the strategies (distraction or reappraisal), and instructions for a control unregulated condition (which includes allowing their feelings). The participants were also asked how they planned to implement the strategy if asked, in order to ensure their comprehension of the instructions and the immediate initiation. Participants were told that their strategy condition would appear via subtitles during the film; that only one type of subtitle would appear and would remain valid thereafter; and were told to allow their feelings to arise before receiving the subtitle instructions. The distraction condition involved asking participants to think about something unrelated to the film content and emotionally neutral. The reappraisal condition (which adhered closely to instructions given by Richards and Gross, 2000) involved asking participants to adopt a neutral, analytical and objective attitude toward the film contents. Participants watched a 403 second film clip taken from the TV documentary “The Real Story”, about Holocaust survivors hospitalized in a mental institution after being abandoned by their families and society. It was previously shown that this film mainly induces sadness (Sheppes & Meiran, 2007). 190 seconds following the film’s onset, the subtitles (containing the regulation condition) appeared at the bottom of the screen and remained throughout the film. After watching the film, participants were given a mood check, followed by a test phase of the Stroop task, and a surprise memory test.