This item is only available as the following downloads:
CROSSING THE THRESHOLD: AN EXAMINATION OF THE DURATION OF SELF REGULATORY TASKS WITHIN THE EGO DEPLETION PARADIGM BY KELSEY COOKE A Thesis Submitted to the Division of Psychology New College of Florida in partial fulfillment of the requi rements for the degree Bachelor of Arts Under the sponsorship of Michelle Barton Sarasota, Florida May, 2012
ii Acknowledgements I would like to extend my gratitude to everyone involved in the thesis process over the past year. This project would not have been possible without the guidance, support, encouragement and interest of my friends and family. Specifically, I would like to thank my mother for being the inspiration behind my thesis topic; without her stories of ambitious attempts to hold out on eating homemade French fries and various chocolaty temptations over long periods of time, my research question would never have existed. This topic became a reality thanks to the hard work and tutelage of my thesis sponsor, Michelle Barton, whose guida nce and input were invaluable. I would not have been as prepared for this project had I not taken classes with the other members of my baccalaureate committee, Steve Graham and Heidi Harley. Thank you to Duff Cooper for all the assistance analyzing my data I owe my confidence in my topic, my (maintained) sanity, and my clarity of writing to my loving boyfriend, best friend, and editor, Daniel Sweet. To those who participated and those who helped me recruit participants, thank you for soothing the thesis cr azy out of me! Finally, I would like to thank Boris the Cat, whose vigilant support in the early stages of this process, by means of purring next to me, on my feet, or on my computer, will be forever remembered, appreciated, and missed, dearly.
iii TAB LE OF CONTENTS ACKNOWLEDGEMENTS ii TABLE OF CONTENTS iii ABSTRACT v INTRODUCTION 1 Evidence for the Strength Model of Self Control 5 Sources of Depletion 11 Predisposing Factors 11 Social Influences 17 Areas Affected By Ego Depletion 22 Affective Influence of Ego Depletion 22 Cognition and Perception 25 Counteracting Ego Depletion 29 Perception 29 Distraction 31 Priming 32 Autonomy 35 Sel f Affirmation 36 Mood 37 Self Control as a Finite Resource 39 The Current Study 44 METHOD 47 RESULTS 52
iv DISCUSSION 56 REFERENCES 68 TABLE 1 74 TABLE 2 75 APPENDIX A 79 APPENDIX B 85 APPENDIX C 91
v List of Tables TABLE 1: Interference in Inzlicht and Kang (2010)'s Original Study and the Current Study TABLE 2: Duration of Self Regulatory Tasks Throughout Ego Depletion Literature
vi CROSSING THE THRESHOLD: AN EXAMINATION OF THE DURATION OF SELF REGULATORY TASKS WITHIN THE EGO DEPLETION PARADIGM Kelsey Cooke New College of Florida, 2012 ABSTRACT Self regulation is the suppression of an idea, emotio n, or urge, often in an effort to meet a goal, or to align with social expectations, morals, and values. Self regulation exertion has been found to result in self control failure, also known as ego depletion. To test whether the amount of time spent self r egulating affects the occurrence of subsequent self control failure, the duration of a self control task was manipulated so that participants self regulated for either a short period of time (1 min), a long period of time (6.5 min), or not at all. Afterwar d, self regulatory ability was measured on the Stroop task. No overall difference was found between groups on Stroop performance. However, there was a significant difference between the short duration and long duration groups on initial Stroop performance. No difference was found in mood between groups. These findings suggest that the duration of a self control task affects self control performance, such that longer periods of self regulation result in worse subsequent self control. However, it is unclear w hether ego depletion occurred because the control group did not differ from the depletion groups. This reinforces the importance of replicating previous findings, and supports the failure of previous studies to report significant effects of ego depletion.
vii Michelle Barton Division of Social Sciences
1 Crossing the Threshold: An Examination of the Duration of Self Regulatory Tasks Within the Ego Depletion Paradigm Effectively controlling the self requires effort. This can be considered problematic, as it may not be easy to make the self behave. For example, in her comedic memoir Bossypants Tina Fey (2011) discusses her success at achieving her ideal weight, describing her efforts as a feat: For a brief time at the turn of the century, I was very skinny. This is what I remember about that period. I once took a bag of sliced red peppers to the beach as a snack. I regularly ate health food cookies so disgusting that when I enthusiastically gave one to Rachel Dratch she drew a picture of a rabbit and broke the cookie into a trail of tiny pieces coming out of the rabbit's butt. I ran three miles a day on a treadmill six days a week. (p. 115 116) Fey (2011) comically depicts the actions she used to reach her ideal weight, eating flavorless foo ds and exercising excessively, as obviously less than enjoyable. This highlights how refraining from the instant gratification of delicious, albeit unhealthy foods is hard; it requires self control. Also referred to as self regulation, self control means e ngaging deliberately in any volitional act (Baumeister, Bratslavsky, Muraven, & Tice, 1998). This requires the suppression or monitoring of a certain behavior, thought, or urge, often in an effort to reach a certain goal (e.g., to be thin); other motivatio ns for the use of self control include relating to certain ideals, values, morals, or social expectations (Baumeister, Vohs, & Tice, 2007). Baumeister et al.
2 (1998) examined the ability to carry out these behaviors by placing a plate of freshly baked cooki es in front of hungry participants and making them eat unappealing radishes instead. The behaviors that Baumeister et al.'s (1998) participants and Fey (2011) engaged in effectively represent the nature of self control as an important facet of everyday goa ls and tasks. However, successfully using self control is not always easy, and despite many great attempts, it is possible to experience self control failure, as Fey (2011) elaborates: For a brief time at the end of that last century I was overweight. This is what I remember about that period. I once left a restaurant in the middle of dessert to get a Krispy Kreme before it closed. Even though I only liked McDonald's fries, I believed it was more nutritious to make a meal of it and have two cheeseburgers as well. If I was really ambitious, I would get a Whopper Jr. at Burger King and then walk to McDonald's to get the fries. The shake could be from anywhere. ( p. 117) The behaviors that Fey (2011) discusses highlight how challenging using self control can be when temptations, like Krispy Kremes and french fries, get in the way of a goal. In support of Fey's (2011) less healthy behaviors, Baumeister et al. (1998) suggested that self control is a finite resource that gets temporarily depleted with use. Specifica lly, Baumeister et al. (1998) found that only the hungry participants who had already used self control (i.e., they ate radishes while seated in front of a plate of tempting chocolate chip cookies) performed significantly worse on a subsequent self
3 control task in comparison to hungry participants who had not already used self control (i.e., they ate chocolate chip cookies while seated in front of a plate of unappealing radishes). This illustrated that using self control to refrain from eating delicious coo kies impaired further use of self control. An additional body of research has extended these findings to various tasks and constructs, indicating that engaging in any task that requires self control results in the temporary impairment of self regulatory ab ility on subsequent tasks (Muraven, Tice, & Baumeister, 1998; Baumeister et al., 1998). This phenomenon is called ego depletion, and has been observed in response to various volitional acts and applied situations. Building onto these findings, there is a range of self control acts that have been found to lead to ego depletion, such as challenging social interactions (Finkel et al., 2006), coping with stereotype threat (Inzlicht & Kang, 2010) and dispositional factors, like social orientation (Balliet & Joi reman, 2010). Additionally, research indicates that ego depletion can affect a variety of self regulatory domains, including overeating (Vohs & Heatherton, 2000), risk taking and decision making (Unger & Stahlberg, 2007), perception and information process ing (Vohs & Schmeichel, 2003; Schmeichel, Vohs, & Baumeister, 2003; Neshat Doost, Dalgleish, & Golden, 2008), as well as affective factors like the regulation of aggression ( DeWall, Baumeister, Stillman, & Gailliot, 2007) and passivity (Baumeister et al., 1998). However, certain factors have been observed to counteract ego depletion, including receiving a gift (Tice, Baumeister, Smueli, & Muraven, 2007), making autonomous decisions (Moller, Deci, & Ryan, 2006) and, interestingly, replenishing blood glucose levels (Gailliot et al., 2007). While there is great depth to the ego depletion literature, the
4 duration of tasks used in previous methods has varied from study to study. It is unclear whether the amount of time spent self regulating affects the occurrence of self control failure. Thus, the goal of this study was to examine whether there is a threshold at which one becomes ego depleted. In order to explore the idea of a threshold in ego depletion, it is important to understand how self regulation and ego de pletion can impact seemingly unrelated tasks, both in the laboratory and in everyday life. Thus, the following literature review will provide an overview of self regulation and ego depletion findings, focusing on methodological ways to strengthen this rese arch. To do so, it is necessary to establish self control as a limited resource by examining early studies regarding this phenomenon. The strength model of self control, which dictates that self control is a "muscle," in that it experiences fatigue followi ng exertion (Muraven et al., 1998), will be discussed, along with literature supporting this model. Next, the literature review will move on to factors that cause ego depletion, followed by the areas impacted by ego depletion. Ways to counteract the occurr ence of ego depletion following the use of self control will be explored, leading up to research attempts to identify the exact resource that depletes during self regulation. Identifying the limited resource that depletes during self regulation necessitate s research looking into how long one needs to self regulate to become depleted. Comparing self control performance after various durations of self regulatory tasks will indicate whether there is a threshold at which ego depletion occurs.
5 Evidence for the Strength Model of Self Control As mentioned briefly earlier, ego depletion may be explained in terms of the strength model of self control, which posits that self regulation requires physiological arousal, effort and exertion (Muraven et al., 1998). Based on this model, Muraven et al. (1998) argue consequently that self control must also be a limited resource, such that self regulatory exertion will impair further self regulatory performance. The exertion of self regulatory ability results in fatigue in a similar fashion to muscular strength; when this energy is expended, reduced capacity for the behavior follows, although this is not permanent. To test the strength model of self control, Muraven et al. (1998) used the dual task paradigm, in which self regu latory performance was measured immediately after self regulatory exertion. This method has been replicated throughout self regulatory research as an effective way to create an ego depleted state. In Experiment 1, self regulatory exertion was manipulated t hrough emotional regulation; two experimental groups suppressed or exaggerated their emotional response (i.e., performed an emotional regulation task) to an upsetting, 3 min movie, while control participants were instructed to react naturally to the movie. To ensure that mood did not influence findings, the Brief Mood Introspection Scale (BMIS) was distributed before the dependent measure. To measure self regulatory depletion, the researchers had participants perform a self regulatory task requiring physica l strength; participants had to squeeze a handgrip together, keeping a wad of paper between the tongs for as long as possible. Because physical stamina varies from person to person, the dependent measure was performed twice, once at the beginning of the ex periment as a
6 baseline and once after self regulatory exertion. The baseline handgrip duration score was then subtracted from the second handgrip duration score and used for analysis. As was hypothesized based on the strength model of self control, partici pants who had engaged in self regulatory exertion (i.e., participants in the exaggerate/suppress emotions groups) performed significantly worse on the second measurement of handgrip stamina; meanwhile, control participants did not exhibit a significant dif ference in handgrip stamina between the baseline and the second measurement. Additionally, the results confirmed that mood was not a main effect of self regulatory performance on the handgrip task. In Experiment 2 of Muraven et al.'s (1998) study, particip ants exerted self control first by performing a 6 min thought control task. The experimenter told participants to write down all of their thoughts on a piece of paper, which they were told would later be analyzed for its word usage. The participants were d ivided into one of three groups, a control group and two experimental groups; the control group received no further instruction on the task, while the experimental groups were given advice to guide their thoughts. One experimental group was told to think a s much as they could about a white bear, while the other was told to try not to think about a white bear. The dependent measure was the amount of time that the participants spent working on a set of anagrams. Once the participants quit, they filled out a q uestionnaire in which they rated how difficult the thought control task was. The results indicated that the participants who performed the thought suppression task quit significantly sooner than the control group and the experimental group that thought as much as they could about a white bear. In accordance with these findings, participants
7 in the thought suppression group perceived the thought control task to be significantly more difficult than the other experimental group or the control group. These find ings reiterate the occurrence of ego depletion using the dual task paradigm. Together, the results from Experiments 1 2 of Muraven et al.'s (1998) study support the strength model of self control, in that physical stamina and task persistence were impaired following self regulation exertion, similar to the way a muscle is fatigued after a work out (Muraven et al., 1998). Moreover, this paradigm of self regulation exertion, followed by a measure of self regulatory ability, has been repeatedly, conceptually r eplicated in subsequent ego depletion studies (Baumeister et al., 1998; Inzlicht & Kang, 2010; Vohs & Heatherton, 2000). Based on the strength model of self control, Baumeister et al. (1998) conducted four experiments using various methods to examine wheth er performance on a self control task was impaired when participants had already self regulated. Consistent with the idea of self control as a resource that depletes with exertion, the logic behind the methods employed within each experiment was to encoura ge self regulation by forcing participants within the experimental group to ignore an urge or impulse. Different methods were used in each experiment to create a depleted state, illustrating that ego depletion is not the result of one specific, self regula tory task. In Experiment 1 of the Baumeister et al. (1998) study, participants were seated in front of two plates, one with a stack of chocolate chip cookies, and one full of radishes. To ensure that participants were hungry, participants were instructed to skip one meal and to fast for 3 hr prior to the experiment. Additionally, the chocolate chip cookies used in the manipulation were baked in the lab where the experiment took
8 place to fill the room with the tempting aroma. Participants were assigned to o ne of three groups; in the chocolate group, participants could eat only the chocolate chip cookies on the plate, whereas in the radish group, participants could eat only the radishes on the plate. The control group did not participate in this section of th e experiment. After spending 5 min tasting these foods, the amount of time spent working on an unsolvable problem solving task was measured until the participants quit. The results indicate that participants in the chocolate group did not vary from those i n the control group on task persistence and number of attempts, though participants in the radish condition were significantly more likely to quit sooner than those in the chocolate and the control groups. In Experiment 2, Baumeister et al. (1998) hypothe sized that adopting counter attitudinal behavior results in ego depletion, as well. Participants, who were college students, read a persuasive speech for the university's Board of Trustees that either agreed with a raise in college tuition or disagreed wit h it. The logic behind this manipulation was that most students would disagree with the implementation of a tuition raise; therefore, reading a pro tuition raise speech would be counter attitudinal. The experimenter assigned participants in the no choice g roup to read only the pro tuition raise speech. Participants in the high choice conditions could choose which persuasive speech to read aloud. The experimenter, however, urged participants in the high choice group to choose a specific speech by explaining that too many participants had already chosen a certain topic; thus, participants in the high choice conditions could be in the counter attitudinal group or the pro attitudinal group. In the control condition, participants did not read a speech. As in Expe riment
9 1, task persistence on an unsolvable task was measured. Persistence dropped when participants had to adopt counter attitudinal behavior without being given a choice. Participants in the counter attitudinal high choice and pro attitudinal high choice groups exhibited significantly more persistence in comparison to the counter attitudinal no choice and control groups, thus indicating that being forced to adopt counter attitudinal behavior results in ego depletion. Similar to Muraven et al.'s (1998) st udy, Baumeister et al. (1998) had participants monitor their emotional response while watching either a sad or a funny, 10 min video clip in Experiment 3. However, this time, Baumeister et al. (1998) had participants use only emotion suppression, and did n ot have an emotion exaggeration manipulation as Muraven et al. (1998) did. Thus, participants in the suppressed emotion group were instructed to suppress all internal and external emotions while watching the clip; meanwhile, the control group was instructe d to react naturally to the video clip. To ensure that the manipulation worked, participants were told that their reactions would be videotaped. Afterward, participants were given an anagram task in which they had to unscramble letters to form English word s. The results indicate that performance on the anagram task was significantly worse in participants who suppressed emotions while watching either a sad or a funny video clip than in the control groups. In the final experiment of the Baumeister et al. (19 98) study, participants performed a complicated letter recognition task. Specifically, participants performed the letter e task, in which they had to cross out the letter e in one page of text copied from a statistics book. Whereas the control group simply crossed out the letter e every
10 time they came across it in the text, the experimental, depletion group was instructed to cross out the letter e according to a specific set of rules (i.e., it could only be crossed out when it was not adjacent to another vo wel or one letter away from a vowel). The goal of this set of rules was to force the participants to self monitor while performing the task, requiring self control. Additionally, the copy of the text that the experimental group received was very poor, maki ng the task more difficult. Afterward, passivity was measured while participants watched a deliberately boring movie. They were told to indicate when they wanted to stop watching the movie by using a buzzer; however, half of the participants had to press t he buzzer actively when they wanted to stop the movie, while the other half had to hold the buzzer down passively while watching the movie. Once the buzzer was released, the movie would stop. Those in the experimental, depletion condition stopped the movie significantly sooner when quitting required a passive response rather than an active response, and watched the movie for significantly longer than those in the control condition when stopping the movie required an active response. No differences were obse rved in the control group on movie duration when stopping the movie required a passive or an active response. This suggests that ego depletion may increase passivity. Together, the findings from Baumeister et al.'s (1998) four studies not only illustrate t hat self regulating results in ego depletion, but also that this is not limited to one specific act of volition. Based on these findings, Muraven and Baumeister (2000) compared self regulatory ability to a muscle, because it depletes after self regulation as a muscle fatigues after physical exertion. In support of this notion, Baumeister et al. (2007)
11 suggested that ego depletion may be an adaptation in response to the depletion of self regulatory resources; because self control is a valuable commodity, les s effort may be expended when in a depleted state because it is important to conserve self regulatory strength. Still, self control failure is not a result of learned helplessness; rather, the continuous use of self control strengthens self regulatory abi lity, just as a muscle strengthens with repeated exercise (Baumeister et al., 2007). Additionally, because self control affects so many different domains, it is suggested that the processes affected tie to the idea of free will, a concept applying to the a bility to override impulses and make moral decisions (Baumeister et al., 2007). Regardless, there are a vast number of tasks requiring self control. Sources of Depletion While the majority of ego depletion studies are performed within a laboratory setting it is important to understand that self regulation applies to real, every day situations. This can include internal constructs, like predispositions and characteristics (Vohs & Heatherton, 2000; Balliet & Joireman, 2010; DeWall, Baumeister, Mead, & Vohs, 2010), as well as external ones, like social encounters and behaviors (Finkel et al., 2006; Ciarocco, Sommer, & Baumeister, 2001; Inzlicht & Kang, 2010). Regardless of the nature of the source of ego depletion, research indicates that there are many facto rs that are present in every day life requiring and thus, depleting one's self regulatory resources. Predisposing Factors. There are a multitude of daily activities requiring self control, which can easily deplete the self's resources. In an effort to exam ine how the chronic use of self regulation may influence ego depletion, Vohs and Heatherton
12 (2000) hypothesized that ego depletion would be affected by the chronic delay of gratification associated with dieting. In Experiments 1 2, chronic dieters and non dieters (as determined by the Restraint Scale) were recruited to participate in a study similar to Baumeister et al.'s (1998) cookie manipulation, where initial self regulation involved watching a boring video for 10 min while trying to resist eating tempt ing snacks. In all experiments, participants were instructed not to eat within 2 hr before the experiment to ensure that they would be hungry. The plate of snacks was manipulated so that it was placed in front of the participant (high temptation) or on the other side of the room (low temptation). The experimenter told half of the participants that they could help themselves to the snacks, whereas the other half of the participants was instructed not to touch the snacks. In Experiment 3, a different depletio n manipulation was set up to examine how utilizing self control in general may affect eating habits in chronic dieters versus non dieters. Thus, in Experiment 3, the first self regulatory task required all participants to watch the same sad video clip whil e being recorded. Whereas half of the participants suppressed all internal and external emotional responses to the film, the other half simply watched the video without further instruction. The dependent variables measured in Vohs and Heatherton's (2000) study varied per experiment, as well. After the first self regulatory task, in Experiments 1 and 3, participants were taken into another room to taste and rate 3 different kinds of ice cream; the amount of ice cream eaten in grams was compared between grou ps. In comparison, Experiment 2 utilized a dependent measure of self regulatory ability that was not food related; rather, time spent on an unsolvable geometric task was
13 recorded. The results from Experiment 1 of the Vohs and Heatherton (2000) study illust rate that dieters in the high temptation manipulation ate significantly more on average than dieters in the low temptation manipulation. Additionally, an interaction was found, such that high temptation dieters who were told they could help themselves ate significantly more ice cream on average than low temptation dieters who were told to help themselves (Experiment 1). Experiment 2 built onto these findings by illustrating that dieters in the high temptation manipulation persisted at the unsolvable task fo r significantly less time on average than low temptation dieters. The results from Experiments 1 2 illustrate that perceived availability and proximity to snacks undermined self control in chronic dieters. Finally, when initial self regulation was not food related, but rather required emotional suppression, dieters who suppressed emotion ate significantly more ice cream on average than dieters that did not suppress emotion (Experiment 3). Together, the results from the three experiments indicate that chroni c inhibition from dieting depletes self regulatory ability. Adding to the idea that predisposing factors requiring chronic self regulation can influence ego depletion, Balliet and Joireman (2010) proposed that certain social orientation characteristics cou ld require the self to allocate more resources than others. Specifically, it was hypothesized that ego depletion would predict reduced concern for others based on participants' cooperative orientation, which may be pro social or pro self. In particular, eg o depletion was expected to affect the allocation of self regulatory resources in those who were pro self, based on the idea that cooperation in social situations requires self control. With this in mind, the
14 researchers proposed that those who are pro soc ial have likely strengthened their self regulatory ability by being cooperative, but that those who are pro self need to conserve their resources (Balliet & Joireman, 2010). To test this, trait self control was compared between pro social and pro self part icipants by examining performance on a set of games requiring them to choose outcomes based on the number of points awarded to the self or to others. Pro social participants were those who awarded the most points to the self and others in the game, while p ro self participants were those who gave the least number of points to others. Responses to a self control questionnaire indicated that pro social participants exhibited higher trait self control than pro self participants thus suggesting that self regula tory ability is more limited in those who exhibit pro self tendencies. Pro social and pro self participants performed the emotion suppression manipulation mentioned earlier, during which they had to retain neutral expressions by suppressing their emotional responses to a funny, 5 min video (Muraven et al., 1998; Baumeister et al., 1998). To ensure that those in the depletion condition self monitored emotional responses during the video, all participants were recorded. Lastly, participants completed a modifi ed version of the social orientation game, where the number of points allocated to the other person was measured (as in Experiment 1). The results indicated that pro self participants in the ego depletion condition allocated fewer points to others than pro self control participants and all pro social participants in both the depleted and the control conditions. Thus, these findings suggest that social cooperation requires self control, a task that is impaired when those who are pro self oriented are ego dep leted.
15 Conversely to Balliet and Joireman's (2010) research, DeWall et al. (2010) examined social dispositions that could improve self regulatory ability in a series of experiments. It was hypothesized that power from leadership positions could improve sel f regulation. In Experiment 1A, pairs of participants were randomly assigned to leadership and subordinate positions and were instructed to work together to replicate a structure created by the researchers using Legos. A control group of partners simply wo rked together on the same task without receiving leadership instructions. In Experiment 1B, participants were assigned to groups, so that the high power group recalled a time they had power over someone else, low power participants recalled a time when som eone else had power over them, and control participants recalled a neutral situation. Afterward, participants' performance was recorded for either a set of anagrams or a dichotic listening task in which participants had to attend to only the words beginnin g with m or p presented to their left ears while different recordings were presented to both ears. The results from Experiments 1A and 1B indicate that participants in high power positions significantly outperformed low power participants, such that high p ower participants recorded on average significantly more words on the dichotic listening task and correctly completed significantly more anagrams than low power and control participants, thus indicating that power positions strengthened self regulation. Fu rthermore, when participants were assigned high power, low power and control manipulations following a self regulatory suppression task (Experiment 2), performance on the dichotic listening task illustrated that high power participants in the depletion con ditions resembled the control group, whereas a difference between the depletion and control conditions in the low power
16 and control groups was observed consistently. Moreover, high power, depleted participants performed better on the dichotic listening tas k than low power, depleted participants and control power, depleted participants. These findings illustrate that power mediated ego depletion, thus counteracting the occurrence of impaired self regulatory ability. When, in Experiment 3, DeWall et al. (2010 ) found that previous findings from Experiments 1 2 were not replicated, such that high power manipulations did not strengthen self regulation or counteract ego depletion, it was hypothesized that leadership positions influenced perceptions of disdain for certain self regulatory activities. Based on participants' ratings of different tasks as unsuitable for a person of power, the experimenters examined performance on a task that had been rated disdainful (multiplication) versus a task that had been rated ac ceptable (the game Operation) in Experiment 4. The results indicated high power participants outperformed the control and low power groups on the self regulatory task only when it was considered acceptable for persons of power. Together, these findings sug gest that power has a strong effect on self regulation, such that it can strengthen self regulation, counteract ego depletion, and prioritize self control expenditure, such that leaders disdain tasks they consider unworthy. This contributes to the idea of dispositional factors contributing to self regulation, in that power strengthened self regulatory ability and also weakened it when the task was considered disdainful. This adds to the body of research on varying dispositional factors and situations that c an influence the use of self control and the utilization of the self's resources.
17 Social Influences. In addition to dispositional factors, research indicates that there are many different social situations that require self regulatory ability and thus, lea d to ego depletion. For example, Finkel et al. (2006) hypothesized that engaging in high maintenance social interaction impaired self regulatory ability. This was tested through a series of studies, in which participants were forced to deal with difficult social situations requiring the adequate allocation of self regulatory energy in order to cope. In Studies 1 5, a high or low maintenance social interaction was staged using a confederate. Although the methods for staging this form of social interaction v aried per study, all of the studies required experimental participants to self regulate during a social interaction, and this was followed by a measure of self regulatory ability. The staged social situations varied to include a 3 min task in which the con federate provided error filled (high maintenance group) or error free (low maintenance group), scripted instructions to the participant while solving a maze (Experiments 1 and 3), a 5 min task in which the confederate read scripted instructions for data en try that were either out of sync (high maintenance group) or in sync (low maintenance group) with what the participant was entering (Experiment 2), a 6 min task in which the confederate provided pessimistic (high maintenance group) or optimistic (low maint enance group) responses while the participant tried to solve a scripted problem that the confederate was supposed to have (Experiment 4), or a 30 min task in which the confederate and the participant took turns describing pictures while the confederate ant i mimicked (high maintenance group) or mimicked (low maintenance group) the participants' physical mannerisms and gestures (Experiment 5). In Experiments 1 3, control participants worked along on the same tasks as the
18 experimental participants. Afterward, a dependent measure was provided to gauge depletion of self regulatory ability. Dependent measures included rating the level to which they wanted to complete an easy, but fun anagram or a challenging, but rewarding anagram (Experiment 1), accuracy in compl eting GRE analytical problems (Experiments 1 3), pre and post test handgrip stamina (Experiment 4), or the number of failures and lack of attempts to remove a body part while playing the game Operation (Experiment 5). Additionally, participants in all exp eriments responded to a questionnaire regarding perception of depletion and perception of the interaction as high maintenance or low maintenance. Finkel et al. (2006) found that high maintenance participants rated the interaction as significantly more high maintenance than low maintenance participants in Experiments 1 2, although this was not observed in Experiments 3 5. Additionally, high maintenance interactions predicted significantly higher ratings of exhaustion. In Experiment 1, high maintenance partic ipants exhibited lower task motivation, in that they were significantly more likely to choose the easy anagram task than low maintenance and control participants. Additionally, high maintenance participants solved significantly fewer GRE problems and were less accurate (Experiments 1 3) and showed significant decrements in post test handgrip stamina (Experiment 4) and performance on the Operation game (Experiment 5) in comparison to low maintenance and control participants. These findings indicate that high maintenance social interaction impairs self regulatory ability, regardless of whether or not this exhaustion is subjectively perceived.
19 Social situations can also be high maintenance based on the self's actions toward others. For example, Ciarocco et al. (2001) hypothesized that the act of ostracizing another person requires active self monitoring and the suppression of the urge to respond to or interact with a person. Therefore, the researchers argued that the act of ostracizing someone should result in e go depletion. In both studies, all participants spent 3 min conversing with and getting to know a confederate; afterward, the experimenter provided staged positive or negative feedback about the confederate's impression of the participant. To create a stat e of depletion, participants engaged in a second 3 min interaction requiring those in the silent condition to ostracize the confederate by ignoring the person completely, and those in the conversation condition to speak freely. However, to give the illusio n that participants had a choice in ostracizing or conversing with the confederate, the experimenter explained that they already had enough participants conversing with the confederate, but the decision was up to the participant. In both conditions, the co nfederate was given four questions regarding the previous conversation with each of the participants, although in the silent condition, the confederate made a comment about herself when the participant ignored her after each question. Afterward, the Positi ve Affect Negative Affect Schedule (PANAS) was completed in order to gauge mood in participants. In Study 1, persistence on a series of anagram tasks (some of which were impossible) was measured following the ostracism manipulation. In Study 2, handgrip st amina (while trying to keep a sponge between the tongs of a handgrip) was measured before and after the ostracism manipulation. Participants then filled out a questionnaire asking about the positivity of the evaluations they received, whether
20 they liked th e confederate, and whether their task of conversing with or remaining silent toward the confederate was difficult. The results of the Ciarocco et al. (2001) study indicate that the confederate received significantly higher liking scores from participants who received positive feedback from the confederate as opposed to negative feedback, although there was no significant effect of liking the confederate (as a result of positive or negative feedback) on task persistence or handgrip stamina. Additionally, pa rticipants gave significantly higher ratings of liking the confederate when they conversed with the confederate, as opposed to ostracizing the confederate. Participants gave up significantly sooner on the anagram task (Study 1) and on the second handgrip m easure (Study 2) when they were in the silent condition as opposed to the conversation condition. Additionally, ostracizing was rated as significantly more difficult than conversing with the confederate, particularly when the ostracizer had received positi ve feedback from the confederate. Those who ostracized the confederate also experienced significantly more negative affect in comparison to participants who conversed with the confederate; however, mood and handgrip measurements were not significantly corr elated, indicating that the findings were attributed to the manipulation, and not to negative affect associated with ostracism. These results indicate that the act of ostracism is difficult and requires self regulation to complete. Strengthening the argume nt that social situations can be ego depleting, Inzlicht and Kang (2010) sought to examine the physiological effect of coping with stereotype threat. The researchers used an electroencephalogram (EEG) to monitor
21 brain activity during self regulation. Based on the hypothesis that dealing with stereotype threat requires self control, which would subsequently impair self regulatory skills in non stereotyped areas, Inzlicht and Kang (2010) had participants cope with stereotype threat or prejudice, which require s self monitoring. Female students completed a math test; these participants were recruited based on their reported awareness of the stereotype that women are worse at math than men. Half of the female participants were given a coping mechanism in which th ey looked at the math test objectively while taking it. After completing the test, participants received fake scores with negative feedback before performing the Color Stroop task, which required participants to read the color of ink in which a list of wor ds for colors were written (e.g., "red" was written in the color "green"). Errors on the Stroop task were measured, as was brain activity by the EEG. The researchers found that participants confronted with stereotype threat made significantly more errors o n the Stroop task than participants using a coping mechanism to deal with stereotype threat. Additionally, the EEG indicated that women confronted with stereotype threat were more vigilant on the Stroop task, in that they experienced significantly greater differences in waveforms in comparison to women in the coping group on congruent Stroop task trials (where the color of the word was the same as the word) than on incongruent trials. This illustrates that the experimental group monitored responses more clo sely due to having experienced stereotype threat. Together, these findings illustrate that stereotype threat impairs self regulatory performance by increasing self monitoring, a factor that is common in various other social situations.
22 The factors mentione d in this section illustrate how everyday tasks and experiences can demand self regulatory ability, whether through coping, monitoring, or suppressing thoughts or urges. Moreover, using self control to tackle these tasks and encounters has led to self cont rol failure on a variety of subsequent tasks. While there are many factors that can cause ego depletion, there are also many factors that can be influenced by ego depletion. Areas Affected by Ego Depletion Given the nature of self control as an important f acet for everyday tasks, it is necessary to examine the effect that an ego depleted state can have on certain behaviors. Since ego depletion has been shown to impair performance on a number of tasks (Muraven et al., 1998; Baumeister et al., 1998), the foll owing section seeks to explore how depleted resources affect everyday experiences. Specifically, ego depletion has been shown to influence several affective behaviors, including aggression (DeWall et al., 2007) and optimism (Fischer, Greitemeyer, & Frey, 2 007). Additionally, ego depletion can affect cognitive performance in several domains, as will be discussed. Affective Influence of Depletion. DeWall et al. (2007) suggested that the regulation of certain behaviors might become difficult when ego depleted. Using a conceptual replication of Baumeister et al.'s (1998) cookie experiment, DeWall et al. (2007) hypothesized that refraining from eating a delicious donut while being forced to eat a radish would result in greater acts of aggression. The experiment b egan by having participants write a brief essay about abortion, which would later be evaluated by a partner; additionally, participants rated their preference for different types of
23 food, including spicy, sweet, salty, and creamy. Afterward, participants e valuated the quality of a bogus essay supposedly written by a partner of matching attitude and gender. At this point, the same manipulation from the Baumeister et al. (1998) study was used, in which participants in the experimental group had to refrain fro m eating a donut and control participants had to refrain from eating a radish for 5 min. The experimenter then provided the participant with the partner's list of taste preferences and bogus evaluation of the essay on abortion; the partner's taste preferen ces and evaluation were always rigged so that the partner did not like spicy food and the participants' essay was considered to be "one of the worst essays I've ever read." Participants were asked to prepare a snack of chips and hot sauce for their partner s according to their partner's taste preferences. The amount of hot sauce each participant used, knowing that his or her partner did not like spicy food, was the dependent measure of aggression; this was done by recording the weight of the hot sauce bottle before and after the participant distributed the hot sauce. The results indicated a significant difference in the amount of hot sauce used between ego depleted and control participants, such that ego depleted participants used significantly more hot sauce than control participants. These results support the hypothesis that ego depletion results in increased aggression after an insult provocation. In addition to difficulty regulating socially undesirable behaviors like aggression, Fischer et al. (2007) sugg ested that ego depletion might diminish positive outlook. The researchers employed different methods in a series of five studies to examine the effect of ego depletion on positive attitude toward the self, optimism
24 about the future, and belief in the self' s control over the environment. Methods of self regulation varied to include both an affective and a cognitive self regulation task while watching a video, such that experimental participants either suppressed internal and external responses to a funny car toon or ignored words scrolling across the television screen while watching a 5 min video (Studies 1, 2, and 4). Control participants simply watched the videos without receiving further instruction. Other methods employed to create a state of ego depletion include the letter e task from Baumeister et al.'s (1998) study, which requires those in the experimental group to cross out the letter e according to a challenging set of rules (Study 3), as well as a thought suppression task, in which participants wrote out all of their thoughts, although those in the experimental group were asked not to think of a white bear (Study 5). After the self regulation manipulation in all five studies, participants provided self report measures of self relevant optimism, which includes overestimation of the self's abilities and characteristics, the self's control over the environment, and positive outlook for the future. These questions required participants to make self assessments in which they judged themselves as better or w orse at certain abilities (i.e., intelligence, ability to learn new things, memory, and work organization) in comparison to the average person (Study 1), indicate the level to which they believed they could win a gambling game (Study 2), report on belief i n future outcomes (Study 3), and describe positive and negative self attributes (Studies 4 5). Fischer et al. (2007) found that control participants reported significantly more above average ratings on self abilities than ego depleted participants (Study 1 ).
25 Additionally, ego depleted participants were less confident that they would win the game than control group participants (Study 2), and were significantly less optimistic about fulfillment in the future and rated significantly higher probabilities of th e occurrence of a severe disease in the future (Study 3). Studies 4 5 indicated that although all participants recalled significantly more positive self attributes than negative self attributes, ego depleted participants recalled significantly fewer positi ve and more negative attributes than participants who were not ego depleted. Together, the findings from Studies 1 5 indicate that ego depletion resulting from a variety of different volitional acts impairs self relevant optimism. This adds to the current body of research by applying ego depletion findings to affective areas, providing evidence for the fact that ego depletion can affect everyday behaviors and experiences. Cognition and Perception. The present section examines how the depletion of self regul atory resources can apply to every day constructs and resources. Research has indicated that diminished self control can also impact cognitive performance. For example, Schmeichel, Vohs, and Baumeister (2003) found that ego depletion affected some forms of information processing, but not others. Experiments 1 3 required participants to perform cognitive or affective self regulation (i.e., word or emotion suppression) while watching a video. Afterward, problems from different standardized test questions were distributed to examine general cognitive ability and higher order information processing. These problems were extracted from the Analytical and Verbal sections of the Graduate Record Examination (GRE); the Graduate Management Admission Test (GMAT), a test of general math ability; and the Common Entrance Test (CET), a cognitive test requiring greater mental effort.
26 The number of items answered correctly, as well as the proportion of correct responses and the number of items attempted, were measured. All par ticipants were given 10 min to work on these problems. Experimental participants attempted significantly fewer problems, provided more incorrect answers, and had a lower proportion of correct answers than the control group in all three studies. Moreover, p erformance on problems that required higher order reasoning was worse in ego depleted participants than in control participants. These results indicate that ego depletion had a detrimental effect on intellectual reasoning, particularly when this involved h igher order information processing. Adding to research about impaired information processing due to ego depletion, Unger and Stahlberg (2011) explored how risk aversion was affected by ego depletion when decision making tasks required information processin g. The set up for the experiment required that participants begin by performing a self regulatory task. Tasks varied so that the experimental group had to relearn and recall new meanings for common English words for 5 min, while the control group simply le arned 20 new Turkish words (Studies 1 2); additionally, the experimental group had to suppress internal and external expressions of emotion during a showing of an aversive film, while the control group watched the film without receiving instruction (Study 3). After this manipulation, participants in all studies performed an investment game in which they were managers and had to make a decision about which country to invest money in. Two countries were coded in a pilot study to be safe, but not as rewarding investment options, and two countries were coded to be risky, but more rewarding investment options. Descriptions of the risks and profits for varying
27 countries were provided, and participants were informed that they would receive money for participating i n the experiment, which they could increase or lose depending on their investment decision. Thus, the task required information processing in order to make a decision and involved a real life result. The results indicated that ego depleted participants wer e less likely to make high risk investment decisions, indicating that ego depletion results in risk aversion on self control tasks that require greater information processing. Further research on information processing indicates that ego depletion has an e ffect on memory, as well (Neshat Doost et al., 2008). Specifically, the researchers hypothesized that impaired self regulatory ability could prevent the retrieval of autobiographical, emotional memories. Participants were placed in one of three groups, one depletion condition and two control conditions. The depletion group performed the Color Stroop task for 6.5 min, in which a list of colors was presented, but written in ink that varied in color from the semantic meanings of the words presented (e.g., the word "red" was written in yellow); the task required participants to state the color of each word. One control group read aloud color names that were all written in black ink, while the other control group did not perform any task at all before the depende nt measure. Afterward, all participants performed the Autobiographical Memory Test (AMT), in which participants were provided with cue words, from which they had 30 s to recall an autobiographical memory. Cue words could be positive, negative, or neutral. This task was recorded for later coding of specific and nonspecific memories, the results of which indicated a significant difference in memory recall between participants that completed the Color Stroop
28 task and both control groups, such that Color Stroop participants remembered fewer specific, autobiographical memories and more nonspecific, categorical memories than both of the control groups. These results indicate a significant effect of self regulatory condition on autobiographical recall, such that de pleted resources reduced specificity of autobiographical memories, but allowed for the recall of over generalized memories. Aside from information processes, perception may be influenced by the occurrence of ego depletion. Vohs and Schmeichel (2003) used v arying methods in a series of studies to test the idea that perceptions of time increased when participants were ego depleted. To ensure that participants could not track how much time had passed during the study, all participants in Studies 1 4 were instr ucted to take off all jewelry and watches. A state of ego depletion was created in the experimental groups through emotion suppression/emotion exaggeration while watching a 10 11 min, emotional video (Studies 1 2), acting excited and interested while readi ng aloud text from a psycholinguistics book for 4 min (Study 3), or writing out thoughts freely while refraining to think of a white bear for 6 min (Study 4). Control participants in each of the studies performed the same tasks, but without having to self monitor or self regulate (i.e., they did not receive instructions as to how to watch the film or read the text in Studies 1 3, and there were no guidelines as to what they couldn't think about in Study 4). The results indicated that participants made signi ficantly greater estimations and over predictions than control participants about the time they spent on a task when they had self regulated, regardless of the task. Together, these findings indicate that ego depletion influences the perception of time spe nt self
29 regulating, such that people overestimate the amount of time they spent self regulating when ego depleted. While ego depletion has been found to affect both affective behaviors and cognitive, perceptual performance, it is important to note the vari ety of implications these findings can have. For example, impaired regulation of aggressive behaviors as a result of ego depletion may negatively influence social encounters and experiences (DeWall et al., 2007). This highlights the importance of ego deple tion in everyday tasks and interactions. Counteracting Ego Depletion Despite research indicating that self regulation results in self control failure, there are certain factors that may counteract this occurrence. In order to test this, it is necessary to insert an additional manipulation into the dual task paradigm (Hagger, Wood, Stiff, & Chatzisarantis, 2010). Specifically, before the dependent measure of self regulatory ability, the variable needs to be introduced in order to examine how it affects the o ccurrence of ego depletion. The nature of certain variables used to counteract ego depletion has varied throughout ego depletion research, indicating that there is not one, sole factor that boosts self regulatory ability. Perception. While previous ego d epletion findings are substantial, researchers have suggested that perception may influence the occurrence of ego depletion, such that belief that one's resources are depleted is enough to cause ego depletion (Clarkson, Hirt, Jia, & Alexander 2010). To exa mine this, Experiments 1, 2, and 4 required participants to perform a version of letter e task used in Baumeister et al.'s (1998) study; for the purposes of Clarkson et al.'s (2010) study, the letter e task
30 was adjusted so that participants had to break a habit. Participants began by crossing out the letter e on a page of meaningless text. Those in the low depletion condition continued this task, while those in the high depletion condition must break this newly formed habit by following a new set of rules w here the letter e is only crossed out when it is adjacent to or one letter away from another vowel. In Experiment 3, the first self regulatory task varied from the other experiments, in that participants performed a thought suppression task while writing o ut all of their thoughts freely on a sheet of paper. For this manipulation, participants in the depletion condition were instructed not to think of a white bear, and to indicate on the sheet if they thought of a white bear. After self regulating, participa nts in all experiments were assigned to one of three groups: a control group, a group informed that the color of the page of text or the page they had written their thoughts on replenished mental resources, and a group told that it depleted them. Next, par ticipants completed a second self regulatory task in which either reaction time and task persistence on a computerized letter or number task (Experiments 1 3) or discrimination between persuasive arguments was measured (Experiment 4). Finally, participants filled out a questionnaire after the feedback manipulation to gauge individual perception of mental depletion. Clarkson et al. (2010) observed that participants in the low depletion group reported higher levels of depletion when they received the depletio n manipulation and participants in the high depletion group reported higher levels of depletion when they received the replenished manipulation, though these differences were not significant. The results indicated that participants in the low depletion gro up persisted significantly longer on the letter and number tasks when they believed their mental
31 resources were replenished, whereas participants in the depletion group persisted significantly longer on the letter or number task when they believed their me ntal resources were depleted. No difference was found between the persistence of the depleted group who received depleted feedback and the high depleted/low depleted groups that did not receive feedback at all. Additionally, in Study 4, participants in the low depletion condition that were given the replenished feedback perceived themselves as less depleted and could discriminate better between strongly and weakly persuasive arguments, while participants in the high depletion condition that were given the d epleted feedback perceived themselves as less depleted and could discriminate between strongly and weakly persuasive arguments. The results of the four experiments illustrate an interaction between perception and ego depletion, such that receiving a situat ional attribution for internal state counteracted the occurrence of ego depletion. Distraction. In an effort to examine factors that strengthen the ability to resist urges in favor of delayed gratification, Alberts, Martijn, Nievelstein, Jansen, and de Vri es (2008) acknowledged how distracting the self led to greater self regulatory performance. This idea was present in Mischel and Ebbeson's (1970) findings, in which the children who were able to preoccupy themselves while waiting for a better reward exhibi ted greater self regulatory ability. Based on this finding, Alberts et al. (2008) hypothesized that attention control could counteract ego depletion. Specifically, the experimenters examined how shifting attention away from the self regulatory task perform ed could influence physical stamina. A baseline measurement of time spent holding 1.5 kg weights aligned with the shoulders was recorded before
32 instating a 3 min distracted or a non distracted state. This first measurement of physical stamina was used as t he main self regulatory task to create a depleted state. Afterward, those in the distracted state performed a counting task, whereas the non distracted participants were instructed to focus entirely on the muscles and tendons used to hold up the weights. A third, control group did not receive any instruction before performing the physical stamina task again after 3 min. The results indicate that participants who had performed a distraction task performed significantly better on the second measure of physica l stamina in comparison to those in the non distracted condition. These findings illustrate that shifting attention away from, rather than concentrating on, the self regulatory task at hand can improve self regulatory performance, such that it counteracts the occurrence of ego depletion. Priming. Alberts, Martijn, Greb, Merckelbach, and de Vries (2007) suggested that performance on subsequent self regulatory tasks within the dual task paradigm may be boosted when participants were primed to persist on them. The researchers hypothesized that certain automatic processes could counteract the occurrence of ego depletion. Specifically, priming participants with task persistence prior to performing a self regulatory task in which task persistence would be required was expected to unconsciously circumvent the occurrence of ego depletion. Thus, a prime was provided between self regulatory tasks to examine whether depleted participants could replenish self regulatory resources. First, participants performed a self reg ulatory task requiring them to complete easy (control) or difficult labyrinths (depletion). Next, persistence was primed by having participants unscramble sentences that were all either related or unrelated to persistence and self control. A
33 dependent meas ure of physical stamina was used, thus requiring a baseline measurement in a pre and post test design; physical stamina was gauged by using a handgrip exercise, in which participants had to keep a coin in place between the tongs of a handgrip. As was expec ted, not only did the depletion condition predict poor performance on the second handgrip measurement, but those who received the persistence prime (versus a neutral prime) performed significantly better on the second handgrip measurement. Additionally, de pletion participants who received the persistence prime exhibited little change between the pre and post handgrip measurements. These results were replicated similarly in the second experiment, thus indicating that priming for task persistence increased persistence on a dependent measure of self regulatory ability. This study contributes to the body of literature examining factors that counteract ego depletion, such that when primed to perform an automatic process (i.e., persistence), performance is bette r. Similarly, Martijn, Alberts, Merchelbach, Havermans, Huijts and de Vries (2007) hypothesized that priming participants with an exemplar illustrating perseverance after self regulating would counteract the occurrence of ego depletion. Time measurements o f handgrip stamina while keeping a match in place between the tongs of the handgrip were recorded in a pre and post test design. All participants completed a series of labyrinths, although task difficulty varied so that the control group worked on easy la byrinths, whereas the depletion group worked on difficult labyrinths. Afterward, both groups were divided so that half of the control group and half of the depletion group were primed with a person exemplar that illustrated perseverance; these participants sorted through the text to pick out five out of context
34 words within a series of newspaper clippings regarding an Olympic runner who had succeeded after many trials and setbacks. The other halves of the control group and the depletion group performed the same word sorting task with a neutral prime, which included texts about the International Olympic Committee (IOC). Afterward, a second measure of handgrip stamina was recorded using the same method as the pretest handgrip recording. The results indicated that depleted participants who were primed with a perseverance exemplar maintained handgrip stamina, whereas depleted participants primed with a neutral exemplar had significantly lower handgrip times. Conversely, control participants who were primed with a perseverance exemplar performed significantly worse on the second handgrip stamina measurement in comparison to control group participants who experienced a neutral prime. This finding illustrates that perseverance priming enacts different results depend ing on whether or not one is ego depleted. Thus, using a perseverant person exemplar strengthens self regulatory ability only if self regulatory ability is impaired due to ego depletion. The researchers provided one possible explanation as to why this assi milation occurs in those that are ego depleted is because the ability to compare oneself with the exemplar may be impaired due to the lack of self regulatory resources. This aligns with previous research on contrasting priming results due to social compari son. Thus, the researchers suggest that those who are not ego depleted may compare themselves more intensively than those who are ego depleted, resulting in different effects of exemplar priming.
35 Autonomy. As mentioned earlier, the use of self control can be tied to the use of free will, which influences moral behaviors and decisions (Baumeister et al., 2007). Previous research indicates that forcing participants to adopt a counter attitudinal stance results in ego depletion ( Baumeister et al., 1998). Howev er, Moller et al. (2006) hypothesized that this does not occur when choice is autonomous. Through a series of experiments, Moller et al. (2006) conceptually replicated the method from Study 2 of the Baumeister et al. (1998) study by adding a group that cou ld make an autonomous decision, while the other groups made decisions that were either controlled or assigned. In the first stage of Experiment 1, participants were told to create a persuasive speech arguing whether or not psychology should be taught in hi gh school. Participants in the autonomous choice and controlled choice groups were assigned which side to write about by having two participants at a time; one participant chose their topic (autonomous choice), leaving the other participant stuck with the leftover choice (controlled choice). Participants in the no choice condition were not given the chance to make a decision; instead, the experimenter assigned participants to a specific topic. This same method of controlled, autonomous, and no choice condit ions was used in Experiments 2 3, where instead of creating a persuasive speech, participants were given a list of potential activities they could perform during the experiment. Afterward, persistence at a self regulatory task, including working at retraci ng an unsolvable geometric shape without lifting the pen from the paper (Experiment 1), searching for differences between two different, dense matrices of letters and symbols while holding their non dominant hands above their
36 heads (Experiment 2), or perfo rming the letter e task while holding the spacebar on a computer keyboard down (Experiment 3). In all experiments, Moller et al. (2006) found that participants in the autonomous choice group persisted for significantly longer than the controlled choice gr oup. Additionally, participants in the autonomous choice group made significantly more attempts to solve the geometric task in Experiment 1 and were more accurate on the letter e task in Experiment 3 than the controlled choice group. Finally, the autonomou s group resembled the no choice group in persistence, number of attempts, and accuracy on all dependent measures, whereas the controlled choice group was significantly lower. Together, these findings indicate that forced choice situations are ego depleting whereas making a decision autonomously is not. Thus the use of autonomy in decision making counteracts the occurrence of ego depletion. Self Affirmation. Schmeichel and Vohs (2009) hypothesized that reaffirming the self would replenish depleted resources thus counteracting ego depletion. In a series of experiments, participants performed a self affirmation task between self regulatory tasks. The self regulatory task varied in each experiment to include a task in which participants had to write a story wi thout the use of the letters a or n (Experiment 1), or a thought suppression task in which experimental participants had to suppress the urge to look at subtitles in a movie (Experiment 2). The self affirmation manipulations included reaffirming values by ranking them according to personal importance, whereas those in the control conditions ranked the values according to another person's point of view (Experiment 1 2). As a dependent measure, persistence on the cold presser task, in which participants subme rge their
37 hands in a container of ice water for as long as possible, was recorded (Experiment 1); additionally, persistence at solving a difficult puzzle was measured (Experiment 2). The results indicated that a difference between the depleted and non depl eted groups was only observed when depleted participants did not perform the self affirmation task. Thus, self affirmation was found to counteract ego depletion (Schmeichel & Vohs, 2009). Mood. Tice et al. (2007) hypothesized that boosted mood or positiv e affect may mediate ego depletion. In order to do so, four experiments were conducted using different manipulations to examine whether positive emotion after self regulating could influence self regulatory abilities on a second task. First a state of ego depletion was created using a variety of different methods. In Experiment 1, all participants wrote down every thought that occurred to them, although those in the ego depletion group were instructed not to think of a white bear while performing this task. Similarly, in Experiment 2, participants performed the letter e task, in which they begin by crossing out the letter e on a page of meaningless text; however, those in the depletion group must break this newly formed habit by crossing out the letter e onl y when it is adjacent to or two letters away from a vowel. Finally, Experiment 3 required participants to perform a difficult handgrip task and Experiment 4 required participants who had fasted for 3 hr to eat radishes while sitting in front of a plate of freshly baked cookies. Afterward, different methods were utilized to manipulate mood. In Experiments 1 2, this was done by giving participants either a small bag of candy as a thank you gift (to create positive affect) or a participation receipt (to create neutral affect). As a manipulation check, participants rated their arousal and
38 emotional states (e.g., tired, happy, worn out, excited). In Experiments 3 4, mood was manipulated by having participants watch different movie clips, including a funny movie c lip and a neutral video clip; in Experiment 3, however, a sad mood component was added by having a third group of participants watch a sad video clip. After the mood manipulation, performance on a self regulatory task was measured. In Experiment 1, the num ber of 1 oz cups of Kool Aid and vinegar mixtures consumed was measured. In Experiment 2, the amount of time spent playing a poorly constructed game in which participants have to balance two metal rods while rolling a ball uphill was measured. In Experimen t 3, a second measure of handgrip stamina was recorded as the dependent measure, whereas in Experiment 4, persistence at an unsolvable task was measured. The results were similar across all experiments, such that positive affect counteracted ego depletion. Thus, ego depleted participants who had experienced a mood boost due to positive mood manipulations (and not sad mood manipulations as in Experiment 3) exhibited similar performance to control groups. In Experiment 1, ego depleted participants who receive d the gift drank on average about as many ounces of the mixture as control group participants who did not receive a gift. Results from Experiment 2 illustrate that ego depleted participants who had received a gift did not exhibit detriments in time spent p laying the poorly constructed game in comparison to ego depletion participants who did not receive a gift. Additionally, in Experiment 3, participants who watched the neutral and sad movie clips gave up more quickly on the second handgrip task, whereas par ticipants who watched the positive clip performed about the same as in the first trial. Similarly, persistence on an
39 unsolvable task in Experiment 4 revealed a significant difference between the two groups, such that participants who watched the positive movie clip worked for significantly longer than participants who watched the neutral movie. The findings from these four experiments indicate that positive mood consistently counteracts ego depletion, regardless of the mood enhancing behavior or experience The various findings regarding counteracting ego depletion indicate that the self can be replenished when depleted. Additionally, this can occur using a variety of different methods to boost the self through persistence (Alberts et al., 2008; Martijn et al., 2007) or mood (Tice et al., 2007), or the affirmation of values (Schmeichel & Vohs, 2009). These findings illustrate that self control is a limited resource that can be replenished for optimal use. Still, it is necessary to examine the actual, finite resource that depletes duration self regulation. Self Control as a Finite Resource As mentioned earlier, ego depletion research has indicated that self regulatory ability functions similarly to a muscle; the strength model predicts that self control is a finite resource (Muraven et al., 1998). Specifically, this resource depletes with exertion. Although extensive research has been performed on sources of ego depletion, areas impacted by ego depletion, and factors that counteract this effect, the question o f what actually fuels this ability still needs to be answered. Gailliot and Baumeister (2007) assert that the success of the dual task paradigm in supporting the strength model of self control indicates that self control is a finite source. This is because it illustrates how engaging in self control consumes or depletes self regulatory energy on a subsequent task. More specifically, the
40 researchers suggest that the finite energy source that depletes during self regulation might be blood glucose. This implie s that ego depletion occurs because blood glucose is metabolized during self regulation; the logic behind this hypothesis is that normal brain functions (i.e., certain cognitive functions and reasoning) rely on blood glucose and, when there is an inadequat e amount of blood glucose, these functions are impaired (Benton et al., 1996). Moreover, lower levels of blood glucose have been cited as predicting poor performance on certain behaviors that require self control, including regulating emotions, acts of vio lence and aggression, attention control, coping with stress, and impulsivity (Gailliot & Baumeister, 2007). Based on these connections to self regulatory performance, Gailliot and Baumeister (2007) suggest that inadequate levels of blood glucose should be present during ego depletion. In order to fully examine whether this is true, it is necessary to illustrate that lower levels of blood glucose are present when self regulatory ability is impaired, that lower levels of blood glucose predict self regulatory performance, and that the presence of blood glucose counteracts the occurrence of ego depletion. While there are presumably a multitude of factors contributing to self regulatory performance, researching blood glucose as a potential predictor within the eg o depletion paradigm is an effective way of shedding light on what the current body of literature broadly defines as a finite source of self regulatory energy. Gaillot et al. (2007) explored the caloric expense of self regulation by examining whether blood glucose levels may contribute to ego depletion in a series of six studies. In Study 1, participants fasted for three hours prior to taking blood samples before and after completing a self regulatory task. Those in the depletion
41 group refrained from lookin g at words at the bottom corner of the screen while watching a 6 min video, while participants in the control group watched the same video without having to exercise restraint. The data revealed that blood glucose was significantly lower in the group that had to exert self control. Study 2 expands on these findings, in that researchers found that errors on a subsequent self regulatory task increased when participants were given a beverage with artificial sweetener as opposed to a beverage sweetened with sug ar. Thus, replenishing blood glucose levels with real sugar resulted in better self regulation. Together, the findings from Studies 1 2 provide evidence that blood glucose depletes during self regulation and that self regulation improves when blood glucose is replenished (e.g., by drinking something sweetened with sugar). These findings confirm that something actually depletes when we self regulate blood glucose. Study 3 indicated that lower levels of blood glucose predicted poorer performance on a second s elf regulatory task when blood glucose was measured before and after refraining from looking at words in a video. Similarly, Studies 4 6 employed different self regulatory tasks than Studies 1 3 to show that low blood glucose predicts less perseverance on an unsolvable task afterward, regardless of the self regulatory task. Together, these findings provide strong evidence that blood glucose affects self regulatory ability. Specifically, the use of different manipulations in Studies 4 6 while producing simil ar results further makes the case that depleted blood glucose after self regulating is not task specific. These findings confirm that something actually depletes during self regulation, and that when this resource is replenished, self regulatory ability is no longer impaired.
42 Since the discovery that blood glucose depletes during self regulation, several studies have conceptually replicated prior studies to include a blood glucose manipulation. The success of doing so strengthens the validity of recent rese arch findings relating blood glucose to the ego depletion phenomenon. DeWall, Deckman, Gailliot, and Bushman (2011) tested the impact of blood glucose on aggression regulation, an area that was previously researched to find that ego depletion was associate d with significantly greater aggression on a variety of tasks ( DeWall et al., 2007). Specifically, previous findings indicates that refraining from acts of aggression requires self monitoring; when the self's resources are depleted, the ability to refrain from aggression also depletes (DeWall et al., 2007). Based on recent blood glucose findings relating to ego depletion, DeWall et al. (2011) hypothesized that replenished blood glucose would similarly be linked to lower aggression. Thus, in alignment with p revious blood glucose research, participants received a drink that was sweetened with either sugar or artificial sweetener before a measure of aggressive behavior. The dependent measure required participants to play a rigged game in which they believed the y were competing against a partner to respond the fastest; before each trial, participants determined how long the blast of white noise would be and the volume of the blast of white noise their partner would receive if he or she lost. The results indicated that participants who received a drink with real sugar behaved significantly less aggressively toward their partners in comparison to those who received a drink with artificial sweetener. These findings illustrate that blood glucose affects the regulation of aggressive behavior, such that increasing blood glucose levels results in a decrease in aggressive behavior.
43 Similarly, adding to previous findings regarding how self regulatory resources deplete depending on social orientation (Balliet & Joireman, 201 0), DeWall, Baumeister, Gailliot, and Maner (2008) hypothesized that blood glucose would counteract the occurrence of ego depletion in a social situation. Specifically, the researchers suggested that depleted participants would exhibit fewer prosocial, hel ping behaviors (which is consistent with previous research); however, this should only occur when blood glucose levels are not restored by the ingestion of a sugary drink. In order to test this, researchers gave participants a 14 oz drink made with either sugar or artificial sweetener (a placebo) before an attention control task. All participants watched a video of a woman talking; the depletion group was instructed to ignore words scrolling across the bottom of the screen and to refocus on the woman talkin g if they looked at the words, whereas the control group watched the video without further instruction. Afterward, participants listened to a radio broadcast in which a girl talked about how her parents died and she would need to drop out of college to rai se her younger siblings unless she could find some volunteers to help out or raise some money. Participants received a bogus debriefing, followed by a request from another professor that all participants that listened to the radio broadcast fill out some i nformation regarding volunteering; on this sheet, participants indicated the number of hours they were willing to dedicate to helping out the young woman through volunteer work. The number of hours willing to volunteer was compared across groups to find th at those in the depletion group volunteered significantly fewer hours than the control group only when blood glucose was not replenished by the ingestion of a sugary drink. These results build on previous ego depletion research by
44 applying recent blood glu cose findings to a previously researched area, prosocial behavior. The ability to apply Gailliot et al.'s (2007) blood glucose findings to previously researched domains within the ego depletion literature confirms the depiction of ego depletion in terms of the strength model of self control (Muraven et al., 1998). These findings strengthen the argument that self control is a limited resource and indicate that something actually depletes when self control is exerted. The Current Study The phenomenon of ego d epletion has been widely accepted as the result of self regulation exertion. As mentioned earlier, these findings require the use of the dual task paradigm (Hagger et al., 2010), in which participants use self control, and then self regulatory ability is m easured on a subsequent, non related task. This paradigm supports the strength model of self control, which dictates that the use of self control temporarily depletes self regulatory ability, just as a muscle becomes fatigued after a work out (Muraven et a l., 1998). Still, it is important to pay attention to factors that might influence or be influenced by the occurrence of ego depletion. Recent findings that blood glucose depletes during self regulation provide additional support to the theory that self co ntrol is a finite resource (Gailliot et al., 2007). Moreover, replications of previous ego depletion studies to include a blood glucose manipulation in the research have strengthened the hypothesis that using self control metabolizes blood glucose. Specifi cally, DeWall et al. (2008) found that replenishing blood glucose levels increased prosocial, helping behavior in ego depleted participants. DeWall et al. (2011) found that replenishing blood glucose with a sugary
45 drink decreased aggressive behavior, indic ating that it strengthened or replenished the ability to regulate emotions. These findings provide evidential support to a central theme in ego depletion research that the self is a finite resource. Since blood glucose depletes during self regulation and c an replenish self regulatory ability when ego depleted, it becomes necessary to look at previous ego depletion methodology more critically. While the strength model of self control indicates that ego depletion results from self control exertion because of the depletion of a finite resource (Muraven et al., 1998), the literature does not indicate whether the amount of time spent self regulating affects this occurrence. In review of previous ego depletion research, the amount of time spent performing a self r egulatory task is not often mentioned in previous methods. Moreover, when the duration of self regulation has been mentioned, the different self regulatory tasks utilized have varied between 3 min (Muraven et al., 1998) to longer times, like 10 min (Baumei ster et al., 1998). In extreme cases, the self regulatory tasks have lasted for 30 min (Finkel et al., 2006). Still, whether time plays a role in the occurrence of ego depletion has not yet been researched. Specifically, it is unclear whether there is a th reshold at which one becomes ego depleted, but under which ego depletion does not occur. If blood glucose is consumed during self regulation, then presumably the amount of time spent self regulating will impact the amount of this energy source that gets de pleted during self regulation. Looking at a threshold at which one becomes ego depleted is important in order to understand the real world impact of self regulation in everyday tasks.
46 The present study tried to access this gap in the research by exploring the assertion that ego depletion follows as a result of self regulation. With this in mind, the present study sought to answer the question, how does the amount of time spent self regulating affect the occurrence of ego depletion? To answer this question, it was necessary to manipulate the duration of the first self regulatory task in the dual task paradigm. Thus, two experimental groups performed the letter e' task, in which they had to follow a challenging set of rules while examining text (Baumeister e t al., 1998); however, one group self regulated for a short period of time and one group self regulated for a long period of time. Afterward, self regulatory ability was measured by performance on the Stroop (1935) task, which required participants to self monitor when indicating the color of ink that words are presented in. In addition, it was necessary to compare self regulatory performance in both experimental groups with a control group that only self regulates on the dependent measure. The following wa s hypothesized: 1. If the duration of the task affects the occurrence of ego depletion, then the long duration group should perform significantly worse on the Stroop task than both the short duration group and the control group. 2. If the duration of the task do es not affect the occurrence of ego depletion, then the long duration and short duration groups should perform significantly worse on the Stroop task in comparison to the control group. Duration of self regulation is an important area to address, in that i t will indicate whether there is a threshold when self regulating at which ego depletion occurs. Researching the duration of self regulation could potentially shed light on appropriate
47 methods to use in future ego depletion studies, in terms of how long it is necessary to have participants self regulate in order to produce the desired effect. Additionally, these findings will indicate whether or not everyday acts of self control will always result in ego depletion. Method Participants Sixty participants ( 23 males, 34 females, 3 other) over the age of 18 ( M = 20.167, SD = 1.317) were recruited. Students from a small, liberal arts college in southwest Florida were either approached in public areas on campus or responded to an online forum to participate in a n experiment regarding performance on recognition tasks. All participants met with the experimenter alone to complete the self regulatory tasks. Participants were randomly assigned to one of three groups, so that one third of the participants were in the c ontrol group, one third of the participants were in the short duration group, and one third of the participants were in the long duration group. Materials Letter e Task. As in previous ego depletion research (Baumeister et al., 1998), all participants per formed the letter e task, in which they received five stapled pages of text copied from a statistics textbook. A set of instructions was attached to the top of each pack of text, which varied depending on the group. The rules varied so that the experimenta l groups had to use self control on this task, but the control group did not. Thus, the control group performed an easy task, where they simply crossed out the letter e every time it occurred in the text on the following pages. In contrast, the
48 experimenta l groups were told to cross out the letter e only when it was not adjacent to a vowel or one extra letter away from another vowel (Baumeister et al., 1998). The participants were reminded that this rule applied even if the adjacent vowel was not within the same word. For the experimental groups, the photocopy of the text was lightened for the experimental groups, making it more difficult to see; in contrast, those in the control group received a legible copy of the same text. A copy of the legible text that the control group received is attached in Appendix A; additionally, a copy of the lightened version of the text that both experimental groups received is attached in Appendix B. Positive Affect Negative Affect Schedule (PANAS). Mood was measured using th e Watson, Clark, and Tellegen's (1998) PANAS prior to recording the dependent measure. The PANAS is a 20 item self report questionnaire that measures positive and negative mood. Using a 5 point Likert scale (where 1 = Very Slightly or Not at All and 5 = Ex tremely), participants rated the level to which 10 positive emotions (e.g., interested, excited) and 10 negative emotions (e.g., distressed, upset) described their current internal state at the moment. Based on responses to both 10 item scales, participant s receive a summed score for positive affect, as well as a summed score for negative affect. Thus, participants receive two scores, in which it is possible to receive a minimum score of 10 and a maximum score of 50 for both the positive scale and the negat ive scale. A copy of the PANAS is provided in Appendix C. Color Stroop Task. Using SuperLab 4.0 a Color Stroop task (Stroop, 1935) was set up in which participants responded to specific word stimuli on a computer
49 screen, exactly as in previous research ( Inzlicht & Kang, 2010). The word stimuli were the words red and green ; however, the color of the words varied so that it was either in red or green letters Each word stimulus was presented randomly, one at a time. Before each word stimulus was presented, participants saw a fixation cross for 500 ms; afterward, the stimulus word appeared for 200 ms and participants were given 800 ms to respond before a new fixation cross appeared (Inzlicht & Kang, 2010). To respond, participants hit the right key on the res ponse pad if the word was written in red letters and the left key on the response pad if it was written in green letters. There were 10 blocks of 48 trials, including 32 congruent words (i.e., the color of the letters matched the semantic meaning of the wo rd) and 16 incongruent words (i.e., the color of the letters did not match the semantic meaning of the word). At the end of each block, the instructions for the task appeared again and participants could begin the task by pressing one of the keys on the re sponse pad. Procedure All participants were told that they were participating in a study examining performance on perceptual, recognition tasks. All interactions with participants were scripted. To begin, participants performed the letter e task exactly a s it was performed in the original Baumeister et al. (1998) study, in which participants are given copies of text from a statistics book and instructed to cross out all cases of the letter e but the experimental group had to follow a challenging set of gu idelines while doing so. The logic behind the letter e task is that it requires participants in the experimental group to self monitor their decisions when performing the task. Additionally, the text
50 provided to the experimental group was very poorly copie d and hard to see, requiring self regulation to monitor responses on a difficult task. The amount of time spent performing the letter e task was manipulated, so that the short duration group performed the task for 1 min, whereas the long duration group and the control group performed the task for 6.5 min. It was determined that a control group that would perform the task for 1 min was not necessary, as the control participants should not be depleted after any period of time, whether short or long in duratio n. Because 6.5 min was one of the longer periods of time mentioned within previous research used for any self regulatory task, it was chosen as an acceptable time for the long duration group to perform the task (see Neshat Doost et al., 2008; DeWall et al. 2007). While the longest self regulatory task used in previous research was 30 min (Finkel et al., 2006), this strays from the norm, which ranges from 3 min (Muraven et al., 1998) to 10 min (Baumeister et al., 1998). Thus, 6.5 minutes is more appropriate for the purpose of the current study because it is a longer time within the range of most ego depletion research where time was reported. Finally, 1 min was decided upon as the designated amount of time for the short duration group because it fell below t he shortest amount of time (3 min) listed in previous research for any type of first self regulatory task within the dual task paradigm (Baumeister et al., 1998). After completing the letter e task, the PANAS (Watson et al., 1988) was distributed as a mani pulation check, to ensure that mood did not influence any findings. The PANAS has been used frequently in previous ego depletion research as an effective way of controlling for mood as a confounding variable (Inzlicht & Kang, 2010; Balliet & Joireman, 2010 ; DeWall et al., 2010; Fischer et al., 2007; Gailliot et
51 al., 2007; Schmeichel & Vohs, 2009); Schmeichel et al., 2003; Vohs & Schmeichel, 2003). For the purposes of the present study, using the PANAS is necessary because those in the long duration group pe rformed the self regulatory task for such a long period of time, it was important to ensure that an affective component (i.e., negative mood) did not contribute to a lack of persistence on the dependent measure. Finally, performance on the Color Stroop tas k (Stroop, 1935) was measured. The Color Stroop task was set up exactly as in a previous study (Inzlicht & Kang, 2010), so that participants indicated the colors of word stimuli presented on a computer screen. Similarly, Stroop performance was evaluated ex actly as in the Inzlicht and Kang (2010) study, in which the mean response time for congruent trials was subtracted from the mean of incongruent trials for correct responses only. Additionally, Stroop errors (i.e., when a participant misidentified the colo r of letters) were measured. After completing the Stroop task, participants completed a 2 item demographic questionnaire asking for their age and gender. Upon completion of the demographic questionnaire, participants were debriefed and thanked for their pa rticipation. As part of the debriefing process, participants were asked not to discuss the experiment with other students until the end of the semester, as this could potentially skew the data. All participants received a $5 Starbucks gift card as compensa tion for their time.
52 Results Mood Participants' self control ability as determined by the Stroop task was measured following self control exertion (or lack thereof) on the letter e task. Thus, mood should not affect performance on the Stroop task. As a manipulation check, a one way ANOVA was performed to ensure that positive mood did not differ across conditions. The analysis revealed no significant difference between the control ( M = 23.350 SD = 6.923 ), short duration ( M = 26.800 SD = 8.501 ), and long duration ( M = 21.300 SD = 6.967 ) conditions on positive mood, F (2, 57) = 2.750, p = 0.073. Additionally, a one way ANOVA was performed to ensure that negative mood did not differ across conditions. This was determined to be an important measure, as the nature of the tasks completed was rather challenging. The analysis revealed that there was no significant difference between the control ( M = 13.450 SD = 2.762 ), short duration ( M = 15.350 SD = 6.158 ), and long duration conditions ( M = 14.300 SD = 4.169 ) on negative mood, F (2, 57) = 0.860, p = 0.427. Together, these results indicated that neither positive nor negative mood scores varied significantly across conditions. Notably, however, positive mood scores were almost two times higher than negative moo d scores. Stroop Performance As mentioned in the hypotheses, the duration of self regulatory tasks should either affect or not affect the occurrence of ego depletion. Specifically, if the duration of the task affects the occurrence of ego depletion, it was anticipated that Stroop performance would 1) be similar in the control group and the short duration group
53 and 2) be significantly worse in the long duration group. In contrast, if the duration of the task does not affect the occurrence of ego depletion, i t was anticipated that Stroop performance would be significantly worse in both long duration and short duration groups in comparison to the control group. To test these hypotheses, a one way ANOVA examined Stroop interference across conditions. Stroop inte rference was calculated by subtracting the mean correct reaction times for congruent trials from the mean correct reaction times for incongruent trials. The results revealed no difference between the control ( M = 19.874, SD = 25.281), short duration ( M = 2 2.694, SD = 17.578), and long duration groups ( M = 23.482, SD = 17.878) in Stroop interference, F (2, 57) = 0.170, p = 0.844. Additionally, a one way ANOVA examining Stroop errors (i.e., incorrect responses) for both congruent and incongruent trials across conditions revealed no difference between the control ( M = 31.650, SD = 25.351), short duration ( M = 30.050, SD = 22.908), and long duration groups ( M = 41.250, SD = 31.291), F (2, 57) = 1.03, p = 0.365. Additionally, a one way ANOVA examining Stroop erro rs for congruent trials only across conditions revealed no difference between the control ( M = 17.00, SD = 15.488), short duration ( M = 13.800, SD = 14.674), and long duration groups ( M = 20.600, SD = 14.858), F (2, 57) = 1.03, p = 365. A one way ANOVA exa mining Stroop errors for incongruent trials only across conditions revealed no difference between the control ( M = 14.650, SD = 10.879), short duration ( M = 16.250, SD = 10.192), and long duration groups ( M = 20.650, SD = 17.536), F (2, 57) = 1.09, p = .34 2. These results indicated that all participants performed similarly to
54 one another on the Stroop task, and that these errors did not differ on congruent or incongruent trials. Additional Analysis Because the Stroop task was very long (about 12 min), it was suspected that participants stopped applying themselves or responding as the task wore on. This was determined to be true based on an examination of the data; performance on the later blocks of trials did not match earlier trials (i.e., as the task wo re on, responses were sporadic and reaction times doubled). Thus, the reaction times from blocks 1 and 2 of the trials were analyzed using the same method, excluding all other trials. Afterward, a one way ANOVA examined Stroop Interference across condition s. The results revealed that, again, there was no difference in Stroop Interference between the control ( M = 22.213, SD = 39.971), short duration ( M = 28.509, SD = 28.004), and long duration ( M = 39.984, SD = 40.411) groups, F (2, 57) = 1.210, p = 0.305, a lthough the means were in the expected direction 1 Additionally, a one way ANOVA was performed examining Stroop errors across conditions for blocks 1 and 2 only. The results revealed that there was no difference between the control ( M = 6.650, SD = 4.923), short duration ( M = 6.300, SD = 4.118), and long duration ( M = 8.700, SD = 8.761) groups, F (2, 57) = .860, p = 0.431. A one way ANOVA examining block 1 and 2 Stroop errors for congruent trials only across conditions revealed no difference between the con trol ( M = 3.650, SD = 3.360), short duration ( M = 3.200, SD = 2.483), and long duration groups ( M = 3.200, SD = 3.365), F (2, 57) = .14, p = .869. A one 1 Because the Stroop interference means were in the expe cted directions for the ANOVA on blocks 1 and 2, a trend analysis using orthogonal polynomials was performed. The results revealed that, again, there was no significant difference in Stroop interference across conditions for blocks 1 and 2, F (2, 57) = 2.3 6, p = .130.
55 way ANOVA examining block 1 and 2 Stroop errors for incongruent trials only across conditions revealed no difference between the control ( M = 3.00, SD = 2.200), short duration ( M = 3.050, SD = 2.982), and long duration groups ( M = 5.450, SD = 7.366), F (2, 57) = 1.73, p = .186. This indicates that, again, participants Stroop performance (in terms of interfe rence and errors) did not differ on blocks 1 and 2. Finally, an additional one way ANOVA was performed on only Block 1 of Stroop Interference scores across conditions. The results illustrated that differences between the control ( M = 19.851, SD = 50.874), short duration ( M = 6.226, SD = 50.461), and long duration groups ( M = 46.097, SD = 63.938) were approaching significance, F (2, 57) = 2.67, p = .078. An independent samples t test revealed that there was no significant difference between the control and s hort duration groups on block 1 Stroop interference, t (38) = .85, p = .401. Additionally, there was no significant difference between the control and long duration group on block 1 Stroop interference, t (38) = 1.44, p = .159. However, there was a signific ant difference between the short duration and long duration groups on block 1 Stroop interference, t (38) = 2.19, p = .035. This indicates that Stroop interference on block 1 differed depending on the amount of time spent self regulating. Comparing Stroop errors across conditions in block 1, a one way ANOVA revealed no difference between the control ( M = 4.000, SD = 3.146), short duration ( M = 3.700, SD = 2.155), and long duration groups ( M = 5.000, SD = 4.834), F (2, 57) = .73, p = .485. A one way ANOVA ex amining block 1 Stroop errors for congruent trials only across conditions revealed no difference between the control ( M = 2.050, SD = 2.064), short duration ( M = 2.050, SD = 1.356), and long duration
56 groups ( M = 1.850, SD = 2.277), F (2, 57) = .07, p = .93 2. A one way ANOVA examining block 1 Stroop errors for incongruent trials only across conditions revealed no difference between the control ( M = 1.950, SD = 1.791), short duration ( M = 1.650, SD = 1.599), and long duration groups ( M = 3.150, SD = 3.469), F (2, 57) = 2.12, p = .129. This indicated that, again, all participants performed similarly to one another on the Stroop task in terms of correct and incorrect responses, although the differences in Stroop interference across conditions were larger and in the expected direction. Discussion As illustrated by the results, there was a trend in Stroop performance across conditions in the current study. Although the original results were non significant, when only the data from block 1, as well as blocks 1 and 2 combined, were retained for analysis, the means were still in the expected direction. The Stroop interference difference between the short duration group and the long duration group for block 1 was significant, such that the short duration group experien ced significantly less Stroop interference than the long duration group. This supports the hypothesis that self regulating for a longer period of time should result in worse Stroop performance; however, this was only apparent in the trials from block 1. Mo reover, the short duration group also performed better in comparison to the control group in terms of Stroop interference on block 1 (although this was not significant), suggesting that spending a short amount of time performing any task always resulted in better initial Stroop performance.
57 A normal distribution was observed for all Stroop interference analyses (i.e., scores across all blocks of trials, across blocks 1 and 2, as well as across block 1 only). Additionally, the manipulation check for mood ind icated that neither positive nor negative mood differed significantly between groups, although positive mood was higher than negative mood. Taking this into account, it is curious that the current study's original data did not initially display the ego dep letion effect. This raises potential issues with the tasks used in the current study, the replication of ego depletion findings, and the possible variability of duration across self regulatory tasks. Stroop Task Issues with the computerized Stroop task, a s replicated from Inzlicht and Kang's (2010) study, may have contributed to the lack of initial significant results. As mentioned previously, this task was designed so that the words red' and green' appeared very quickly in letters of varying colors, giv ing the participant only 800 ms to respond. During debriefing, the majority of participants reported that this task was very difficult, and many in the depletion conditions stated that it was much harder in comparison to the letter e task. Inzlicht and Kan g (2010) report, "We designed this task to be difficult for everyone," indicating that this feedback aligns with the original researchers' goals (p. 12). Thus, it is possible that the current study's difficulty replicating previous findings reflects a floo r effect, in which all participants performed poorly on the Stroop task. The data revealed that participants in the current study definitely performed differently in comparison to the original study, as can be seen in Table 1. The intent
58 of Inzlicht and K ang's (2010) study was to compare self control ability between females who were coping with stereotype threat, a control group of females who faced stereotype threat while practicing a coping strategy, and an additional control group of male participants w ho did not face stereotype threat at all. The researchers found that females coping with stereotype threat performed worse in terms of Stroop interference in comparison to females who faced stereotype threat while using a coping strategy and males who did not face stereotype threat at all. In looking at the current study's means for Stroop interference across conditions, it is apparent that participants' interference scores are higher at times in comparison. Because the current study had 8 more participants than the Inzlicht and Kang (2010) study, this disparity is not due to dissimilar sample sizes. Over the course of 10 blocks of trials in the current study, the control group, the short duration group, and the long duration group exhibited scores that were slightly different from Inzlicht and Kang's (2010) participants'. Specifically, the control group scored higher in comparison to both of Inzlicht and Kang's (2010) two control groups, while the current study's depletion groups both scored lower than the p revious researchers'. When only blocks 1 and 2 were included in the current study's analysis, Stroop interference scores increased even more across conditions. The scores from blocks 1 and 2 in particular are much higher in comparison to those of Inzlicht and Kang's (2010) participants, indicating that each group in the current study performed worse overall in terms of Stroop interference. Finally, when only block 1 was observed, the control, short duration, and long duration groups were even more variable in comparison to Inzlicht and Kang's (2010), with a much lower score for the short duration group and an even
59 higher score for the long duration group. Notably, the current study's control group exhibited a mean that was slightly higher than Inzlicht and K ang's (2010) male control group. Notably, the significant differences that Inzlicht and Kang (2010) observed on the Stroop task between the depletion group and the two control groups were very small (with a difference between the depletion group and female control group of 15.480 ms, and a difference between the depletion group and the male control group of 13.100 ms ). In blocks 1 and 2 of the current study, there was a greater difference between the control group and the long duration group (17.771 ms), an d an almost equal difference to Inzlicht and Kang's (2010) between the short duration and long duration group (11.475 ms). These differences expanded when only block 1 was retained for analysis, with a difference between the control group and the long dura tion group of 26.246 ms and an even larger difference between the short duration and long duration groups (39.871 ms), yet statistical analysis revealed that these differences were significant when only the short duration and long duration groups were comp ared. One possible reason for this may be that participants in the current study performed differently overall. For example, to eliminate outliers, Inzlicht and Kang (2010) winsorized their data; this converts outliers to any chosen standard deviation. Whi le the number of standard deviations converted was not reported, Inzlicht and Kang (2010) did indicate that the range of Stroop interference scores had a low end of 10 ms and a high end of 100 ms. Meanwhile, Stroop interference across all blocks of trials in the current study exhibited a low end of 13.034 and a high end of 73.791.
60 This indicates that, over the course of all trials, participants in the current study experienced less Stroop interference than Inzlicht and Kang's (2010) participants, and some even performed better. When only the first two blocks of trials were used for analysis, the data ranged from a low end of 41.376 to a high end of 124.666. Additionally, performance on the first block of trials ranged from a low end of 149.667 to a high end of 195.693. This indicates that, on the first block of trials, as well as the first two blocks of trials, there were participants who performed worse in terms of Stroop interference (i.e., they took much longer to respond to incongruent trials in compa rison to congruent trials) than Inzlicht and Kang's (2010) participants, as well as participants who performed much better (i.e., they responded much faster to incongruent trials than congruent trials). Based on these findings, it is apparent that there wa s greater variability in Stroop interference across conditions in the current study in comparison to Inzlicht and Kang's (2010) study. Additionally, Inzlicht and Kang (2010) did not report Stroop errors across conditions in their ego depletion study; rathe r, they stated that these did not vary across conditions, and reported a mean of 5.82 errors. This means that, while participants' reaction times may have changed as they monitored their responses on congruent and incongruent trials, actual task performanc e (in terms of correct and incorrect responses) was not affected by the ego depletion effect. The current study also found no difference in errors across conditions; however, across all blocks of trials, the mean errors in the control the control ( M = 31.6 50, SD = 25.351), short duration ( M = 30.050, SD = 22.908), and long duration groups ( M = 41.250, SD = 31.291) were much higher Since Stroop Interference was calculated by looking at
61 correct responses only (meaning all reaction times for incorrect respons es were excluded), it is possible that Inzlicht and Kang's (2010) Stroop Interference results were not replicated here because participants in the current study made so many errors in comparison. This means that, when calculating Stroop Interference, Inzli cht and Kang (2010) excluded a much smaller portion of reaction times for incorrect responses in comparison to the current study. Only in Block 1 were Stroop errors about the same as the Inzlicht and Kang (2010) study; there were no differences between th e control ( M = 4.000, SD = 3.146), short duration ( M = 3.700, SD = 2.155), and long duration groups ( M = 5.000, SD = 4.834) in this respect. The similarity of Stroop errors between the original study and block 1 of the current study provide support for the fact that the differences between groups in terms of Stroop interference were approaching significance for this block of trials only. It also supports the significant differences in Stroop interference between the short duration and long duration groups. Overall, these observations about the Stroop task illustrate that the results from Inzlicht and Kang's (2010) study were not ultimately replicated for a variety of reasons; there was greater variability in scores, different Stroop Interference means, and a much higher number of errors in all analyses except for block 1. Replicability Given that Stroop performance did not match the Inzlicht and Kang (2010) study, the results from the current study illustrate an important methodological issue within ego depl etion research the inability to replicate the ego depletion effect. As cited by Hagger et al. (2010) in a meta analysis of ego depletion research, there is a
62 sample of ego depletion studies that, similarly, have reported no significant differences between groups, suggesting that ego depletion did not occur. Specifically, Hagger et al. (2010) mention that these studies produced non significant findings between control and depletion groups on a variety of tasks, including thought suppression, challenging coun ting tasks, and persistence. The researchers suggest that the failure of this group of ego depletion studies provide support that "the ego depletion effect is not unequivocal, and there are many variations across the literature" (p. 497). The current study is one such case in which participants in both depletion conditions still resembled the control group in self regulatory performance following a self control task. Although block 1 illustrated a significant difference between the short duration and long d uration groups, the results from the analysis of block 1 Stroop interference scores between the control, short duration, and long duration groups reflect this fragility overall. This is because there were large differences that were observed between groups but additional analysis still could not make these differences across all conditions significant. Task Variability One possible explanation for Stroop performance across groups may be that there are variations across different self regulatory tasks. Per haps there are inconsistencies throughout the literature because the necessary duration of self control expenditure (in order to observe the ego depletion effect) is task dependent. If this is the case, then it is possible that participants did not self re gulate for a long enough time on the letter e task in order to see an effect. For example, in Baumeister et al.'s (1998) study, participants were given one sheet of text copied from a statistics
63 textbook and asked to cross out the letter e either by follow ing an easy or a challenging set of rules. Participants finished this task when they had gotten to the end of the page; the amount of time spent working on the task was not measured (Baumeister et al., 1998, Experiment 4). In contrast, the duration of the letter e task for the current study was determined by choosing a median time that fit within the range of most self regulatory tasks used in previous research. As mentioned earlier, most of these times have ranged from 3 min ( Muraven et al., 1998) to 10 mi n (Baumeister et al., 1998, Experiment 3; Vohs & Heatherton, 2000), with some as long as 30 min (Finkel et al., 2006). Numerous self regulatory tasks fit within this range, and a full outline of self regulatory tasks with varying durations within the ego d epletion literature is depicted in Table 2. The duration of the task for the current study was chosen based on the assumption that all self regulatory tasks will be equally demanding; thus, the duration of any self regulatory task that successfully resulte d in ego depletion should apply to all. Upon looking at the packets of text that participants completed for the current study, however, it seems that 6.5 minutes was not long enough for most participants in the long duration group to complete the entire fi rst page of text. Rather, participants in the long duration group completed on average about one half to three quarters of a page in this amount of time, which is less than what Baumeister et al.'s (1998) participants completed. Meanwhile, control particip ants in the current study were able to complete a full page of text (following easy rules) within this time. Since participants in the long duration group did not complete a full page of text for the letter e task, it is possible that they did not self reg ulate for long enough to see an ego
64 depletion effect for this specific, self regulatory task. If there is a threshold, and if it takes a full page to observe the ego depletion effect, then the data from the current study contains two depletion groups who d id not cross the threshold and hence, equal the control group on Stroop performance across blocks. This raises the issue of whether or not duration varies across self regulatory tasks, and suggests that the threshold at which ego depletion occurs depends u pon the task. Future Research To ensure the replication of the ego depletion phenomenon in future research, it may be beneficial to lengthen the duration of the letter e task. Given that long duration participants only completed between one half and three quarters of a page of text in the 6.5 minutes provided, it would be best to extend the long duration group to a time that will allow participants to complete at least one full page of text. This may be done by allowing participants to complete the full pag e of text before stopping, or by setting a longer time for the task. Additionally, future research may also use different self control tasks aside from the letter e task using the current study's set up. Possible longer tasks that may be used in this set u p might include the longer tasks depicted in Table 2: thought control (Muraven et al., 1998), emotion suppression while watching a movie (Muraven et al., 1998; Baumeister et al., 1998; Fischer et al., 2007), completing challenging labyrinths (Martijn et al ., 2007), resisting tasty snack foods (Baumeister et al., 1998; Vohs & Heatherton, 2000; DeWall et al., 2007), and variations of the Stroop task (Neshat Doost et al., 2007), to name a few. Using these other tasks that have had longer durations in previous research may indicate whether the exact duration in order to produce an ego depletion effect varies across tasks.
65 Since there was only one control group, it is possible that adding a second control group that performs an easy task for a short period of tim e (in addition to a control group that performs an easy task for a long period of time) would be beneficial. Because the control group never uses self control, it was assumed that a control group that performs an easy task for a long period of time still s erves as a powerful comparison for depletion groups for the purposes of the current study. However, Block 1 analyses, wherein the short duration group experienced less Stroop interference ( M = 6.226, SD = 50.461) than the control group ( M = 19. 851, SD = 5 0.874), indicated that less time performing the letter e task, regardless of whether or not it required self control, affected initial Stroop performance. Thus, adding a short duration control group could provide a better comparison group within the curren t study's set up. As mentioned earlier, recent studies have linked blood glucose to the occurrence of ego depletion, in terms of counteracting ego depletion and boosting self regulatory performance (Gailliot et al., 2007; DeWall et al., 2011; DeWall et al. 2008). This suggests that controlling for blood glucose prior to participation is a necessity for future research. Still, numerous ego depletion studies, including those that were replicated in the current study (Baumeister et al., 1998; Inzlicht & Kang, 2010), have reported significant ego depletion findings without controlling for blood glucose (Muraven et al., 1998; Alberts et al., 2007; Alberts et al., 2008; Balliet & Joireman, 2010; Ciarocco et al., 2001; Clarkson et al., 2010; DeWall et al., 2007; D eWall et al., 2008; DeWall et al., 2010; Finkel et al., 2006; Fischer et al., 2007; Martijn et al., 2007; Moller et al., 2006; Neshat Doost et al., 2008; Schmeichel et al.,
66 2003; Schmeichel & Vohs, 2009; Tice et al., 2007; Unger & Stahlberg, 2011; Vohs & H eatherton, 2000; Vohs & Schmeichel, 2003). This indicates that controlling for blood glucose is not the sole solution to observe significant ego depletion findings, and that blood glucose likely did not interfere with the current study's results. However, it may be beneficial in future research to control for blood glucose, considering its powerful impact on self regulatory ability (Gailliot et al., 2007; DeWall et al., 2011; DeWall et al., 2008). General Discussion Despite issues with the Stroop task, rep licability, and task variation, the significant finding that the short duration group experienced significantly less Stroop interference than the long duration group on block 1 provides support for the idea that self regulating for longer periods of time n egatively impacts subsequent self control ability. However, it is unclear whether this is only an initial effect, as these results were not observed in Blocks 1 and 2, or across all blocks (although the means were in the expected direction). Additionally, the fact that the analysis of block 1 as a whole was only approaching significance makes it difficult to assume that the long duration group in the current study crossed the threshold to become ego depleted, because the difference in Stroop interference be tween the control and long duration groups was not significant. Ultimately, the current study's difficulty in replicating the ego depletion effect reinforces the methodological issue that the current study sought to address: controlling the duration of sel f regulatory tasks across ego depletion research is a necessity. There is a very clear paradigm (i.e., the dual task paradigm) that is used
67 throughout ego depletion research to induce and measure a depleted state; however, in order to replicate and build o nto previous findings, it is necessary to add a specific duration of self regulatory tasks to be used throughout research on the topic. The observation that the long duration group experienced more Stroop interference than the short duration group on block 1 reinforces that the duration of prior self regulatory tasks can affect subsequent self control performance. Thus, controlling the duration of tasks in future research is beneficial for practical purposes, in that thresholds may vary across self regulato ry tasks. Identifying specific durations for self regulatory tasks will strengthen the validity of the research on this phenomenon, and lead the way to more intricate research regarding self control, which is an important contributor to everyday activities
68 References Alberts, H. J. E. M., Martijn, C., Greb, J., Merckelbach, H., & de Vries, N. K. (2007). Carrying on or giving in: The role of automatic processes in overcoming ego depletion. British Journal of Social Psychology, 46 383 399. doi: 10.1348/014466606X130111 Alberts, H. J. E. M., Martijn, C., Nievelstein, F., Jansen, A., de Vries, N. K. (2008). Distracting the self: Shifting attention prevents ego depletion. Self and Identity, 7 322 334. doi: 10.1080/15298860801987583 Balliet, D., & J oireman, J. (2010). Ego depletion reduces proselfs' concern with the well being of others. Group Processes & Intergroup Relations, 13 227 239. Baumeister, R. F. (2002). Ego depletion and self control failure: An energy model of the self's executive functi on Self and Identity, 1 129 136. Baumeister, R. F., Bratlavsky, E., Muraven, M., & Tice, D. M. (1998). Ego depletion: Is the active self a limited resource? Journal of Personality and Social Psychology, 74 1252 1265. Baumeister, R. F., Gailliot, M., DeW all, C. N., & Oaten, M. (2006). Self regulation and personality: How interventions increase regulatory success, and how depletion moderates the effects of traits on behavior. Journal of Personality, 74 1773 1801. doi: 10.1111/j.1467 6494.2006.00428.x Bau meister, R. F., Schmeichel, B. J., DeWall, C. N., & Vohs, K. D. (2007). Is the conscious self a help, a hindrance, or an irrelevance to the creative process? Advances in Psychology Research, 53 137 152.
69 Baumeister, R. F. & Vohs, K. D. (2007). Self regulat ion, ego depletion, and motivation. Social and Personality Psychology Compass, 1 1 14. doi: 10.1111/j.1751 9004.2007.00001.x Baumeister, R. F., Vohs, K. D., & Tice, D. M. (2007). The strength model of self control. Current Directions in Psychological Scie nce, 16 351 355. Benton, D. (1990). The impact of increasing blood glucose on psychological functioning. Biological Psychology, 30 13 19. Ciarocco, N. J., Sommer, K. L., & Baumeister, R. F. (2001). Ostracism and ego depletion: The strains of silence. Per sonality and Social Psychology Bulletin, 27 1156 1163. Clarkson, J. J., Hirt, E. R., Jia, L., & Alexander, M. B. (2010). When perception is more than reality: The effects of perceived vs. actual resource depletion on self regulatory behavior. Journal of P ersonality and Social Psychology, 98 29 46. doi: 10.1037/a0017539 DeWall, C. N., Baumeister, R. F., Stillman, T. F., & Gailliot, M. T. (2007). Violence restrained: Effects of self regulatory capacity and its depletion on aggressive behavior. Journal of Ex perimental Social Psychology, 43, 62 76. doi: 10.1016/j.jesp.2005.12.005 DeWall, C. N., Baumeister, R. F., Gailliot, M. T., & Maner, J. K. (2008). Depletion makes the heart grow less helpful: Helping as a function of self regulatory energy and genetic rela tedness. Personality and Social Psychology Bulletin, 34 1653 1662. doi: 10.1177/0146167208323981
70 DeWall, C. N., Baumeister, R. F., Mead, N. L., & Vohs, K. D. (2010). How leaders self regulate their task performance: Evidence that power promotes diligence, depletion and disdain. Journal of Personality and Social Psychology Advance online publication. doi: 10.1037/a0020932 DeWall, C. N., Deckman, T., Gailliot, M. T., & Bushman, B. J. (2011). Sweetened blood cools hot tempers: Physiological self control and aggression. Aggressive Behavior, 37 73 80. doi: 10.1002/ab.20366 Fey, T. (2011). Bossypants New York, NY: Reagan Arthur Books. Finkel., E. J., Campbell, W. K., Brunell, A. B., Dalton, A. N., Scarbeck, S. J., & Chartrand, T. L. (2006). High maintenance in teraction: Inefficient social coordination impairs self regulation. Journal of Personality and Social Psychology, 91 456 475. doi: 10.1037/0022 35126.96.36.1996 Fischer, P., Greitemeyer, T., & Frey, D. (2007). Ego depletion and positive illusions: Does the c onstruction of positivity require regulatory resources? Personality and Social Psychology Bulletin, 33 1306 1321. doi: 10.1177/0146167207303025 Gailliot, M. T., Baumeister, R. F., Dewall, C. N., Maner, J. K., Plant, E. A., Tice, D.M., Brewer, L. M. & Schm eichel, B. J. (2007). Self control relies on glucose as a limited energy source: Willpower is more than an energy source. Journal of Personality and Social Psychology, 92 325 336. doi: 10.1037/0022 35188.8.131.525
71 Gailliot, M. T. & Baumeister, R. F. (2007). The physiology of willpower: Linking blood glucose to self control. Personality and Social Psychology Review, 11 303 327. doi: 10.1177/1088868307303030 Hagger, M. S., Wood, C. Stiff, C. Chatzisarantis, N. L. D. (2010). Ego depletion and the strength mode l of self: A meta analysis. Psychological Bulletin, 136 495 525. doi: 10.1037/a0019486 Inzlicht, M., & Kang, S. K. (2010). Stereotype threat spillover: How coping with threats to social identity affects aggression, eating, decision making, and attention. Journal of Personality and Social Psychology Advance online publication. doi: 10.1037/a0018951 Martijn, C., Alberts H. J. E. M., Merckelbach, H., Havermans, R., Huijts, A., & De Vries, N. K. (2007). Overcoming ego depletion: The influence of exemplar prim ing on self control performance. European Journal of Social Psychology, 37 231 238. doi: 10.1002/ejsp.350 Mischel, W., & Ebbesen, E. B. (1970). Attention in delay of gratification. Journal of Personality and Social Psychology, 16 329 337. Moller, A. C. Deci, E. L., & Ryan, R. M. (2006). Choice and ego depletion: The moderating role of autonomy. Personality and Social Psychology Bulletin, 32 1024 1036. doi: 10.1177/0146167206288008 Muraven, M., Tice, D. M., & Baumeister, R. F. (1998). Self control as a limited resource: Regulatory depletion patterns. Journal of Personality and Social Psychology, 74 774 789.
72 Muraven, M. & Baumeister, R. F. (2000). Self regulation and depletion of limited resources: Does self control resemble a muscle? Psychological Bull etin, 126 247 259. doi: 10.1037//0033 2909.126.2.247 Neshat Doost, H. M., Dalgleish, T., & Golden, A. J. (2008). Reduced specificity of emotional autobiographical memories following self regulation depletion. Emotion, 8 731 736. Schmeichel, B. J., Vohs, K. D., & Baumeister, R. F. (2003). Intellectual performance and ego depletion: Role of the self in logical reasoning and other information processing. Journal of Personality and Social Psychology, 85 33 46. doi: 10.1037/0022 35184.108.40.206 Schmeichel, B. J. & Vohs, K. D. (2009). Self affirmation and self control: Affirming core values counteracts ego depletion. Journal of Personality and Social Psychology, 96 770 782. doi: 10.1037/a0014635 Stroop, J. R. (1935). Studies of interference in serial verbal react ions. Journal of Experimental Psychology, 18, 643 662. Superlab (Version 4.0) [Computer software]. San Pedro, CA: Cedrus. Tice, D. M., Baumeister, R. F., Smueli, D., & Muraven, M. (2007). Restoring the self: Positive affect helps improve self regulation fo llowing ego depletion. Journal of Experimental Social Psychology, 43 379 384. doi:10.1016/j.jesp.2006.05.007 Unger, A., & Stahlberg, D. (2011). Ego depletion and risk behavior: Too exhausted to take a risk. Social Psychology, 42 28 38.
73 Vohs, K. D. & Heat herton, T. F. (2000). Self regulatory failure: A resource depletion approach. Psychological Science, 11 249 254. Vohs, K. D., & Schmeichel, B. J. (2003). Self regulation and the extended now: Controlling the self alters the subjective experience of time. Journal of Personality and Social Psychology, 85 217 230. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS Scales. Journal of Personality and Social Psychology, 47 10 63 1070.
74 Table 1 Interference in Inzlicht and Kang (2010)'s Original Study and the Current Study Stroop Interference (ms) 10 Blocks Blocks 1 2 Block 1 Groups Means (SD) Means (SD) Means (SD) Original Study Control (F) 12.91 (8.42) Control (M) 15.29 (13.78) Depletion (F) 28.39 (27.22) Current Study Control 19.87 (25.28) 22.21 (39.97) 19.85 (50.87) Short (1 min) 22.69 (17.58) 28.51 (28.00) 6.23 (50.46) Long (6.5 min) 23.48 (17.88) 39.98 (40.41) 46.10 (63.94)
75 Table 2 Duration of Self Regulatory Tasks Throughout Ego Depletion Literature Previous Self Regulatory Tasks Experiment Duration Task Type Muraven et al. (1998) 1 3 min Emotion regulation 2 6 min Thought control 3 18 min Emotion regulation 4 ? Writing task Baumeister et al. (1998) 1 5 min Resisting snacks 2 ? Adopting counter attitudinal perspective 3 10 min Emotion regulation 4 ? Letter e task Vohs & Heatherton (2000) 1 2 10 min Resisting snacks Balliet & Joireman (2010) 2 5 min Emotion regulation DeWall et al. (2010) 2,4 6 min Attention control 3 10 min Letter e task (habit breaking) Finkel et al. (2006) 1 3 min High maintenance social interaction
76 2 5 min Same as above 3 6 min Same as above 4 30 min Same as above Ciarocco et al. ( 2001) 1 2 3 min Ostraci sm Inzlicht & Kang (2010) ? Experiencing stereotype threat DeWall et al. (2007) 5 min Resisting snacks Fischer et al. (2007) 1, 2, 4 5 min Emotion regulation 3 ? Letter e task 5 ? Thought control S chmeichel et al. (2003) 1, 3 6 min Attention control 2 10 min Emotion regulation Unger & Stahlberg (2011) 1, 2 5 min Word task 3 4.35 min Emotion regulation Neshat Doost et al. (2008) 6.5 min Stroop task Vo hs & Schmeichel (2003) 1, 2 10 11 min Emotion regulation 3 4 min Feigning interest 4 6 min Thought control
77 Clarkson et al. (2010) 1, 2, 4 ? Letter e task 3 ? Thought control Alberts et al. (2008) Varied Physical stamina Alberts et al. (2007) ? Labyrinth task Martijn et al. (2007) 5 min Labyrinths Moller et al. (2006) 1 ? Adopt counter attitudinal perspective 2 3 ? Forced activities Schmeic hel & Vohs (2009) 1 5 min Letter monitoring task 2 5 min Attention control Tice et al. (2007) 1, 3 5 min Thought control 2 ? Letter e task (habit breaking) 4 10 12 min Resisting snacks Gailliot et al. (2007) 1, 3, 4, 7 6 min Attention control 2 5 min High maintenance social interaction 5 4 min Stroop task
78 6 2 min Emotion regulation 8 ? Coping 9 Varied Exam DeWall et al. (2008) 6 min Attention control
79 Appendix A Please cross out the letter e every time you come across it in the text on the following pages I will tell you when to stop. Thank you for your participation!
85 Appendix B Ple ase cross out the letter e when you come across it in the text on the following pages if it is not adjacent to a vowel or one letter away from a vowel. For example, you would not cross out the letter e in the word "vowel." I will tell you when to stop. Th ank you for your participation!
91 Appendix C Positive Affect Negative Affect Schedule (Watson et al., 1988) This scale consists of a number of words that describe different feelings and emotions. Read each item and then list t he number from the scale below next to each word. Indicate to what extent you feel this way right now, that is, at the present moment. 1. Interested 11. Irritable 2. Distressed 12. Alert 3. Excited 13. Ashamed 4. Upset 14. Inspired 5. Strong 15. Nervous 6. Guilty 16. Determined 7. Scared 17. Attentive 8. Hostile 18. Jittery 9. Enthusiastic 19. Active 10. Proud 20. Afraid 1 2 3 4 5 Very Slightly or A Little Moderately Quite a Bit Extremely Not at All