Table of Content
List of Figures
2. Theoretical Foundations.
3. Antecedents and Outcomes of Ego Depletion
3.1 Evidence of the effect
3.1.1 The state of ego depletion.
3.1.2 Ego depletion by making choices.
3.1.3 Effects of ego depletion..
3.1.4 Overcoming the effect
3.2 Concerns about the theory..
4.1 Managerial Implications.
4.2 Limitations and Future Research
Appendix A: Tables
Appendix B: Figures
Appendix C: Literature Review Tables
Appendix D: Comparative Literature Table
List of Figures
Figure 1: Ego depletion undermines dieting
Figure 2a + 2b: Ego depletion increases dishonesty
Figure 3a + 3b: Glucose counteracts ego depletion
Figure 4: Limited-resource theory versus nonlimited-resource theory
Figure 5a + 5b: Severe ego depletion by four self-control tasks
Figure 6: The process model of ego depletion
The ego depletion paradigm has been investigated in many experiments. There is a lot of evidence for the effect but somehow there are interestingly still doubts about it. Furthermore, the strength model is the selected explanation for the effect. It assumes that the capacity of self-regulation is based on a limited resource akin to strength or energy which implies a resource depletion after a self-control task – ego depletion. These findings could shed light on how self-regulatory success and failure are related and may be helpful to protect yourself from failures towards reaching long-term goals. This thesis provides a literature review over the antecedents of ego depletion and the outcomes of the effect. Moreover, the raising concerns are inspected. A brief discussion considers on managerial implications and future research.
Self-regulatory success is related to enabling people to delay gratification and thus to reach long-term goals (Hagger et al. 2010, p. 495; Muraven, Tice, and Baumeister 1998, p. 774). In contrast, self-regulatory failure is associated with smoking, crime, divorces, depression, obsessive thoughts and school underachievement (Muraven, Tice, and Baumeister 1998, p. 774). Responsible for such acts of volition? The Self (Schmeichel, Vohs, and Baumeister 2003, p. 33). Therefore, it is highly desirable to achieve a better understanding of how and when the self uses self-regulation and why it breaks down some times (Muraven, Tice, and Baumeister 1998, p. 774).
Over the past two decades, a growing body of social psychology research has examined how all these favourable or maladaptive outcomes are related (Hagger et al. 2010, p. 495, 2016, p. 547; Inzlicht and Schmeichel 2012, p. 450). Baumeister et al. (1998) concentrated on the “controlling aspect of the self” and suggested that all acts of volition draw on a limited resource (p. 1252). Making use of this resource for self-regulatory responses means to deplete it, which indicates to have less of it for the following task involving self-regulation (Baumeister et al. 1998, p. 1252; Muraven, Tice, and Baumeister 1998, p. 774). Muraven, Tice, and Baumeister (1998) proposed a strength model to explain the hypothesis of regulatory depletion (p. 774).
A large number of studies can be traced back to these antecedents and brought supporting evidence. However, in recent years, this paradigm has been questioned (Carter et al. 2015, p. 796; Hagger et al. 2016, p. 556; Inzlicht and Schmeichel 2012, p. 452 f.; Job, Dweck, and Walton 2010, p. 1692). This thesis provides an overview of how the research has progressed, from the roots to the present day. In the next section, the most important terms are defined and related. In a literature review, evidence for and against the underlying theories by visiting many empirical findings from different spheres is presented and linked to the theory. The thesis is completed by implications for marketers how to make use of the findings to boost sales. At the end, several limitations and possibilities for future research are exhibited.
2. Theoretical Foundations
To ensure a better understanding later, the most important terms are defined and explained beforehand.
Self-regulation is the capacity to override, interrupt, control and alter the responses of the human organism like one’s thoughts, emotions, urges and behaviour (Gailliot et al. 2007, p. 325; Muraven, Tice, and Baumeister 1998, p. 774). Self-regulation is involved in many tasks, from “getting out of bed” to maintaining attention in class or running a marathon (Vohs and Heatherton 2000, p. 249). Hence, self-regulation let people overcome the impulses and drives of everyday life (Hagger et al. 2010, p. 495). Otherwise, without self-regulation, “life would become a series of unconstrained impulsive actions” (Hagger et al. 2010, p. 495).
Self-control is interchangeably used with self-regulation and is defined similarly like above (Baumeister 2002, p. 670; Gailliot et al. 2007, p. 325; Hagger et al. 2010, p. 496). For some researchers, self-control is the conscious and effortful form of self-regulation to adjust to societal standards and prescriptive expectations (Gailliot and Baumeister 2007, p. 303). However, in this thesis the terms fulfil the same meaning like in most of the cited articles.
Willpower is the colloquial used word for self-control (Gailliot and Baumeister 2007, p. 304; Inzlicht and Schmeichel 2012, p. 450) and seemingly means regulatory strength (Muraven, Tice, and Baumeister 1998, p. 775). Moreover, from the concept of willpower the strength model was derived (Muraven, Baumeister, and Tice 1999, p. 447).
The dual-task paradigm is an experimental procedure (Hagger et al. 2010, p. 496). It is adopted in most empirical tests of the ego depletion effect using two unrelated self-control tasks. Participants are randomly allocated to an experimental group or a control group (Hagger et al. 2016, p. 547). In the first task, participants in the experimental group are manipulated by an exertion of self-control whereas participants allocated to the control group have to do a simpler task that should require at the most very little of self-control (p. 547). The second task follows subsequently for both groups and involves a measure of self-control performance (p. 496). Most studies proved whether valence of mood or arousal affected the results, but that was seldom the case (Muraven and Baumeister 2000, p. 253). It is noteworthy that participants were not told the exact purposes of the experiments, to avoid additional bias influencing the results. The experiments described in the literature review are mainly based on this methodology of examination.
The strength model is one of two limited resource models and entails that “the capacity for self-regulation is a limited resource” (Muraven, Tice, and Baumeister 1998, p. 775). Unlike the other limited resource model entailing constant capacity, the strength model implies that the available amount of this limited resource will decrease after regulating one response of the self (p. 785). Exerting self-control in one task leads to a diminished performance in a subsequent self-control task, even in quite different spheres. The assumption is that all tasks that require self-control draw on and share the same limited resource, akin to strength or energy (Baumeister et al. 1998, p. 1256; Hagger et al. 2010, p. 496 f.; Muraven, Tice, and Baumeister 1998, p. 775). This limited resource is used “for all acts of volition”, such as active choice, overriding responses, controlled processing, initiating behaviour and to resist temptations (Baumeister et al. 1998, p. 1253). There are individual differences in the extent of the self-regulatory resources (Muraven and Baumeister 2000, p. 248). The metaphor of the strength model claims that the capacity of self-control exertion is analogous to a physical muscle (Gailliot and Baumeister 2007, p. 304; Muraven and Baumeister 2000, p. 254; Sjåstad and Baumeister 2018, p. 128; Wheeler, Briñol, and Hermann 2007, p. 150). Crucially, self-control strength recovers more slowly than it is used (Muraven and Baumeister 2000, p. 249). In this view, a total exhaustion would mean that this inner resource is unavailable then until it is replenished, and self-regulatory failures are likely to occur (p. 250), which leads to the next term and core of the thesis:
Ego depletion is the state of self-regulatory resource depletion. Baumeister et al. (1998) employed this term to describe “temporary reductions in the self’s capacity” to respond on volitional actions like making choices (p. 1253). For instance, ego depletion increased subsequent passivity (p. 1261). How ego depletion comes up, works and the outcomes the effect can have are broadly presented and discussed in the main body of the thesis.
3. Antecedents And Outcomes Of Ego Depletion
Ego depletion has acquired a remarkable importance in the fields of social psychology. This literature review is structured as follows: In section 3.1, papers that prove the effect are presented, beginning from the initial studies to achieve a solid understanding of the effect. Subsequently, it is shown how making decisions during the day evoke the effect and what kind of outcomes ego depletion can cause. Besides that, possible strategies to defend oneself against the effect are portrayed. In section 3.2, concerns about the effect or the strength model are reported.
3.1 Evidence Of The Effect
The state of ego depletion. Baumeister et al. (1998) refered firstly to a temporary reduction in the capacity of the self, calling it ego depletion (p. 1253). Before, “The self-control as limited resource: Regulatory depletion patterns” (Muraven, Tice, and Baumeister 1998) handled the strength model, and then afterwards the paper “Ego-Depletion: Is the active self a limited resource?” (Baumeister et al. 1998) was published at the Case Western Reserve University in the Journal of Personality and Social Psychology. In the ensuing paragraphs the experiments of these two initial papers are presented in detail.
In Study 1 participants were instructed to control their emotional responses while watching an upsetting movie (Muraven, Tice, and Baumeister 1998, p. 776). Thus, after the initial manipulation involved affect regulation, the second task as the dependent measure was quite different, involving the physical stamina on holding a handgrip (p. 777). Measuring the time before and after the exertion of self-control offered the ability to compare the results to the control group, who did not regulate their emotions through the movie. Regarding the results (“Insert Table 1a about here”), participants in the emotional control condition reported more fatigue and a higher perceived load of effort, but there was no significant difference between the ones who enhanced and the ones who suppressed their emotional responses. The time stopped when the participants stopped exerting the handgrip and a significant difference between the emotional response group and the control group has been conducted. While the control group showed almost a similar persistence before and after, there was a significant decline in the emotional response group. The results support the hypothesis that self-regulation operates like a strength, after the subsequent performance on holding the handgrip dropped down in the emotional response group (p. 778).
Study 2 replicated study 1 using different methods. The manipulation of regulatory exertion was induced by suppressing to think of a white bear according to Wegner et al.’s popular procedure from 1987. The subsequent task measured the persistence on unsolvable anagrams (p. 779). The control group could easily express their thoughts, according to the suggested fact that thought suppression is much more effortful than thought expression. For the anagram task, participants did not know that they were unsolvable. Regarding the results (“Insert Table 1b about here”), the suppressing thought participants quit earlier than the ones who could express their thoughts. It was followed that self-regulation involves a limited capacity.
In study 3 the manipulation part involved again the white bear task, but the control group got a mathematical task with difficult multiplication problems. Both tasks were considered and perceived as equal in difficulty, but the suppress thoughts condition should affect more self-regulation (p. 781). For the dependent second measure participants watched a humorous video tape but were not allowed to show any amusements. Participants in the suppressing thought condition were likely to smile much more on the movie than the math condition. It was concluded that depletion of regulatory capacity results specifically in poorer self-control (p. 783).
Study 4 tried to establish greater external validity to generalize the depletion effect. Participants had to write autobiographical narratives about situations they experienced in which they could control emotions and in which they could not (p. 783). Using an independent judge for coding the results has shown that participants mentioned that they felt tired more often in self-regulatory failure stories than in regulatory success stories (p. 784). In contrast, calmness indicated increased regulation and less fatigue and were found more often in regulatory success stories. To summarize, the capacity for self-regulation is depleted by multiple demands (p. 785). This wide range of methods and measures across the studies provides strong support for the limited resource model with a self-regulatory capacity and seems to be the best explanation of the findings in the studies.
Baumeister et al. (1998) published the paper “Ego-Depletion: Is the Active Self a Limited Resource?” and gave birth to the paradigm of ego depletion. The first study illustrates the ego depletion effect very well and provides help to understand the paradigm. The laboratory room was filled with the aroma of fresh baked chocolate cookies (p. 1254). For the manipulation part, participants had to choose radishes while cookies were standing next to them, so impulse control as a form of self-control was required. The control group could easily decide for the cookies or candies. There was also a no food control group that went directly to the second task in which participants had to do a puzzle. Unbeknownst to them, it was impossible to solve the puzzle and the experimenter stopped the time how long they persisted on the task (p. 1255). The results have shown that there was a significant difference between the radishes condition and the chocolate and no food conditions, but the ladder ones did not differ significantly (“Insert Table 1c about here”). The chocolate condition quit sooner and made less attempts to solve the puzzle. Combined with the findings from Muraven, Tice, and Baumeister (1998), the results fit a limited resource model best and let the authors suggest that this resource is a common one for all kinds of self-regulation, namely thought control, emotion control, impulse control, task performance (p. 1256).
Experiment 2 was designed to examine whether choice and volition also draw on the same self-regulatory resource. Having the same problem-solving task as in the previous study 1, the participants had to do some speeches about tuition raises before in the manipulation task. One group had no choice and was assigned to do the pro-tuition speech, while another group was impressively given the choice to decide for themselves (p. 1257). Participants who were in the high choice group showed a significant decline in their persistence on the frustrating, unsolvable task than the no choice condition and the no speech control.
To broaden the theory, different procedures were used in study 3. In this experiment, participants were asked to watch an emotionally evocative videotape but should regulate their emotional responses. Some watched a funny videotape, some a heartbreaking sad one. For the dependent task, participants had to do some anagrams, but this time they were solvable (p. 1258 f.). There was no effect for the type of the video, but a significant difference between both groups who had to suppress (4.94 solved) and the no regulation control group (7.29 solved).
To safe space, experiment 4 is not presented in close detail but provided further support for the hypothesis because the results show that ego depletion increases subsequent passivity. Thus, only active choice draws on the self’s volitional resource.
In general, these findings support the ego depletion effect insofar that an initial act of self-regulation depletes a common resource and therefore the performance in subsequent self-regulatory tasks declines. Furthermore, these findings refute alternative explanations how self-regulation functions as well. If self-regulation was a skill, there would be no changes in the second task (Hagger et al. 2010, p. 497; Muraven, Baumeister, and Tice 1999, p. 447). In addition, if self-regulation was something like a knowledge structure and master schema, there should be increasing performances on the second task (Baumeister et al. 1998, p. 1254). Moreover, as study 2 let participants record a speech about tuition raise, contrary to one’s own belief, cognitive dissonance seems to be an appropriate explanation, howbeit there is no reason that dissonance should decrease the performance on an unrelated task (Baumeister et al. 1998, p. 1262).
Having the findings depicting ego depletion and the explanation of the strength model for it behind, it was a logical conclusion to ask whether self-control resembles a muscle (Muraven and Baumeister 2000, p. 247). A brief look at vigilance tasks, which need a continuous exertion of self-control, exhibited that the longer participants concentrated on detecting any distractions, the greater the decline in accuracy became (Muraven and Baumeister 2000, p. 254). In addition, not only that exertion of self-control or a muscle produces short-term fatigue, it could lead to long-term improvement in both cases (p. 254).
Therefore, a new study was launched, investigating whether repeated exercise would improve the self-control strength for self-regulation (Muraven, Baumeister, and Tice 1999, p. 446). For the results two possible ones were suggested, but only one reached some evidence: There was no improvement of the simple capacity of self-control, but a less vulnerability to ego depletion could be observed (p. 453 f.). The muscle did not have any greater power, but “it had greater stamina and was less prone to suffer from rapid fatigue after the exercises” (Muraven, Baumeister, and Tice 1999, p. 454).
In this line, there is also space to inquire whether it is possible to conserve some of the self-regulatory strength (Muraven and Baumeister 2000, p. 255). Later, an examination exposed an interaction of self-control performance among the amount of self-control exerted previously and the expected amount of self-control to be exerted in the near future (Muraven, Shmueli, and Burkley 2006, p. 536). More specifically, depleted participants who anticipated exerting self-control in the future performed worse on a second self-control task than did participants who were not depleted or who did not anticipate another self-control exertion soon (p. 536). Muraven, Shmueli, and Burkley (2006) posited that in many situations individuals are not unable but instead are unwilling to exert the amount of self-control required facing the approaching future (p. 536). Individuals will do so if they attribute more importance to a future demand than to a comparable gain at the moment (p. 525). Hence, the conservation strategy may be useful to manage the consumption of the limited resources in order to save it for future demands (p. 536).
Ego depletion by making choices. The next paragraph examines how making choices can impair self-regulation. Loads of everyday choices are effortless and “guided by intuitive thinking” (Pocheptsova et al. 2009, p. 345). On the contrary, deliberate choice making is described as effortful and can be difficult (Vohs et al. 2008, p. 895). Making choices apparently depletes the same limited resource as the underlying studies have shown (p. 895). Participants who made choices on class materials, college courses and consumer products performed more poorly in subsequent tasks than the control group (p. 895). Moller, Deci, and Ryan (2006) emphasized the importance to differentiate more and divulged a distinction between controlled choice and autonomous choice (p. 1035). Albeit the ego depletion effect was replicated, referring to Baumeister et al. (1998) study 2, Moller, Decy, and Ryan (2006) pointed out that the labeled high-choice condition was not an autonomous choice as participants had to choose under pressure by briefing (p. 1034). Divided by a significant difference in energy and persistence, controlled choice was depleting whereas autonomous choice turned out as not depleting (p. 1033). Whilst Vohs et al. (2008) agreed to this view, they added that “making many choices becomes depleting even when the activity is viewed as an opportunity for positive self-expression” (p. 895). Pocheptsova et al. (2009) claimed that depleted individuals rely on intuitive processing and simpler decision strategies (p. 352). Resource-depleted participants were furthermore susceptible “to be influenced by the reference point in their choices” (p. 348). These results iterate two empirical findings: First, the capability to censor the intuitive reactions depends on the availability of the self-regulatory resources and the willingness to use them (Baumeister et al. 1998, p. 1253; Muraven and Baumeister 2000, p. 255; Muraven and Slessareva 2003, p. 904; Pocheptsova et al. 2009, p. 345). Second, depleted individuals show more sensitivity that can lead to greater susceptibility to deviations (Pocheptsova et al. 2009, p. 352). However, Wang et al. (2010) went one step beyond and suggested that trade-offs while making choices are the roots of the depleting effects of choices (p. 917). In particular, as “trade-offs require both preference formation and conflict resolution”, resolving conflicts did deplete more executive resources than forming preferences (p. 918). Wang et al. (2010) posited that “the greater the trade-off conflict, the greater are the required executive resources, and therefore the greater is the depletion effect” (p. 914).
Effects of ego depletion. The following paragraphs reveal some of the many different spheres in which ego depletion can have effects. One paper provides support that depletion impairs mental performances like logical reasoning which requires active guidance by the self while simple information processing like rote memory does not exert self-control (Schmeichel, Vohs, and Baumeister 2003, p. 33 f.). To generate the manipulation, participants were instructed to regulate their attention or their emotional responses in another experiment (p. 43). In the following tasks participants in the state of ego depletion showed poorer intellectual performance on the analytical GRE, GMAT and CET tests that require high cognitive operations (p. 43). Ego depletion thwarts the controlled functioning of the central executive (p. 43). As long as the central executive is engaged in the responses to a task, it is regardless whether the task involves cognitive work or muscle movements (p. 44 f.). Furthermore, the authors postulated that the greater the executive requirements are, the more it is harmed by ego depletion (p. 44).
In this line fits an investigation how ego depletion has effects on attitude change processes. Resisting persuasion is also linked to draw from the same limited resource since counterarguing persuasive messages engrosses active control processes (Wheeler, Briñol, and Hermann 2007, p. 151). The study yielded that strong arguments were more persuasive than weak arguments, “but depleted individuals were significantly more persuaded by weak arguments than nondepleted individuals” (p. 152). Thus, although the underlying message was counter-attitudinal, depleted egos generated more favorable cognitive thoughts and responses (p. 153). These results extend the discernment of how situational factors influence information processing and attitude change, videlicet former situational variations could affect subsequent attitude change processes (p. 153).
Another interesting aspect is the behavior of chronic dieters where ego depletion could be a plausible explanation why dieters fail to inhibit urges to eat (Vohs and Heatherton 2000, p. 249). Dieters and nondieters sat down in a study, whereby it was assumed that only dieters exert self-control as they actively monitor and inhibit caloric intake (p. 252), thereby decreasing their ability to self-regulate (p. 254). Like in the first experiment of Baumeister et al. in 1998, tempting cookies were placed next to the participants, in this case either directly beside them for high temptation or across the room for low temptation while they watched a neutral video. Half of both groups were allowed to eat the cookies while the other half were assigned not to touch. As the dependent measure served the amount of ice cream eaten in a subsequent task to taste and rate ice cream flavor (“Insert Figure 1 about here”). A significant interaction between temptation level and availability condition revealed that dieters who were highly tempted and allowed to help themselves to the cookies ate significantly more ice cream (M = 181.7g, SD = 95.4) than dieters who were low tempted and also allowed to help themselves (M = 71.7g, SD = 57.4), whereas no significant differences among the nondieters could be ascertained (p. 251).
In study 2 chronic dieters sat down on the same manipulation task as in study 1, but afterwards they had to work on an embedded-figures task which was made unsolvable. Highly tempted participants spent less time on it (M = 17.0 min, SD = 4.5) than low-temptation participants did (M = 21.8 min, SD = 6.9). Furthermore, this supports the hypothesis that self-regulatory resources are global and not domain-specific (Vohs and Heatherton 2000, p. 252). Moreover, in study 3 chronic dieters were exposed to a sad movie while half of them should inhibit their emotional reactions. Dieters who suppressed their emotions ate significantly more ice cream (M = 211.2g, SD = 123.8) than those who were allowed to express their responses naturally (M = 135.6g, SD = 71.7) (Vohs and Heatherton 2000, p. 253).
A series of experiments disclosed that self-regulatory resources and self-presentations are integrally related (Vohs, Baumeister, and Ciarocco 2005, p. 653). In four studies participants had to do a self-presentation first and became depleted as their performance in a subsequent task like puzzle solving or handgrip persistence dropped down (p. 653). Widdershins, in alternative four experiments participants were manipulated first, for instance, in study 7 by a Stroop task (p. 649). In the typical Stroop test, color words printed in ink colors that are incongruous in their meaning are displayed to participants. For instance, the word ‘red’ is printed in green ink (Webb and Sheeran 2003, p. 280). Therefore, this task requires the self to override automatic responses to answer correctly. After that, participants had to do certain self-presentations and depleted individuals tended to show off less favorable impressions, to offer a greater willingness to speak without monitoring what they said or to behave more narcissistic (Vohs, Baumeister, and Ciarocco 2005, p. 653 f.).
A completely different area Gailliot, Schmeichel, and Baumeister (2006) scrutinized was the assessment of how self-regulation could help the self to defend against fears and thoughts of dying (p. 49). Naturally, after thinking about death, people try to suppress thoughts and emotions about mortality. Reconsidering, the process of thought suppression and emotion regulation requires self-control (p. 50). Like before, the experiments were carried out in both directions (p. 59). First, states of self-regulation and the individual level of trait control had impacts on the susceptibility of thoughts concerning mortality and death-related anxiety (p. 50). Second, attempts to avoid thinking of death led to reductions in self-regulatory resources (p. 55). To put this into relation to ego depletion, participants with low states of self-regulatory resources and low traits of self-control were most vulnerable to intrusive thoughts about death (p. 59). Analogously, having mortality salience in the first part of the experiment, subsequent performances on effortful tasks like the Stroop test or analytical reasoning problems decreased. To conclude, self-regulation is a useful instrument in the management of mortality concerns, although the service should be handled with caution concerning the limited capacity of the resources that could be missing elsewhere (p. 60). In addition, trait and state capacity for self-regulation can serve for a prediction which people will suffer most from threats of death (p. 59).
Another fear people have is being reduced to a stereotype (Inzlicht, McKay, and Aronson 2006, p. 262). The underlying studies detected a clear causal connection among stigma and impaired self-control (p. 265). In the second study, after activating race-based stereotypes for Black participants, Black students took significantly longer in the subsequent Stroop task to complete. However, White students did not show any differences between the conditions (p. 265). To summarize, when people realize that they are the target of prejudice the stigma increases people’s stress (p. 263). Stigmatized people use and deplete their self-regulatory resources to administer their devalued social identity that can lead to intellectual underperformance, as above in avoiding fears of death. Likewise, stigma can fade the fundamental ability to control one’s actions (p. 267).
In another field, it was investigated whether dishonesty depends on self-control (Mead et al. 2009, p. 594). In the second experiment, participants were depleted by the Stroop test and worked subsequently on a quiz about university. At the end, participants had to transfer the answers to a bubble sheet, but could check the correct answers. For every correct answer participants could earn $.10 (p. 595). As shown by the results (“Insert Figure 2a about here”), depleted participants were more likely to bare themselves to the temptation to cheat (p. 596). Moreover, participants in the state of ego depletion cheated significantly more. This leads to the conclusion that – in the state of ego depletion – selfishness and dishonesty may easily ensue (p. 597). The findings of dishonest behavior were replicated by Gino et al. (2011). Furthermore, the authors linked unethical behavior and states and trait of self-control as moral identity was found to moderate the effect of depletion on dishonesty (“Insert Figure 2b about here”; p. 195). Renouncing from unethical behavior consumes self-regulatory resources. In the state of ego depletion combined with low individual traits, the greatest transformation towards unethical behavior was observed. Participants lied about their performance, could not resist the temptation to cheat and were less able to recognize their shift towards unethical behavior (p. 199).