IT TAKES TWO TO TANGO: EXAMINING THE INTERDEPENDENCE OF STATE WORKING MEMORY CAPACITY AND EGO DEPLETION By Joshua Jwala Prasad A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Psychology -Doctor of Philosophy ABSTRACT IT TAKES TWO TO TANGO: EXAMINING THE INTERDEPENDENCE OF STATE WORKING MEMORY CAPACITY AND EGO DEPLETION By Joshua Jwala Prasad Recent research uses ego depletion to ex plain daily variation in organizational phenomena like how sleep deprivation leads to abusive supervision or how the experience of incivility leads to instigation of incivility (Barnes et al., 2015; Rosen et al., 2016). However, ego depletion is typically incorporated as the sole theoretical perspective for mediation. I propose that working memory capacity (WMC) may be a valuable additional mediator in the examination of daily variability in organizational constructs. Incivility and sleep deprivation have b een tied to poorer working memory function (Porath & Erez, 2007; Maltese et al., 2015) and WMC itself has been shown to exhibit meaningful daily variability (Sliwinski et al., 2006; Brose et al., 2012). I hypothesize that daily experienced incivility , slee p deprivation , and engagement in physical activity may impact WMC, whic h may in turn predict daily progress towards short - and long -term academic goals . Ego depletion may also moderate the relationship between WMC and these outcomes. State achievement moti vation is incorporated as a predictor of ego depletion, to distinguish antecedents of motivation and cognition (Locke & Schattke , 2018). Further, Effort and Forgetfulness are investigated as unique outcomes of ego depletion and WMC, respectively. Several m otivational and cognitive individual differences are also examined to explore the independence of the effects of daily variation in WMC and ego depletion. Analyses of data provided by university students in an experience sampling study reveal that daily va riation in WMC is systematic and can be predicted by physical activity and caffeine consumption. Little evidence is provided for the consequences of variation in WMC or its interplay with self -regulatory resources. Further investigation of the nature of wi thin -person variation in WMC would be fruitful. iv TABLE OF CONTENTS LIST OF TABLES ....................................................................................................................... v LIST OF FIGURES .................................................................................................................... vi INTRODUCTION ........................................................................................................................ 1 Daily Variation in Working Memory Capacity ....................................................................... 6 Predictors of State WMC ......................................................................................................... 9 Sleep Deprivation ................................................................................................................ 9 Incivility ........................................................................................................................... 12 Physical Activity .............................................................................................................. 15 Ego Depletion and its Distinction from State WMC ............................................................ 16 The Role of State WMC and Ego Depletion during Performance Episodes ........................ 22 Between -Person Differences in Within -Person Relationships ............................................. 33 METHOD .................................................................................................................................. 37 Participants and Recruitment ................................................................................................ 37 Procedures ............................................................................................................................. 37 Measures ............................................................................................................................... 38 Sleep Deprivation ............................................................................................................. 38 Achievement Motivation ................................................................................................. 39 Physical Activity .............................................................................................................. 39 Experienced Incivilit y ...................................................................................................... 40 Working Memory Capacity ............................................................................................. 40 Ego Depletion .................................................................................................................. 44 Short - and Long -Term Academic Goal Progress ............................................................. 46 Forgetfulness .................................................................................................................... 47 Effort ................................................................................................................................ 47 Implicit Theories of Will power ....................................................................................... 48 Trait Self -Control ............................................................................................................. 48 General Mental Ability .................................................................................................... 48 Time Management Strategies .......................................................................................... 49 Control Variables ............................................................................................................. 49 Analytic Strategy .................................................................................................................. 50 Model Specification ......................................................................................................... 51 RESULTS .................................................................................................................................. 60 Test of Hypotheses ................................................................................................................ 63 DISCUSSION ............................................................................................................................ 74 Theoretical Implications ....................................................................................................... 74 Working Memory ............................................................................................................. 74 Ego Depletion .................................................................................................................. 78 v Antecedents of Performance ............................................................................................ 81 Practical Implications ............................................................................................................ 84 Limitations and Future Directions ........................................................................................ 87 Conclusion ............................................................................................................................ 91 APPENDI X ................................................................................................................................. 93 REFERENCES ........................................................................................................................ 104 vi LIST OF TABLES Table 1. Between -Person Descriptives and Intercorrelations among Studied Variables ..............61 Table 2. Within -Person Correlations among Daily Variables ......................................................63 Table 3. Percentage of Variance Within -Person among Daily Variables .....................................64 Table 4. Path Analytic Results from Model 1 ...............................................................................65 Table 5. Indirect Effect Estimates from Model 1 .........................................................................68 Table 6. Cross -Level Interaction Results from Models 2 and 3 ...................................................71 vii LIST OF FIGURES Figure 1. Theoretical Model. H1 through H17 denote the hypotheses posed in the Introduction . ........................................................................................................................................................10 Figure 2. Graphical representation of the numerical updating task . .............................................42 Figure 3. Graphical representation of Model 1 estimation. Variables grouped by their location in the theoretical model. An arrow from one group to another represents estimated paths between all variables in those groups. Autoregression boxes depict regres sions of each variable in a group regressed onto its measurement from the prior day . .....................................................................55 Figure 4. Graphical representation of estimation of Models 2 and 3. The outcomes for Models 2 and 3 are Short -Term and Long -Term Academic Goal Progress, respectively. Variables grouped by their location in the theoretical model. An arrow from one group to another represents estimated paths between all variables in those groups. Autoregression boxes depict regressions of each variable in a group regressed o nto is measurement from the prior day. ...............................57 1 INTRODUCTION Studies of ego depletion, or one™s fluctuating capacity to engage in self -control, typically incorporate depletion as the sole theoretical perspective for mediation between constructs in organizational settings. Recent research draws on ego depletion to ex plain phenomena like how sleep deprivation leads to abusive supervision, or how the experience of incivility leads to instigation of incivility (Barnes, Lucianetti, Bhave, & Christian, 2015; Rosen, Koopman, Gabriel, & Johnson, 2016 ). Ego depletion is thoug ht to impact outcomes over time through the regulation of attention (Beal , Weiss, Barros, & MacDermid , 2005), but it is unrealistic to assume that attention is solely shaped by ego depletion. Further, it is also unlikely that ego depletion is the only medi ating factor that helps tie experiences like sleep deprivation and incivility to subsequent outcomes. I propose that working memory capacity (WMC) may be a valuable construct in furthering this line of research . By examining WMC alongside ego depletion, no t only can we explore how WMC may be relevant in organizational settings but we may also gain a clearer understanding of the importance of ego depletion. WMC was originally developed by Baddeley and Hitch (1974) as a cognitive workspace for tasks that an individual is consciously aware of. WMC has been shown to predict an individual™s ability to constrain or control a variety of behaviors, such as carrying out a task with out being distracted (McVay & Kane, 2012). Relatedly, this relationship between WMC and control has been shown to have positive real -world associations, such as predicting SAT scores (Engle, Tuholski, Laughlin, & Conway, 1999), learn ing in a military training environment (Baddeley, 1996), and reading comprehension (Engle, 2002). Given the breadth of potential 2 outcomes , understanding why variation in WMC occur s is important for researchers and practitioners concerned with improving the effectiveness of people at work . However, broad application of WMC to the workplace has been limited in two ways. First, WMC is studied almost exclusively as a predictor of organizationally relevant outcomes, with little attention paid to potential antecedents. Recent work has shown that the experience of uncivil behaviors can negatively impact task performance through hurting WMC functioning (Rafeali, Erez, Ravid, Derfler -Rozen, Triester, et al. 2012 ; Porath & Erez 2007) . Investigations like these are scarce, leaving little understanding of what factor s may impact one™s WMC in the workplace. Although research identifying outcomes of WMC is important, further examination of antecedents of this construct may broaden its utility within organizations. A second limitation of past work examining WMC in the wo rkplace is that this research typically takes a static approach, assuming the only meaningful variation in this construct occurs between individuals . However, t his approach ignores the possibility that all individuals are susceptible to having good and bad days concerning their cognition . The present investigation examines the possibility t hat WMC may fluctuate within -person in a meaningful way. Doing so may help to clarify a critical aspect of this construct. My research offers a complementary perspectiv e on the typical individual differences approach to WMC by focusing on daily fluctuation s in WMC. Disparate pieces of evidence have emerged to suggest the presence of within -person variation in WMC. First, past work has demonstrated that WMC may be negativ ely affected by instances of sleep deprivation (Banks & Dinges, 2007; Goel, Rao, Durmer, & Dinges, 2009 ) and uncivil behavior from others ( Porath & Erez, 2007; Rafeali et al. 2012). Second, both sleep deprivation and the experience of incivility have been shown to exhibit within -person variability that can explain organizational 3 consequences ( Cristian & Ellis, 2011; Rosen et al. 2016). Taking both points together, sleep deprivation and incivility may also influence WMC and result in meaningful within -person variability across days . Building on this idea , I expand the conceptualization of WMC to include state WMC, which I define as a time bound deviation in one™s capacity to maintain and manipulate information. I argue that there is a worthwhile am ount of predictive validity within -person and that a view of WMC across time opens the door to interventions that m ay be useful to current employees. Further, state WMC may be useful in future efforts connecting sleep deprivation and incivility to broader consequences tied to each phenomenon. The RAND corporation recently reported that the United States loses an estimated $411 billion annually as a result of insufficient sleep among American employees ( Hafner, Stepanek, Taylor, Troxel, & Van Stolk, 2016). Incivility, on the other hand, saps the equivalent of seven work weeks per year among managers at Fortune 1000 companies, impacts 98% of U.S. employees, and may cost organizations millions of dollars per year (Porath & Pearson, 2013). An examination of stat e WMC may broaden the usefulness of WMC as a construct that has primarily been applied to personnel selection (e.g. Bo sco, Allen, & Singh, 2015 ). Further, a within -perso n approach to WMC allows for further connection to a broader trend in organizational scholarship of increased focus on day -to-day variability in psychological constructs relevant to the workplace. Self -regulation has played a primary role in this trend by illustrating the processes that occur over time and explain construct change (Lord, Diefendorff, Schmidt, & Hall, 2010; Neal, Ballard, & Vancouver, 2017) . Organizational scholars have revealed that one™s ability to self -regulate is also susceptible to thing s like sleep deprivation and incivility , which can impact workplace performance through the self -regulatory construct of ego depletion (e.g. Cristian & Ellis, 2011; Rosen et al. 2016). However, variability in ego depletion 4 has been studied in the absence o f WMC , a construct that may bear similar causal inputs and outcomes . By exploring variability in WMC alongside ego depletion , the relative explanatory power of each mediator may be explored, and a broader understanding of self -regulation may be reached . Past scholars have argued that cognitive and self -regulatory resources may interact in their influence on performance (Beal et al., 2005) . Ego depletion has been incorporated in past work as a construct describing the available resources one has to engage in self -control, often in the pursuit of goals (Cristian & Ellis, 2011; Rosen et al. 2016). Thus, ego depletion may describe one™s volition al capacity to pursue a goal, whereas state WMC captures an individual™s moment -to-moment capability to maintain goal -relevant information in mind during goal pursui t. Building from Beal et al.™s (2005) theory, I expect that ego depletion and state WMC cap ture daily variation in cognitive and self -regulatory resources and that these constructs may have joint and unique impacts on goal -pursuit. I examine this prospect directly by examining two goal progress outcomes that differ regarding important task chara cteristics . I examine short -term and long -term academic goal progress , such as completing regular homework assignments or writing term papers . The timescale of deadlines associated with each form of goal progress should modify the extent to which the task itself fosters motivation (Hendy, Liao, & Milgram, 1997; Steel & König, 2006 ). As a result , potential interactive effects between state WMC and ego depletion may be examined across outcomes that differ in terms of important task features . My study also i ncorporates antecedents , outcomes , and between -person differences to help distinguish ego depletion from state WMC. Regarding antecedents, Ludyga , Gerber, Brand, Holsboer -Trachsler, and Pühse (2016) meta -analytically summarize d studies of physical activity and working memory, finding that engagement in physical activity produces temporary benefits 5 in measures of WMC. However, the relationship between ego depletion and physical activity requires a longer period to stabilize , give n that a single instance of physical activity may be depleting but can lead to improvement in self -control over repeated engagements (Oaten & Cheng, 2006). Locke and Schattke (2018) recently clarif ied the construct of state achievement motivation, which sh ould help explain variation in ego depletion more so than state WMC. For outcomes, I investigate daily variation in forgetfulness (Rast, Zimprich, Van Boxtel, & Jolles, 2009) as an outcome determined by the quality of one™s state WMC. Daily variation in ef fort (Pintrich & DeGroot, 1990), on the other hand, is incorporated to demonstrate the unique consequences of depletion. Additionally, individual differences in trait self -control, implicit theories of willpower, general mental ability, and use of time man agement strategies are used to evaluate whether the within -person relationships between state WMC , ego depletion , and their outcomes are dependent on specific traits. As a whole, the theoretical model I investigated is used to thoroughly compare and contra st ego depletion and state WMC. Thus, the purpose of this investigation is to extend theory on WMC by examining the antecedents and outcomes of within -person variance. Specifically, I reevaluate past studies relating sleep quality and incivility to WMC as evidence for the need to study the relationships between these constructs within -person . I draw on theory by Beal et al. (2005) to suggest that when individuals suffer cognitive setbacks due to poor sleep or experience s of incivility , performance may suf fer for reasons other than ego depletion . Further, ego depletion may exacerbate the impact that temporary deficits in state WMC may have on performance outcomes . Consistent with the theory I have developed here, I propose test ing my model of the antecedents and outcomes of WMC using a sample of students participating in an experience -sampling study . 6 Daily Variation in Working Memory Capacity A key function of WMC is to temporarily store and process relevant information while continuing to experience other environmental stimuli (Miyake & Shah, 1999). This allows for information gained through perception as well as conscious awareness to be processed and contribute to more complex cognition and behavior (Baddeley, 1996). From these seminal works, further investigation of WMC has be en siloed in multiple disciplines. Alongside the development of theories of WMC in the literature of cognitive psychology, theories of ex ecutive function have been advanced in the field of neuropsychology. Broadly, this body of work attempts to align executive functions with the structure of the prefrontal cortex to examine the relationship between neurological functioning and voluntary beh avior (Salthouse, Atkinson, & Berish, 2003). Likely the most widely used model of executive function is the updating, shifting, and inhibition model by Miyake et al. (2000). Updating refers to updating the content of one™s memory to what is relevant to the current moment, a function that is carried out by working memory. Switching refers to changing the direction of one™s voluntary behavior when environmental changes make it beneficial to do so. Lastly, inhibition refers to maintaining the alignment between one™s focus and their intentions in the presence of environmental demands and distractions. Notably, the development of this line of research also draws heavily on the work of Baddel ey (i.e. Baddel ey & Hitch, 1974; Baddel ey, 2003) in the conceptualization of executive function. Neuropsychological research on executive functions have provided meaningful clarity in identifying physiological process es that may influence working memory function. Pertinent to this study , sleep deprivation and engaging in physic al activity both have the potential to temporarily influence WMC through physiological changes (described further below). 7 Investigation of WMC has also occurred in another stream of research, namely the psychometric examination of cognitive ability. This line research derives primarily from Spearman™s (1904) investigation of g, which refers to a general factor that explains performance on a range of cognitive tasks. Given that individuals who perform well on one test tended to perform well on a variety of tests, g is discussed as an individual™s standing on a single latent construct that determines performance across tasks requiring cognitive capability . A modern iteration of this approach is the Cattell -Horn -Carroll (CHC) three -stratum model of general men tal ability (GMA; McGrew, 2009). Rooted in Spearman™s (1904) original work examining g, the CHC model extends the notion of g into a taxonomy of cognitive abilities. A general factor is still retained and thought to underlie the execution of cognitive task s and placed at the highest, third stratum, of the taxonomy. Each lower stratum can be thought of as decomposing human cognitive ability at different levels of specificity. The second -stratum contains as many as 16 broad cognitive abilities such as reasoni ng, general knowledge, short -term memory and long -term memory. Abilities at this level are meant to strike a balance between generality and specificity. Further below are first -stratum abilities, which consist of over 80 specific cognitive abilities as nar row as reading speed, lexical knowledge, and visual memory. Several studies have explored the relationship between WMC and general measures of cognitive ability (e.g. Ackerman et al., 2005; Redick, et al., 2016) concluding the two were highly related but separable constructs. Recent work by Jewsbury, Bowden, and Strauss ( 2016) attempted to integrate WMC and the CHC model, finding that updating measures of WMC are well incorporated into the short -term memory dimension of the CHC model . Thus, the most parsimonious explanation is that WMC captures a specific dimension of gene ral mental ability. Specifying WMC as a dimension of cognitive ability is important given that many of the arguments posed here 8 regarding daily fluctuation in WMC would not align with general conceptualizations of cognitive ability. Theory and research o n the relevance of WMC for organizations has focused more on outcomes than antecedents. However, examples do exist that consider factors that may impact WMC, such as sleep deprivation and experiencing incivility (Maltese, Adda, Bablon, Hraeich, Guervilly, et al. 2015; Porath & Erez, 2007 ). Sleep deprivation and incivility have also been shown to exhibit meaningful within -person variance in organizational settings (e.g. Rosen et al., 2016; Barnes et al. 2015) and models relating these inputs to WMC should focus on daily variability to be adequately specified at the proper level of analysis (Klein & Kozlowski, 2000). Although more work at the between -person level of analysis may be helpful, suggesting that within -person variance in WMC is error may be an insta nce of model misspecification. Beyond potential model misspecification, conceptualization of WMC warrants consideration of within -person variance. Focusing on between -person variability treats WMC as static over time, despite evidence that it may vary . WMC as a construct captures the effective coordination of maintaining memory, updating memory in the presence of relevant novel information, and the inhibition of irrelevant stimuli (Redick, Calvo, Gay, & Engle, 2011). These functions distinguish WMC from simply storing information in mind and contribute to cognition throughout one™s present experience (Baddeley, 1996 ; Engle, 2002). Given the fact that WMC is explicitly conceptualized as a cognitive construct that opera tes over time, it should follow that examination of the influence WMC may have on performance should also occur over time. Some studies have examined WMC across days to assess whether within -person variability is meaningful. Sliwinski, Smyth, Hofer, and S tawski (2006) assessed WMC and stress across six days, finding that WMC task performance was better on low stress days. Brose, 9 Schmiedek, Lövdén , and Lindenberger (2012) found that WMC scores were associated with negative affect at both between - and within -person levels of analysis. These examples demonstrate that daily measurement of WMC captures useful information in that it relates to psychological phenomena such as changes in stress and affect over time. In my study, I refer to daily change in WMC as state WMC , which I define as a temporary deviation from one™s trait level of WMC. Further, these empirical examples demonstrate that the relationships state WMC has with other constructs are similar between - and within -person (i.e. multilevel homology) . I highlight these findings as evidence that multilevel homology is likely and that state WMC should bear relationships with other constructs at the within -person level of analysis that are similar to trait conceptualizations of WMC (e.g. Redick et al. 2011). However , state WMC specifically describes a temporary change in capability. The following sections build the conceptual model I investigate in this study , which is depicted in Figure 1. Predictors of State WMC My investigation examines physiological states and psychological experiences that have the potential to impact state WMC. Specifically, I examine how prior sleep, physical activity, and the experience of incivility may impact cognitive function ing . Each of these antecedents has been investigated for their impact on the effectiveness of individuals in organizational settings as well as their influence on working memory. Sleep Deprivation . Sleep is a state of immo bility and unresponsiveness that serves a broad restorative function on brain activity (Saper, Scammell, & Lu, 2005). Sleep deprivation describes suboptimal sleep which can be conceptualized in terms of poor quantity and quality (Barnes et al., 2015). Defi ciencies in sleep quantity have been operationalized in several ways in past examinations of sleep deprivation, including instances of prohibition of sleep for multiple 10 Figure 1. Theoretical Model . H1 through H1 7 denote the hypotheses posed in the Introduction. 11 days (Caldwell, Caldwell, Smythe, & Hall 2000), waking hours extending beyond a 24 -hour day (Caldwell, Caldwell, Brown, & Smith, 2004 ), and chronic restriction of sleep to fewer than six hours (Van Dongen, Maislin , Mullington, & Dinges, 2003 ). Partial sleep deprivation has been used to distinguish relatively less severe decrements to sleep quantity from these more extreme examples , including any instance of fewer than seven hours of sleep in a single day (Christian & Ellis, 2011; Ferrara & De Gen naro, 2001). Though relatively less severe, partial sleep deprivation has been linked to poorer cognitive functioning among employees (Barnes & Hollenbeck, 2009; Christian & Ellis, 20 11). Given the dai ly within -person effects examined here, sleep deprivation will be examined based on the duration of sleep the prior night . Barnes (2012) also suggests that quality is a distinct aspect of sleep that impacts the restorative effects of sleep alongside sleep quantity. Sleep quality refers to difficulties in falling asleep and staying asleep, which was shown by Barnes et al. (2015) to uniquely predict daily ego depletion and subsequent abusive supervisory behaviors enacted by managers. Thus, I define sleep depr ivation here as poor quantity and quality of sleep prior to a given day. Though sleep deprivation generates a host of negative physiological effects, of main concern here is the negative effect on prefrontal cortex functioning. Sleep deprivation is thoug ht to create a state referred to as wake state instability ( Doran, Van Dongen , & Dinge s, 2001; Goel et al. 2009). Broadly, this state involves frontal areas of the brain that influence voluntary behavior to periodically give way to activation of other neural areas associated with the initiation of sleep. This oscillation of activation between areas related to sleep and areas involved in wakefulness give rise to the common experience of lapsing, or brief moments of intense fatigue that punctuate the experience of alertness (Lim & Dinges, 2008). Sleeping a suboptimal number of hours per night has been shown to have a negative impact on WMC and this effect can 12 accumulate across nights of re duced sleep (Banks & Dinges, 2007). Inhibited cognitive functioning as a result of sleep deprivation has also been explored as an explanation for the increase of medical errors among sleep deprived medical professionals (Maltese et al., 2016) . Multiple mea sures of WMC, as well as measures of processing speed and reasoning, were found to be significantly worsened after a sample of medical professionals had worked an overnight shift. Furthermore, WMC deficits were unrelated to prior work experience and observ able regardless of having the opportunity to sleep during the shift. Given the link between sleep deprivation and working memory functioning (Banks & Dinges, 2007) and past work demonstrating the impact of daily variability in sleep deprivation (e.g. Barne s et al., 2015), it should be the case that daily sleep deprivation impacts state WMC among the students examined here. Additionally, given that past work demonstrates that sleep deprivation can lead to ego depletion (e.g. Barnes et al., 2015), it should a lso be the case that sleep deprivation among students leads to depletion. H1 Œ Daily variation in sleep quality/quantity predicts daily variation in state WMC . H2 Œ Daily variation in sleep quality/quantity negatively predicts daily variation in ego deple tion. Incivility . Ander sson and Pearson (1999) define i ncivility as filow intensity deviant behavior with ambiguous intent to harm the target, in violation of [institutional] norms for mutual respect. Uncivil behaviors are characteristically rude and discourteous, displaying a lack of regard for others™™ (p. 457). Though incivility is commonly studied as a form of deviant workplace behavior, several studies have shown that incivility does occur in the university context and the experience of incivility can lead to negative outcomes among students. Farrell, Provenzano, Spadafora, Marini, and Volk (2016) show that incivility in academic contexts can 13 appear intentional or unintentional in nature. Intentional academic incivility behaviors include calling a classmate names or spreading rumors, whereas unintentional behaviors include eating during class or packing up before class has been dismissed. Farell et al. (2016) found that measures of intentional and unintentional uncivil behaviors were best modelled as independent but highly correlated factors. Focusing primarily on intentional uncivil behaviors, Caza and Cortina (2007) found that over 75% of university students had been the target of incivility in the prior year from eit her peers or faculty. Further, incivility from both sources led to psychological distress, poorer satisfaction with their institution, and poorer academic engagement and performance. Thus, in academic contexts it is important to distinguish faculty versus students as the source of incivility, even if both sources may yield similar negative effects. Jensen, Ahmad, King, and Lee (2016) found evidence for a similar link between the experience of incivility and distress among students, which ultimately led to w orse self -perceptions of competence. Findings from Alt and Itzkovich (2016) indicate that frequently experiencing incivility from faculty led to poorer student adjustment to university life. These studies show that incivility does occur in academic context s and that the negative consequences for students may be far reaching. Pertinent to the present study is the impact experiencing incivility may have on state WMC. Across three studies, Rafeali et al. ( 2012) demonstrate that minor verbal aggression from customers can hurt performance on working memory tasks. Further, reductions in WMC were associated with poorer customer service task performance among both student and customer service employee samples. Th e verbal aggression manipulations used by Rafeali e t al. (2012) could be construed as instances of incivility as they were of low intensity, rude, and portrayed a lack of respect for participants. Porath and Erez (2007) obtained similar findings such that merely imagining an uncivil encounter negatively im pacted task performance and that this effect 14 was fully mediated by a measure of WMC. Both Rafeali et al. (2012) and Porath and Erez (2007) suggest that decrements to WMC are due to the negative affective experience leading to further processing that saps cognitive resources . Porath, Mac Innis, and Folk es (2010) show that witnessing incivility (as opposed to being the target) can lead to rumination, which describes repeated thoughts about a stressful event that are intrusive to thoughts an individual may othe rwise have (Watkins, 2008). Thus, rather than directly constraining WMC, the experience of incivility may place a somewhat persistent demand on WMC by occupying memory with thoughts of the experience rather than information related to the task at hand. Though the experimental manipulations by Rafaeli et al. (2012) and Porath and Erez (2007) provide evidence for the causal impact experiencing incivility may have on WMC, these studies do not examine the relationship between these constructs over time. Cole, Shipp, and Taylor (2016) argue that this is problematic, given that incivility exists across time by nature and a failure to account for the passage of time may bias the observed relationship between constructs. A major challenge is that the consequences of incivility may change across timescales (e.g. after a specific experience versus reflection days or weeks later) and may exhibit a growth or spiral -like effect (Cole et al. 2016). Here, I focus specifically on daily variation in the experience of incivi lity. Rosen et al. (2016) found that university employees who experienced incivility during the day were more likely to enact uncivil behaviors by the end of the day. Further, this effect was mediated by performance decrements on the Stroop task, a measure of attentional control often used to assess ego depletion (Stroop, 1935; MacLeod, 1991). Thus, there is evidence that daily variation in the experience of incivility impacts both the control of attention as well as subsequent workplace behavior. Given the theoretical accounts from the experimental studies described above, the daily variation in the experience of incivility should 15 also impact state WMC. Further, findings from Rosen et al. (2016) that the e xperience of incivility is depleting should also be observable among the students here , such that students who experience incivility should report greater ego depletion as a result. H3 Œ Daily variation in experienced incivility negatively predicts daily variation in state WMC. H4 Œ Daily variation in experienced incivility predicts daily variation in ego depletion. Physical Activity. Engaging in physical activity has been shown to have broad benefits for individuals in organizational contexts as well as for working memory function (e.g. Barber, Taylor, Burton, Bailey, 2017; Johnson & Allen, 2013; Ludyga et al., 2016). However, these literatures are largely independent and take unique perspectives in conceptualizing the value of physical activity . Physical activity is typically viewed from an occupational health perspective when investigated in an organizational context. Jones et al. (2007) examined physical activity as a behavioral health outcome that workers engaged in less frequently on days that they experienced greater negative affect. Johnson and Allen (2013) showed that reduced job demands predicted working mothers™ engagement in physical activity . Further, the behavior of mothers had subsequent benefits for their children™s physical activity and over all health. Other work in this domain views physical activity as a moderator, such as how Barber et al. (2017) demonstrate that negative interactions with supervisors increased the likelihood of negative personal and family spillover effects and this proce ss was weaker among those who engaged in physical activity. Similarly, Toker and Biron (2012) used a longitudinal panel design to show a reciprocal relationship between job burnout and depression over time with a weaker reciprocal effect among those who en gaged in the most physical activity. Thus, physical activity in this literature is typically investigated as a health outcome or for its health protective effects. 16 Neuropsychological research has viewed physical activity as an intervention , aiming to examine the direct effects of physical activity on executive function. Meta -analytic work by Ludyga et al. (2016) captures this area well, using the results from 40 studies to demonstrate that individual bouts of physical activity promoted executive functi on. Results indicated that physical activity did not selectively improve specific executive function s, finding broad benefits across measures of inhibition, shifting, and working memory. Further, benefits in executive function were not dependent on fitness level of participants in each study, suggesting that temporary improvements in executive function may be achieved regardless of experience or capability. An individual difference that did impact the effects of physical activity on executive function was age, whereby preadolescent and older adults experienced the greatest benefit from physical activity whereas changes experienced by adolescents and young adults were relatively modest. Ludyga et al. (2016) note that research on the physiological mechanism un derlying these effects is limited. However, increased blood flow to regions associated with executive functions appears to be a key part of the process. Based on the literature demonstrating the temporary effects physical activity has on executive function s, including working memory, I hypothesize that: H5 Œ Daily variation in physical activity predicts daily variation in state WMC . Ego Depletion and its Distinction from State WMC Past examinations of the impact of sleep and incivility on workplace effec tiveness explain these phenomena via ego depletion. For example, measures of ego depletion mediate the relationship between managers™ poor sleep and engagement in abusive supervisory behaviors (Barnes et al., 2015). Additionally, Rosen et al. (2016) connec t being the target of incivility to subsequent engagement in uncivil behavior through decrements in a performance measure of depletion . Ego depletion describes both the ability and costs of engaging in self -control, such as 17 constraining urges or impulses. Specifically, ego depletion theory describes how one™s ability to engage in self -control fluctuates over time. Engaging in self -control draws from a finite pool of resources and subsequent acts of self -control become more difficult as this pool of resources is depleted. Thus, ego depletion describes the state of the pool of resources used for self -control (Baumeister, Bratslavsky, Muraven, & Tice, 1998; Muraven & Baumeister, 2000). Sleep is thought to replenish this pool of resources, whereas poor sleep leaves one depleted and less able to engage in self -control (Barnes, 2012; Baumeister, Muraven, & Tice, 2000). Incivility, on the other hand, i s thought to consume self -regulatory resources in attempts to make sense of the uncivil experience, form or constrain responses, and manage the emotions that result from the experience (Rosen et al., 2016). Though ego depletion and state WMC are likely susceptible to similar factors and may covary, recen t empirical work suggests they are distinct . For state WMC to be of unique value, it cannot merely be a manifestation of ego depletion. Past empirical evidence does suggest a relationship between ego depletion tasks and working memory. A meta -analysis by Hagger, Wood, Stiff, and Chatzisarantis (2010) shows that tasks commonly administered to create an ego depletion effect can impact measures of attentional control and inhibitory control. Given how closely related attention and inhibition are to working me mory as executive functions (Diamond, 2013), it would be reasonable to suspect that ego depletion may impact working memory function ing . Carter, Kofler, Forster, and McCullough (2015) meta -analytically assessed the relationship between ego depletion and working memory tasks, finding that an effect was observable but did not remain significant after accounting for study size and publication bias. Recently, Singh and Göritz (2018) collected data from a large , heterogenous sample of adults (N = 1,385) to asses s the relationship between WMC and ego depletion. Ultimately, they found no relationship between ego depletion and 18 WMC task performance despite using multiple depletion tasks and controlling for implicit theories of willpower and trait self -control. Overal l, these r ecent meta -analytic and empirical studies show that changes in WMC task performance are not a reflection of a depletion effect. A recent study by Maranges, Schmeichel, and Baumeister (2017) sheds light on potential differences in the effects of ego depletion and state WMC by comparing depletion and cognitive load tasks. Cognitive load refers to procedures or tasks used to occupy part of one™s attention, leaving fewer attentional resources available for focal tasks (Sweller, Van Merrienboer, & Pa as, 1998). Working memory is the primary construct thought to be impacted by increased cognitive load (Engle, 2002; Maranges et al. 2017). Across four studies, Maranges et al. (2017) found that cognitive load and depletion tasks produced distinguishable ef fects. On a pain tolerance test, ego depletion led participants to focus more on pain and persist less in the task, whereas cognitive load produced the opposite effect. For emotional stimuli, cognitive load inhibited processing of stimuli and subsequent re actions whereas depletion led to more semantic processing of negative emotional stimuli. Maranges et al. (2017) interpret ed their findings as indicating that cognitive load undermine s attention and distract s from the task at hand, whereas ego depletion undermines self -control leaving one susceptible to negative emotional reactions and impulses. This study is especially pertinent given that increased cognitive load is the primary explanation for why incivility inhibits working memory via rumination (Porath & Erez, 2007; Porath et al. 2010). Thus, the findings from Maranges et al. (2017) imply that one™s experience s may result in demands that can be distinguished in terms of affecting cognitive or self -regulator y capabilities. Given this likely distinction between ego depletion and state WMC, it would be pertinent to investigate unique antecedents and outcomes of each state. As a motivational construct, ego depletion would likely be influenced by prior motivat ional states whereas state WMC would not. 19 A suitable motivational state to investigate among the students working towards academic goals in my study is state achievement motivation. Locke and Shcattke (2018) recently call ed for a stronger distinction to be made between intrinsic, extrinsic, and achievement motivation. They argue that the study of intrinsic motivation has thus far been too broad. In their view, intrinsic motivation refers specifically to the drive to experience pleasure from engaging in an a ctivity. Achievement motivation, on the other hand, refers to the drive to do well against some standard of excellence. Based on this distinction, Locke and Shcattke (2018) argue that one may not have the drive to improve one™s level of performance based o n intrinsic motivation alone , given that the pursuit of enjoyment can occur without the goal of improvement . However, improvement is a core aspect of achievement motivation, given that achievement motivation is conceptualized as a drive to push one™s curre nt performance closer to the standard of excellence they hold. Though seemingly external to the self, this standard of excellence should not conflate achievement motivation with extrinsic motivation. Locke and Schattke (2018) argue that extrinsic motivatio n describes the drive to do something as a means to an end, regardless of how close one is to excellence. Though students likely experience all forms of motivation throughout their time as a student, achievement motivation is likely a key form of motivatio n to influence daily ego depletion as students pursue educational goals. A state conceptualization of achievement motivation would be best aligned with ego depletion. Achievement motivation has typically been investigated as a trait, originally proposed by McClelland, Atkinson, Clark, and Lowell (1953). This trait conceptualization was originally applied to the study of entrepreneurship by McClelland et al. (1953), but research by Robbins et al. (2004) has shown that this trait has been widely investigate d as a predictor of retention and GPA among university students. Locke and Schattke (2018) describe how achievement 20 motivation has also been conceptualized as a quasi -trait in the form of goal orientation s, in that goal orientations are a somewhat stable i ndividual difference but are often thought of as situation -specific. Goal orientations have also been widely investigated in the academic context, with meta -analytic work demonstrating that mastery and performance -approach goal orientations relate to great er engagement in self -regulated learning strategies and ultimately higher academic achievement (Vrugt & Oort, 2008). Locke and Schattke (2018) point out that achievement motivation may also be conceptualized as a state, taking on both situation - and task - specific qualities. On a day -to-day basis, individuals may set goals related to the completion of specific tasks. Those in a higher state of achievement motivation may make better progress towards , or be more successful in attaining , task -specific goals an d may experience greater satisfaction as a result (Locke & Latham, 1990). State achievement motivation should predict subsequent ego depletion for several reasons. Of the conceptualizations of achievement motivation, a state conceptualization would be at the proper level of analysis to covary with ego depletion (Kozlowski & Klein, 2000). Beyond levels of analysis, s tudents who start the day experiencing a high state of achievement motivation should be more likely to manifest that motivation in the form of greater effort and persistence in the pursuit of course goals ( Kanfer, 1990; Locke & Latham, 1990). However, rather than effort leading to depletion as may be expected as an act of self -control (Baumeister et al. 1998) , students who are in a state of high achievement motivation may be more likely to have their efforts yield progress that they find energizing . Both the state of achievement motivation itself as well as subsequent acts in line with that motivation should l essen the likelihood of the fatigue -like state of ego depletion. For these reasons, it should be the case that high state achievement motivation at the start of the day helps explain why students may be less depleted during the day. 21 H6 Œ Daily variation i n achievement motivation negatively predicts daily variation in ego depletion . In addition to unique antecedents, outcomes that are motivational in nature should also be more aligned with ego depletion than state WMC . A key motivational outcome to investi gate would be the amount of effort put towards coursework throughout the day as effort is a behavioral manifestation of the motivational process (Kanfer, 1990). Additionally, effort is a behavioral outcome that may provide a clear view of the extent to whi ch students act in line with their present motivation. As students work towards course goals, several factors may be at play that determine how well they meet those goals , including the difficulty of the goal, whether the student has enough knowledge and s kill, and the presence of feedback (Locke & Shcattke, 2018). However, students will likely exert effort if they are motivated to do so even if other factors get in the way of meeting goals. Whether or not a student decides to put forth effort is likely det ermined by ego depletion. Effort expended should be determined by one™s state depletion if putting forth effort requires self -control. This should be the case because as students work towards course goals, they likely have other activities that they coul d pursue that are enjoyable but not productive. Findings by Maranges et al. (2017) demonstrating that depletion undermines self -control suggests that depleted students are more likely to pursue impulses and urges rather than putting forth effort towards th eir courses. Because of this, it should follow that: H7 Œ Daily variation in ego depletion negatively predicts daily variation in effort . If motivation is the unique nomological domain of ego depletion, outcomes in the domain of cognitive function ing should be uniquely related to state WMC . Cognitive failures describe slips or errors in routine tasks that require some form of cognition (Broadbent, Cooper, 22 Fitzgerald, & Park, 1982 ). Cognitive failures have been investigated across three broa d classes, false triggering of movement, distractability, and forgetfulness (Rast, Zimprich, Van Boxtel, & Jolles, 2009). Forgetfulness, or the tendency to fail to remember something known or planned, is incorporated into the model investigated here for tw o reasons. First, forgetfulness may be useful in investigating the impact of human error on performance. Human error has been investigated as causing serious performance errors in aviation, product manufacturing, and healthcare ( Hales & Pronovost, 2006 ). F urther, a common method to help reduce errors in these domains has been to employ a checklist. The rationale behind why checklists have been shown to be effective is that they reduce the likelihood of memory failures negatively impacting performance ( Hales & Pronovost, 2006 ). Understanding when memory failures are more likely, i.e. when people are more likely to be forgetful, should then be a useful target for understanding when performance failures may be more likely. The second reason for investigating fo rgetfulness is that if individuals vary in how forgetful they are on a day to day basis , state WMC should help explain why this variation occurs. A key function of WMC is to help individuals update their memory based on changing environmental circumstances (Miyake et al. 2000 ). Thus, when state WMC is poor, one may be more likely to forget to act on relevant information. H8 Œ Daily variation in state WMC negatively predicts daily variation in forgetfulness . The Role of State WMC and Ego Depletion during Performance Episodes Though state WMC and ego depletion are likely distinguishable, these states may also lead to common outcomes. Considering performance from an episod ic perspective highlights the unique importance of state WMC. I draw on Beal et al.™s (2005) notion of performance episodes for two reasons. First, Beal et al. (2005) describe how performance over time can be thought of in terms of episodes , or that individuals may work towards a performance goal during a s pecific 23 period of time . In my study, I focus specifically on performance in terms of academic goal progress , drawing a distinction between short -term and long -term course goals . Academic goal progress can be thought of as students working towards academic goals during specific episodes. Before engaging in these episodes, students may have to choose between completing short -term (e.g. studying for quizzes, completing homework assignments) or long -term coursework (e.g. studying for final exams, writing term p apers). The second aspect of Beal et al.™s (2005) theory I draw on is that an overall assessment of performance can be thought of in terms of a cumulative view of these episodes. Thus, assessments of daily academic goal progress reflect the overall success of episodes during a specific day. Here, I am hypothesizing daily variation in state WMC, thus the effects of state WMC on performance should be assessed using daily measurement of performance (i.e. a cumulative view of performance across episode (s) withi n a day). Though I am not hypothesizing that state WMC should vary meaningfully from episode to episode, thinking through a specific episode reveals how state WMC can contribute to performance. Beal et al. (2005) argue that effectiveness during a perf ormance episode is determined jointly by the available cognitive resources one can apply as well as the allocation of those resources to the task at hand. Cognitive resources are primarily described as attention and that any given time attention may be div ided between on -task and off -task factors (Beal et al., 2005) . Accounts of cognitive control or cognitive resources tend to treat the total amount of resources (or attention) available to an individual as static over time , determined by traits that vary between -person including general mental ability and WMC (Kanfer & Ackerman, 1989; Neal et al. 2017). When considering how the role of cognitive resources for predicting performance can vary within -person, researchers focus on the proportion of resources devo ted to performance 24 (Beal et al., 2005). Thus, these accounts argue that though the total amount of resources may vary between -person, it is only the share or proportion of these resources that may vary within -person over time or across tasks. A proportiona l conceptualization is helpful when considering how self -regulation shapes the application of available cognitive resources, assuming cognitive resources are constant. However, this conceptualization is lacking if cognitive resources vary, which I argue th at they do as a function of state WMC. At its core, my conceptualization of cognitive resources maintains the notion that resources are, at least in part, determined by WMC (e.g. Kanfer & Ackerman, 1989; Neal et al., 2017). The key change I propose is tha t state WMC, and the cognitive resources available for performing tasks during a given day , may vary within -person. If this is true , the same proportion of cognitive resources applied to a task as a function of self -regulation on one day may not represent the same amount of cognitive resources applied on another day. As a hypothetical example, an individual could have relatively poor state WMC on Monday ( perhaps due to a poor night of sleep) and have fewer cognitive resources to apply throughout the work day as a result . However, on Tuesday they could have relatively better state WMC (perhaps due to a good night of sleep), resulting in more cognitive resou rces. In this scenario, if this individual were to apply the same proportion of their total resources on each day, they would likely perform worse on Monday than on Tuesday due to the fewer resources that are available. Thus, state WMC should better repres ent the available cognitive resources that partly determine variability in performance rather than continuing to treat cognitive resources a s a static attentional capacity determined by individual differences (e.g. Kanfer & Ackerman, 1989; Neal et al. 2017). Building from this proposition, a more nuanced view of how state WMC can explain performance variability can be 25 gained by considering how state WMC may influenc e both the initiation of an episode as well as sustained performance throughout the episode. Initiating a performance episode can be difficult. Iqbal and Horvitz (2007) found that among Microsoft employees, disengaging from a primary task to respond to email led to a 25 -minute delay between the prior task episode and subsequent episode in which t hey reengaged in the primary task. These employees, on average, received email alerts four times per hour, suggesting that modern employees experience the process of reengaging in primary work tasks several times throughout the day. This process can be tho ught of in terms of multi -tasking, or the need to perform multiple tasks by shifting between tasks over time (Oswald, Hambrick, & Jones, 2007). Recently, Redick et al. (2016) demonstrated that working memory explains a meaningful amount of variance in a ba ttery of multi -tasking measures , variance that was distinct from measures of attentional control and fluid intelligence. This finding is likely due to functions attributed to working memory, including updating memory with novel relevant information as well as inhibiting memory of previously relevant information (Miyake et al . 2000). Blumberg et al. (2015) demonstrate how differences in WMC can impact multi -tasking using t ranscranial stimulation , a technique that can temporarily enhance neural function of a specific area. When neural areas associated with WMC are stimulated, individuals can return to a financial management task faster after being forced to engage with irrelevant math problems (Blumberg et al. 201 5). In other words , when initiating a performance episode , working memory should function to inhibit previously relevant information (e.g. the content of the work email) and aid in updating memory with information relevant to the performance episode, in thi s case the primary work task. As students choose to engage in performance episodes related to either short -term or 26 long -term academic goal s on a specific day , state WMC should describe their current ability to inhibit prior information and update memory wi th information relevant to the task at hand. State WMC should also aid performance throughout an episode by protecting task -relevant information held in mind from potentially distracting information . A key feature of WMC is the inhibition of task -irreleva nt information (Baddeley, 2003). This has been demonstrated in lab studies that eliminate differences in performance as a function of between -person differences in WMC (Kane, Bleckley, Conway, & Engle, 2001; Kane & Engle, 2003 ). Such studies have participa nts perform simple lab tasks in the presence or absence of distraction . When compared to tasks that incorporate distracting features, performance on distraction -free tasks bear little relationship with WMC. Distraction from the environment can extend beyon d the artificial visual and aural stimuli used in previous research to task -irrelevant information that is more similar to what is experienced in organizational settings. Across four studies, S chmeichel, Volokhov, and Demaree (2008) demonstrate d that between -person differences in WMC predict ed participants™ ability to suppress emotional expression in response to emotional stimuli . Further, WMC was unrelated to emotional expression when participants were not instructed to suppress their emotions, suggesting that WMC is related to successful goal maintenance rather than emotional stability. Hofmann , Gschwendner, Friese, Wiers, and Schmitt (2008) demonstrate d across three studies that those higher in WMC were better able to connect explicit (rather t han implicit) attitudes and personal goals to behaviors , such as resisting temptations and controlling emotion s in the presence of angering stimuli . These results suggest that WMC protects goal pursuit from distracting information in the environment. Addi tionally , distracting information can also influence performance when an individual™s mind wanders and focuses instead on task -irrelevant information. Mind -wandering 27 describes when attention has shifted away from some primary task to unrelated information, often without intention or initial awareness that attention has shifted (Smallwood & Schooler, 2006). Randall, Oswald, and Beier (2014) conducted a meta -analysis relating WMC, measures of mind -wandering, and task performance and found that WMC predicted t ask performance but was partially mediated through mind -wandering. In addition, t hese researchers conducted moderator analyses and demonstrated that longer task durations strengthened the relationship between WMC and task -related thought, suggesting that t hose high in WMC can sustain longer periods of performance without internal interruption. Further, improved performance as a function of task - related thought was stronger for complex tasks versus simple tasks. WMC seems to facilitate staying on task despit e distractions from a broad variety of sources, at least when investigated between individuals . State WMC should relate to daily variation in performance in a similar way if the functions of working memory are maintained at the within -person level of analy sis but may fluctuate in terms of capacity. I should reiterate that the discussion of performance episodes is not to suggest that state WMC should vary meaningfully from episode to episode, but to discuss performance in a way that highlights the role of WMC. Though fluctuation at shorter timescales may be possible, empirical examinations of within -person variation in WMC have revealed daily variation (Schmiedek et al., 201 3; Sliwinski et al., 2006) and the within -person causal effects of sleep and incivility I hypothesize have been examined daily (Barnes et al., 2015; Rosen et al., 2016). Because state WMC may contribute to performance across multiple episodes, I propose that a daily v iew of performance would be properly aligned with state WMC variation (Klein & Kozlowski, 2000). 28 Though initiating and sustaining performance should be uniquely supported by state WMC, these functions should be distinguished from self -control failures as a result of ego depletion and the inhibition of distraction . In fact, both working memory and self -control have been described as supporting performance through shielding goals from potential distractions (Hofmann, Schmeichel, & Baddeley , 2012; Johnson, Ch ang, & Lord, 2006; Johnson, Lin, & Lee, 2018). The spotlight analogy of attention as described by Maranges et al. (2017) helps distinguish the impact of ego depletion and cognitive load on performance. If attention is thought of like a spotlight, cognitive load should determine the circumference of the light (i.e. the amount of information that can be held in attention) whereas ego depletion describes if that light will be directed towards task -relevant features (i.e. the extent to which attention is on -tas k). As I have argued previously, if cognitive load describes the amount of attention currently occupied, state WMC should describe the remaining attention available for relevant tasks. Given the unique contributions of state WMC to performance, I hypothesi ze that: H9 Œ State WMC predicts daily variation in daily (a) short - and (b) long -term academic goal progress. H10 Œ State WMC partially mediates the relationship sleep quantity/quality , incivility , and physical activity have with daily (a) short - and (b) long -term academic goal progress. Though the spotlight analogy helps broadly distinguish the role of state WMC and ego depletion, the theoretical account by Hofmann et al. (2012) incorporates WMC into models of self -regulation . When considering self -reg ulated goal -pursuit, Hofmann et al. (2012) conceptualize executive functions (including working memory) as capabilities that support self -regulation. They summarize how models of self -regulated behavior are organized hierarchically 29 into three main components. The highest -order component describes how self -regulation requires some ideal or goal for an individual to hold to evaluate one™s thoughts, feelings, and behaviors. The second -order component is motivation because individuals mu st be motivated to reduce the discrepancy between their current state and the ideal or goal they hold. Finally, the third -order component is capability, in that individuals must be capable of the actions required to reduce that discrepancy. The hierarchy o f these components is meant to convey that higher order components must be operating for lower order components to have an impact on self -regulated behavior (Baumeister & Heatherton, 1996). For example, the second -order motivation component cannot promote self -regulation if there is no goal to guide behavior, nor can the third -order capability component impact behavior if there is not motivation to apply one™s capabilities. Hofmann et al. (2012) argue that models of self -regulation do not adequately specify what is meant by capability and suggest that research on executive functions can help bring clarity . Broadly, this is because executive function s represent the mental capabilities required to engage in self -regulated behavior, such as holding goals in min d or switching mental tasks in line with changing goals. Hofmann et al.™s (2012) integration of executive function and self -regulation theory, the spotlight analogy of attention (e.g. Maranges et al. 2017), and Beal et al.™s (2005) theory of episodic per formance all imply that ego depletion should moderate the impact state WMC has on performance. Should executive functions, including working memory, describe the capabilities that individuals rely on to engage in self -regulation as Hofmann et al. (2012) ar gue, then state WMC should reside in the lowe r-order component of self -regulatory models. Ego depletion, on the other hand, should be best represented within the motivation component. Should motivation constrain the impact capability has on performance as suggested by models of self -regulation 30 (e.g. Baumeister & Heatherton, 1996), then ego depletion should constrain the impact of state WMC. Via the spotlight analogy, the utility of the spotlight (or state WMC) depends on whether it is directed on -task (Mara nges et al. 2017). State WMC would have little impact on performance, regardless of level, if it is rarely directed toward the task at hand. However , variation in state WMC should bear a stronger relationship with performance when self -regulatory resources are high and attention is directed on -task more often. With attention directed on-task, variation in state WMC should be more tightly coupled with variation in performance. Beal et al. (2005) describe this dependency as episodic performance being jointly determined by cognitive resources and self -regulatory resources. Should one have more self -regulatory resources (i.e. are not depleted) then they should be able to direct more of their available cognitive resources to the task at hand. As I have argued abo ve, state WMC should be a better representation of available cognitive resources than a proportion of attention as argued by Beal et al. (2005). Updating Beal™s et al.™s (2005) theory, it should be the case that ego depletion determines the extent to which state WMC is directed on -task during a performance episode. In other words, s tate WMC represents within -person variation in the total amount of cognitive resources one could apply during a performance episode, whereas one™s self -regulatory resources (i.e. ego depletion) influence the extent to which state WMC is actively focused on -task versus off -task . The proposed moderating effect of ego depletion on the relationship between state WMC and performance has implication s for the hypothesized mediation effects involving sleep deprivation , the experience of incivility , and physical activity . As discussed pre viously, these antecedents are expected to predict performance indirectly through state WMC. If ego depletion moderates the relationship between state WMC and performance, then the mediation effects I 31 propose should also be moderated by ego depletion. If s tate WMC is affected by sleep deprivation , the experience of incivility, or physical activity, the effect on performance through state WMC may not be apparent if attention is rarely directed on -task due to being depleted. Conversely, if self -regulatory res ources are high and attention is directed on -task more frequently, the impact of these antecedents on performance through state WMC should be more apparent . Based on these ideas, I hypothesize that: H11 Œ Ego depletion will moderate t he within -person rela tionship between state WMC and (a) short - and (b) long -term academic goal progress such that this relationship will be stronger on days when students are less (vs. more) depleted . H12 Œ Ego depletion will moderate the indirect effe cts of physical activity, sleep quantity/quality, and incivility on (a) short - and (b) long -term academic goal progress through state WMC such that these indirect effects will be stronger on days when students are less (vs. more) depleted. Though state WM C likely contributes to both short - and long -term goal progress, ego depletion may best capture how students decide between exerting effort on goals of different temporal distances . Students often have multiple goals they can work towards in their courses based on course assignments. Some of these goals may have short -term deadlines, such as completing homework assignments and regular readings, whereas others have long -term deadlines like term papers and studying for final exams. Whenever students find time to work (i.e. begin a performance episode) they should have to decide whether to work towards short - or long -term goals. A key distinction between short - and long -term goals is the presence of an impending deadline. Studies of deadlines show that they can creat e a sense of time pressure and influence the perceived utility of pursuing a goal through temporal discounting (Hendy, Liao, & 32 Milgram, 1997; Steel & König, 2006). Time pressure refers to the sense of urgency in pursuing a goal due to having a limite d amount of time. The perception of time pressure increases as the time available to achieve a goal approaches the amount of time required to do so (Hendy et al., 1997). Temporal discounting, on the other hand, describes how individuals undervalue goal pursuit when the deadline of that goal is further in the future (Steel & König, 2006 ). Recently, Ballard, Vancouver, and Neal (2018) demonstrate d that both time pressure and temporal discounting influence goal prioritization when goals have differing deadlines. They conducted four studies that varied the presence of time pressure and temporal discounting and used the results of those studies to inf orm multiple competing computational models. They found that the model incorporating both temporal influences best described goal prioritization, suggesting that time pressure and temporal discounting have unique impacts on the prioritization process. Spec ifically, Ballard et al. (2018) conclude that time pressure increases the valence of specific goal s whereas temporal discounting influenced the perceived utility of selecting one goal versus another. Though multiple mechanisms may be at play, short -term de adlines seem to inspire motivation by influencing how the goal is perceived. The relative impact of ego depletion in pursuing short - versus long -term goals may then be the extent to which students must be self -motivated to pursue either goal. One would e xpect that depleted students would be less likely to pursue academic goals in general given that their pursuit requires self -control ( Baumeister et al. 1998 ). However, short -term goals may result in motivation among students regardless of a student™s state of depletion through the experience of time pressure and less susceptibility to bias in temporal discounting (Ballard et al. 2018, Henley et al. 1997, Steel & König, 2006 ). Thus, the impact of ego depletion on short -term goal progress among all factors th at influence motivation may be relatively small. The decision to make 33 progress towards long -term goals, on the other hand, should not be influenced by time pressure and should be perceived as having lower utility via temporal discounting. As a result, the impact of ego depletion on long -term goal progress may be relatively high given that there are few characteristics of the goal itself to inspire action. As a result, I would expect: H13 Œ Daily variation in ego depletion will be a stronger negative predictor of (a) long - term academic goal progress than (b) short -term academic goal progress . Between -Person Differences in Within -Person Relationships The relationships between state WMC and ego depletion at the within -person level of analysis likely vary due to between -person differences. This notion has been demonstrated previously in the study of ego depletion and how its influence at the within -person level of analysis may differ across in dividuals. One relevant individual difference is implicit theories of willpower, which describes how individuals differ in the belief that engaging in self -control draws from a limited pool of resources (Job, Dweck, & Walton, 2010). Those who view self -con trol as a limited resource tend to demonstrate greater depletion effects as a result of laboratory manipulations than those who view self -control as stable ( Job, Walton, Bernecker, & Dweck, 2013; Miller et al., 2012 ). Ma, Lin, Johnson, and Chang (2016) recently demonstrated that believing self -control is unlimited weakened the within -person relationship between commute stress, a depleting experience, and subsequent acts of transformational leadership. Johnson et al. (2017) note that recent research has found a similar moderating effect for trait self -control, such that high trait self -control helps buffer against temporary depletion (Johnson, Muraven, Donaldson, & Lin, 2017). The moderating effects of implicit theorie s of willpower and trait self -control should be specific to the within -person relationships between ego depletion and 34 the outcomes studied here. Demonstrating these unique effects of ego depletion should further distinguish the dynamics of ego depletion an d state WMC. H14 Œ Implicit theories of willpower will moderate t he within -person relationship between ego depletion and (a) short - and (b) long -term academic goal progress such that this relationship will be weaker for students who believe willpower is u nlimited (vs. limited ). H15 Œ Trait self -control will moderate t he within -person relationship between ego depletion and (a) short - and (b) long -term academic goal progress such that this relationship will be weaker for students who have high (vs. low ) trai t self -control . Trait self -control is thought to act as a buffer against momentary declines in self -regulatory resources, a mechanism that may also apply to trait and state cognitive resources. Trait self -control is conceptualized as the capacity for self -control or the amount of self -regulatory resources one has. Having a greater pool of self -regulatory resources minimizes the impact that temporary self -regulatory decrements may have (Johnson et al., 2017). This same logic can be applied to cognitive resources, such that those who have greater pools of cognitive resources should be less susceptible to temporary setbacks. Conceptualizations of cognitive resources at the between -person level of analysis describe both general mental ability and working memory as factors that influence the cognitive resources one has at their disposal (Kanfer & Ackerman, 1989; Neal et al. 2017 ). It could be the case that trait levels of WMC determine the extent to which decrements in state WMC impact performance o utcomes. However, as students work towards academic goals, it is more likely that a more general conceptualization of cognitive resources (i.e. general mental ability) would predict the impact of decrements in state WMC. General mental ability includes dim ensions like general knowledge, 35 reasoning, and long -term memory (McGrew, 2009). Thus, a student may rely on other aspects of general mental ability when state WMC is low and when pursuing academic goals. H16 Œ General mental ability will moderate t he wit hin -person relationship between state WMC and (a) short - and (b) long -term academic goal progress such that this relationship will be weaker for students who are high (vs. low ) on general mental ability . Finally, it may also be the case that the use of ti me management strategies may also buffer against the effects of low state WMC. This proposition is based on the research mentioned previously showing that checklists can help mitigate human error during performance by ameliorating the extent to which memor y failures can hurt performance ( Hales & Pronovost, 2006). Though not all goal s that a student may work towards are amenable to a checklist, students can use time management strategies to ensure that their time is being spent wisely. Further, meta -analytic work by Cred é and Phillips (2011) shows that students who frequently use time management strategies tend to achiev e higher GPAs. Part of the reason why students who manage their time well perform better is because they do not need to rely so heavily on state WMC to make progress towards course goals. Rather than consume cognitive resources with information about the t asks that need to be done during the day, the use of time management strategies may make it so that students can focus on their work and efficiently move on to new goal s when necessary. Conversely, students who do not use time management strategies may mak e less progress towards their goals because their cognitive resources are depleted as they process relevant information and maintain their goals in mind, hindering the influence of state WMC on performance . As such: 36 H17 Œ Time management strategies will m oderate t he within -person relationship between state WMC and (a) short - and (b) long -term academic goal progress such that this relationship will be weaker for students who habitually (vs. rarely) use time management strategies. I propose evaluating th e hypotheses I have developed here using the methods and analyses described below. 37 METHOD Participants and Recruitment Participants consisted of university students from a large midwestern university . They were enrolled in a psychology course and participated in this study to satisfy course requirements. 152 participants enrolled in the study . However, participants were excluded from the analyzed sample based on two criteria. First, participants who co mpleted 1 day or fewer of daily measurements (10 participants) were dropped from analyses. Second, scores from the numerical updating task (described further below) were used as an indicator of sufficient effort in responding. Participants whose average numerical updating scores across surveys were two standard deviations below the mean (8 participants, worse than 33% average accuracy on the task) were also excluded from analyses. Thus, responses from 134 participants ( Mage = 19.51, SDage = 1.26, 85.1% Fema le, 74.6% White ) were retained for further analysis. Participants provided responses to 2,921 of 3,752 (77.9%) possible measurement occasions. This resulted in 1,166 of 1,742 (66.9%) occasions where responses to all predictor, mediator, and outcome variabl es were provided. Participants were recruited via a research subject pool at a large midwestern university as part of a course they were taking . To be eligible for the study, participants had to indicate that they had normal (or corrected to normal) vision and were not impaired in terms of color vision . Participants were compensated with subject pool credit towards their course for completing an initial survey and for each day that they completed both the morning and afternoon survey. Procedures Particip ation began with an initial online survey . This survey contained measures of time management, implicit theories of willpower, and trait self -control. Participants were also asked 38 to report their standardized test scores. Included in this initial survey were practice trials of both the numerical updating and Stroop tasks, as well as further detail of the surveys sent out each morning and afternoon. Roughly one week after completing the initial survey, participants be gan receiving surveys emailed to them each morning and afternoon for 14 days. Morning surveys were sent at 7:00 am with a follow -up reminder at 10:00 am. This survey contained the predictor measures of sleep the prior night and achievement motivation. The morning survey also had participants retrospectively assess all criterion measures from the prior day, including effort, forgetfulness, and short - and long-term goal progress. The afternoon survey was emailed every day at 1:00 pm and a follow -up reminder s ent at 4:00 pm for those who ha d not participated. Incivility and reports of physical activity were included in the model as predictors but w ere assessed during this afternoon survey. Mediators were assessed at this point as well including performance meas ures of WMC and ego depletion, as well as subjectively reported ego depletion. Given the variability in student s™ schedules, participants were instructed to complete the survey as soon as was convenient after receiving the survey email. Measures 1 Sleep Deprivation . Sleep quality and quantity were measured by adapting four items from the Pittsburgh Sleep Diary (Monk et al. 1994). These items ask ed participants to report the time they went to bed, how long they thought it took to fall asleep, how long they spent awake after falling asleep initially, and the time they woke up. The item asking about time spent awake after falling asleep include d an ex ample from Barnes et al. (2015) to help participants respond, fiFor example, if you were asleep until 1 a.m., woke at 1 a.m. and fell back asleep at 1:20 a.m. for the rest of the night, your answer would be 20 minutesfl (p. 1425). This item assesse d 1 All self -report measure s and associated instructions are provided in the Appendix. 39 wakefuln ess after sleep onset and was used to compute total time asleep along with the reported times of sleep and waking. Barnes , Schaubroeck, Huth, and Ghumman (2011) provide evidence of self -reported measures of sleep quantity being useful indicators of objecti vely measured sleep quantity. The operationalization of sleep quantity and quality followed the procedure by Barnes et al. (2015). Sleep quantity was incorporated as the total time asleep (in minutes) whereas sleep quality was operationalized by reverse co ding the wakefulness after sleep onset question, which reflects the approach used in prior research (Wagner, Barnes, Lim, & Ferris 2012). Achievement Motivation. Achievement Motivation was measured using three items from the need for achievement subscal e in a measure of achievement motives ( Lang & Fries , 2006). Items were selected based on the factor loadings reported by Lang and Fries (2006) to identify the items that best assessed the Achievement Motivation construct. Participants were instructed to consider how they felt about the current day and rate the extent to which they agreed to each statement. An example statement is, fiToday, I look forward to tasks that let me test my abilities.fl This scale was highly reliable across days ( M = .97, SD = .01). Physical Activity. Physical activity was assessed using Godin and Shepard™s (1985) three item Leisure Time Exercise Questionnaire . Sallis and Saelens (2000) reviewed several self -report measures of physical activity and found that the measure developed by Godin and Shepard (1985) had similar validity evidence as other self -report measures . This measure was adapted to ask participants to report the number of minutes they spen t engaging in strenuous, moderate, and mild intensity exercise. Each le vel of intensity was provided a long with a brief description of how difficult it was as well as several example activities. Some e xample activities were modified in the present study because the original measure included activities that college students do not engage in frequently during the semester (e.g. cross -country skiing). Typically, this measure asks 40 participants to report the number of times they engage in each activity during a typical week and then responses are aggregated according to the followi ng formula: 9 × strenuous activity occasions + 5 × moderate activity occasions + 3 × light activity occasions (Godin, 2011). This scaling scheme was applied to the number of minutes participants reported from each intensity level of physical activity. Thus, physical activity was operationalized as the scaled number of minutes of physical activity each day according to the formula by Godin (2011) . Experienced Incivility. As discussed above, previous research demonst rate d that incivility in academic contexts may be either intentional or unintentional and is conceptually distinct when coming from faculty or peer actors (Caza & Cortina, 2007; Farrell et al., 2016). The present research targeted intentional uncivil behav iors to correspond with previous research on daily variability in experienced incivility (e.g. Rosen et al., 2016), but did not distinguish between faculty or peer actors. Experienced i ncivility was assessed using four items from Lim and Cortina (2005), wi th instructions to focus on the events of that day. An example item is, fiToday, someone put me down or was condescending to me.fl This measure showed high reliability across days (M = .86, SD = .05) . Working Memory Capacity. Working memory capacity was a ssessed daily using a numerical updating task (Miyake et al., 2000), administered according to the procedure investigated by Wilhelm, Hildebrandt, and Oberauer (2013). See Figure 2 for a graphical representation of this task. Each trial of this task presen ted participants with a list of digits to remember, serially presented updates (i.e. replacements) to individual digits, and ended with participants reporting the updated list of digits. Each trial began with 2 Œ 6 boxes containing a list of digits that pa rticipants were to remember presented on the screen for 2 seconds. The number of boxes on screen represents the load of the task, with trials containing more boxes representing 41 a higher load (higher difficulty). 12 trials were administered in total includi ng two trials where the initial number of digits were 2, 3, or 6 and three trials where the initial number of digits were 4 or 5. Updates involved an individual digit being presented on screen in the location of one of the existing digits from the list the participants had to remember. Participants then had to replace the number in mind at that location with the number presented during the update. For all trials across all su rvey administrations, the number of updates to specific digits were variable and random, ranging from 2 Œ 6. All updates were presented for 2 seconds. After all updates for a trial were presented, participants were prompted to report the updated list of di gits. Scores for each trial consisted of the proportion of correctly reported digits for each position. As an ability measure, modelling within -person variance in WMC requires additional modelling considerations to account for practice effects. In additio n to the theorized effects on variance in WMC, it is possible that scores improve over time as a function of greater familiarity with the task. Sliwinski et al. (2006) argued that failure to account for practice effects when examining within -person varianc e in ability measures may produce spurious relations with other variables at the same level of analysis. Daily variation in working memory performance was operationalized using a three -level multilevel modelling approach proposed by Schmiedek et al. (2013) , which is similar to the approach taken in multiple recent examinations of daily variation in WMC (Brose et al. 2012, 2014). This approach allows for the isolation of daily variation in WMC from practice effects and random differences in task difficulty. Further, this approach allows for the estimation of systematic and unsystematic variability in observed scores given that the numerical updating task is not amenable to internal consistency approaches to examining reliability due to variation in how the ta sk is constructed across measurement occasions. 42 Figure 2 . Graphical representation of the numerical updating task 43 Broadly, the three -level multilevel model involves modeling numerical updating performance by creating even -odd spl it halves for each measurement occasion at Level 1, within -person variation at Level 2, and between -person variation at Level 3. The following Level 1 equation describes how performance on each test half was estimated: WMC ijk = b0 + b1 (session ij) + b2 (session 2ij) + + b3(mobile ij) + eijk WMC ijk denotes performance on the numerical updating task for each individual i, on each day j, for each test -half k. Given the method of randomizing the number of updates across trials in this task, it could be the case that the difficulty of the task varied meaningfully across measurement occasions. To control for variation in task difficulty, Schmeidek et al. (2013) propose the inclusion of categorical dummy variables for each test half as fixed effects so that an individual™s performance is then estimated relative to the average performance of a specific test half. This is included in the above equation as . This term describes the fixed effects for a set of 27 categorical variables ( Zk) as dummy codes for the 28 test halves across 14 measurement occasions. Performance is further modeled via a random intercept b0, random slope b1 corresponding to the line ar number of sessions of the numerical updating task the participant has completed, and random slope b2 that consists of the squared number of completed sessions. The random slopes b1 and b2 control for practice effects using a quadratic polynomial trend similar to Sliwinski et al. (2006). It should be noted that a fisession fl here does not reflect the specific day of the study but the number of occasions that an individual has engaged with the task. This was implemented given the variety of response rates ac ross participants following the logic that practice effects should follow the number of times a participant has completed the task rather than the number of days pas sed in the study. Additionally, a categorial variable was included to control for the use o f either a mobile device or computer to complete a specific 44 afternoon survey and included as a fixed effect b3. Finally, the error variance associated with each test half is included as the term eijk. The intercept and slope terms for practice were composed of fixed and random effects as follows: b0 00 + u0i + v0j b1 10 + u1i b2 20 + u2i 00 10 , a20 represent the fixed intercept, linear, and quadratic terms for WMC score improvement, respectively. The random between -person effects for the intercept, linear, and quadratic terms are represented by u0j , u1j , and u2j , respectively and are estimated at Level 3 . The term of interest used to isolate within -person variation in WMC across days is v0j at Level 2 , which represents an individual™s average performance across test halves on each day j, after controlling for all the effects described above and modelling individual differences at Level 3 . In addition to using this model to estimat e variance components, daily variation in WMC was operationalized as the predicted daily variation term v0j and was incorporated into the substantive model s estimated to assess hypotheses. Though this multilevel modeling approach was used to operationalize WMC and assess reliab ility, this measure exhibited high reliability at each measurement occasions ( M = .94, SD = .02). Ego Depletion. Ego depletion was assessed using both self -report and performance measures. Via self -report, ego depletion was measured subjectively using t he 4-item scale developed by Lanaj , Johnson, and Barnes (2014), which used items from Twenge, Muraven, and Tice (2004). Participants were asked to indicate the extent to which they agree to items that describe how they currently feel using a 5 -point Likert scale (1 = very slightly or not at all, 5 = 45 very much). An example item is fiMy mental energy is running low.fl This scale exhibited high reliability across measurement occasions (M = .93, SD = .01) Ego depletion was also assessed based on performance on the Stroop task (Stroop, 1935) which has been used as a measure of ego depletion in past work (Job, Dweck, & Walton, 2010; Rosen et al., 2016). For this task, a single word is presented on screen that semantically signifies a color but is also presented in colored font. Participants are asked to report the color of the font while ignoring the semantic meaning of the word. For example, if the word fiGREENfl were presented on -screen in black -colored font a correct response would be to indicate the color black. This task engages self -control because the semantic meaning of the word is automatically processed when viewing the stimulus and participants must effortfully suppress this processing to produce the c orrect answer. In this task, ego depletion is operationalized using a participant™s reaction time when making a response as a reflection of the difficulty in engaging in self -control (Galliot, Schmeichel, & Baumeister, 2006) . The task characteristics wer e set up following the procedure used by Rosen et al. (2016). Four colors were used (black, blue, green, and orange) for the semantic meaning and font color of stimuli. All combinations of words and font colors were used as stimuli resulting in 16 unique items. Stimuli were presented sequentially, the order of which was randomized across measurement occasions. Once a word was presented on screen, participants were instructed to indicate the color of the word while ignoring its meaning as quickly as possible . The r eaction time for each trial was logged by the survey platform as the difference in time between when the stimuli was presented to when the response was made. Using reaction time data from participants, ego depletion was operationalized using a simil ar three -level modelling method described above for the numerical updating task (from Schmeidek et al., 2013). Trials were 46 randomly divided to create test halves to model . Unlike the numerical updating task three -level model, categorical variables were not incorporated to control for differences in difficulty as stimuli were similar across measurement occasions. However, practice effects and survey completion method were controlled for in a similar manner and the predicted daily variation in performance ter m was used to operationalize ego depletion. Thus, performance -based ego depletion was represented as one™s predicted deviation in reaction time on a given day after modelling within - and between -person differences in performance and controlling for practic e effects and administration method . The measure also demonstrated high observed reliability across measurement occasions ( M = .88, SD = .02). To distinguish these methods of assessing ego depletion in the following sections, responses to the self -report measure of ego depletion will be referred to as subjectively assessed ego depletion, whereas performance on the Stroop task will be referred to as performance -based ego depletion. Interpretation of the effects on either measure is the same in that higher scores reflect greater depletion. Short - and Long -Term Academic Goal Progress . Short - and long - term a cademic goal progress was assessed by adapting the three -item perceived work goal progress scale used by Rosen et al. (2018), based on the full measure by Koopman, Lanaj, and Scott (2016). When considering short -term goal progress, participants were instru cted to think about course activities with short -term deadlines and provided examples such as studying for quizzes and completing homework assignments. They then were asked to consider how much progress they made towards short -term goals the day before. Items consisted of statements reflecting goal progress such as , fiYesterday , I made good progress toward my short -term course goals ,fl to which participants rated the extent that they agreed. This scale exhibited high reliability across 47 measurement occasions (M = .85, SD = .04) . Similar items were used to assess progress toward long -term goals, but with modified instructions and item -wording to focus participants on these different goals. Specifically, participants were instructed to think of course activitie s with end -of-term deadlines and given examples such as studying for final exams and writing term papers. They were then asked to consider how much progress they made towards long -term goals the prior day using the same response format. An example item for this scale is fiYesterday, I made good progress toward my long -term course goals.fl This scale also exhibited high reliability (M = .87, SD = .03) . Forgetfulness. Items from the Cognitive Failures Questionnaire were adapted to assess daily variation in F orgetfulness (Broadbent, Cooper, Fitzgerald, & Cooper, 1982). Items related specifically to Forgetfulness were identified by Rast et al. (2009) and the three items were selected based on their high factor loadings and their potential for daily variability . For example, one item was excluded because it asked about difficulty remembering the names of people the participant recently met and participants in the present study may not have met new people each day of the study . Participants were instructed to report how often the y made the mistakes described in each item during the prior day. Examples of mistakes related to forgetfulness were provided as items such as , fiDid you find you couldn™t quite remember something although it's ‚on the tip o f your tongue ™?fl This measure showed acceptable reliability across days (M = .72, SD = .07) . Effort. Effort was assessed daily using four items from the Effort Regulation scale by Pintrich and DeGroot (1990). Items were adapted to assess effort directed towards coursework completed the prior day . Statements used to describe effort included, fiI worked hard to do well in my classes even if I didn™t like what we were doing.fl Participants rated how well each 48 statement described them the prior day . Acceptable levels of reliability were observed across days (M = .72, SD = .08). Implicit Theories of Will power. Beliefs about the nature of willpower were assessed using Job, Dweck, and Walton™s (2010) 12 -item Implicit Theories of Willpower scale. Participants were informed that the scale was designed to assess their own beliefs about willpower and that there were no right or wrong answers. Several statement s describing beliefs one could hold about the nature of willpower were provided for participants to respond to. An example item is , fiWhen you have been working on a strenuous mental task, you feel energized and you are able to immediately start with anothe r demanding activity .fl Participants then indicated the extent to which they agreed with each statement. Responses to this scale demonstrated good reliability 79). Trait Self -Control. Participants responded to the brief version of Ta ngney et al.™s (2 004) measure of trait self -control. Participants were asked to read several statements that could describe how they typically behaved. Each statement described different behaviors that reflect self -control such as, fi I refuse things that are bad for me ,fl to which participants would indicate how well each statement described them. This scale exhibited high reliability 85). General Mental Ability. Students™ standardized test scores were used as an indicator of general mental ability. Participants were f irst asked to report the highest composite ACT score they received before admission to their university. For students who did not take the ACT, SAT composite scores were requested and converted to the ACT metric using a concordance table provided by the Co llege Board (College Board, 2018). These scores were used to operationalize General Mental Ability. 49 Time Management Strategies. Use of Time Management Strategies was assessed using the seven -item scale developed by Drzakowski et al. (2005) . Participants were instructed to consider their time as a student during the current semester. They then rated how frequently they engaged in the behavior described by each question. An example question is, fiHow often do you start working on major school projects early (e.g., a final paper)? fl This scale did not exhibit good reliability 56), though follow up reliability analyses did not reveal specific items that could be omitted. As such the scale was included as intended. Control Variables . Class and work attendance, caffeine intake, and negative affect were assessed daily to ensure the veracity of the hypothesized effects. Given that students likely had to attend either class or work to experience incivility, measuring attendance was used to disti nguish the effect of experiencing incivility on ego depletion and state WMC from merely being an indicator of being in class. Participants were asked to report the number of hours of class and work they had scheduled, as well as how many they attended the prior day. Assessment of caffein e consumption was included given that caffeine intake may be more likely when students are sleep deprived or feel depleted. Further , caffeine consumption can impact working memory function (Chai, Abd Hamid, Abdullah, 2018). Caffeine consumption was assessed using one item, fiPrior to taking this survey, how many caffeinated drinks have you had today? Consider one caffeinated drink to be the equivalent of one 8 oz. cup of coffee or one small (8.4 oz) can of Red Bull Energy Drin k.fl Lastly, negative affect was also assessed daily as a control variable. Baumeister et al. (1998) argue that ego depletion should be distinguished from negative emotion. Further, Brose et al. (2012) found that WMC and negative affect covaried within -pers on. Finally, Beal et al. (2005) argue that negative affect may impact performance through hurting both cognitive and self -regulatory resources. Negative affect was measured using three negative affect 50 items from the Positive and Negative Affect Schedule (P ANAS; Watson, Clark, & Tellegen, 1988). Participants rated the extent to which they have specific feelings or emotions using a 5 -point scale (1 = very slightly or not at all, 5 = extremely). Feelings presented include d finervous, fl fiafraid, fl and fiupset ,fl whi ch were selected based on strong factor loadings provided by Thompson (2007). This scale exhibited high reliability across measurement occasions (M = .81, SD = .04) . Analytic Strategy The data I collected is nested (i.e. days within individuals), so multilevel path analysis was used to assess my hypotheses. For the substantive model s as well as all prerequisite steps I used Mplus version 8 (Muthén & Muthén, 2017). All within -person variables were assessed for within -person variability using a null model to ensure that multilevel analyses were appropriate, though multilevel modelling of WMC and performance -based ego depletion via the Stroop task were evaluated using a three -level model as describe d previously . Multilevel confirmatory factor analysis (CFA) of all measures using Likert scales was conducted to verify that items serve as meaningful indicators of the intended constructs ( e.g. Rosen et al., 2016). Model fit was assessed using the followin g rules of thumb: SRMR < .05, NNFI > .90, CFI > .90, and RMSEA < .08. It should be noted that these rules of thumb were used as guidelines as opposed to strict cutoffs as conditions may arise where an individual fit index may signal misfit unnecessarily (N ye & Drasgow, 2011; Schermelleh -Engel, Moosbrugger, & Müller, 2003). Thus, model fit was assessed holistically. Two multi -level CFAs were estimate d as simultaneous multi -level CFA of all Likert -based measures would result in a prohibitively complex model . Such a model cannot be identified as the number of parameters needed to estimate the model (191) exceeded the number of participants in the sample. The first model contained the predictor measures of state 51 achievement motivation and experienced incivility, negative affect (used as a control in further analyses), and the mediator of subjectively assessed ego depletion. This model fit the data well 2(142) = 311.82, RMSEA = .03, CFI = .98, NNFI = .98, SRMR [within] = .03, SRMR [between ] = .05) suggesting tha t these constructs were adequately distinguished using their respective items and that scale scores were appropriate to use for each construct in further analyses . The second model contained all self -report outcomes including forgetfulness, effort, and bot h short - and long -term academic goal progress. Initial estimation of this model resulted in a Heywood case, whereby the residual variance of one of the long -term academic goal progress items was negative. This variance was fixed to zero and the model was re -estimated, resulting in model fit that fell short of the aforementioned fit guidelines 2(119) = 1227.99 , RMSEA = .0 8, CFI = . 85, NNFI = . 81, SRMR [within] = .0 8, SRMR [ betwe en] = . 15), suggesting that item responses were not modelled well by their respective constructs as expected. However, examination of standardized residuals and modification indices suggested that the two reverse coded items for the effort scale may be the cause of model misfit. Inclusion of a residual correlation between these items at both levels of analysis meaningfully improved model fit 2(117) = 695.20, RMSEA = .06, CFI = .92, NNFI = .90, SRMR [within] = .06, SRMR [between] = .10) . Further, the resid ual correlation between these items seemed appropriate because of the reverse coding and the fact that both items discussed figiving up, fl which participants may have interpreted in a unique way 2. Thus, scale scores were used to operationalize outcome constr ucts . Model Specification. The multilevel path analytic models used to assess hypotheses were estimated using the Dynamic Structural Equation Modelling (DSEM) approach 2 The content of these items was, fiWhen class work was difficult, I gave up or only studied the easy parts,fl and fiI often felt so lazy or bored when studying for my classes that I quit before I finished what I planned to do.fl 52 implemented in Mplus version 8 ( Muthén & Muthén, 2017). This approach facilitates the analysis of time series data using the existing multilevel path analytic capabilities of Mplus. It should also be noted that Mplus mandates the use of Bayesian estimation for these analyses (Muthén & Muthén, 2017 ). Bayesian estimation can incorporate past information in the form of model prior s, though the analyses conduc ted here use the Mplus default of noninformative priors , resulting in estimated parameters based only on the collected data and specified model (Mut hén & Muthén, 2017 ). Of relevance to the present analyses, Bayesian estimation treats model parameters as random and the observed data as fixed, resulting in parameter values that are described by a probability distribution. This distribution is referred t o as a posterior distribution and the standard deviation of this distribution can be used much in the same way as standard errors are used in frequentist approaches ( Kruschke , Aguinis, & Joo , 2012). In these analyses, it is the standard deviation of the po sterior distribution that is used to form 95% Bayesian credibility intervals around parameter estimates, which can then be used to assess statistical significance. Further, this approach does not assume normality of estimated parameter distributions and ca n be used to assess indirect and moderated indirect effects without the need for additional procedures (Yuan & MacKinnon, 2009). This quality of Bayesian analysis applies to all the estimated effects, including indirect effects estimated here using Mplus™s Model Constraint function ( Muthén & Muthén, 2017). A broader discussion of Bayesian estimation and its implications can be found in Rosen et al. (2018). Using Bayesian estimation allows for models to be estimated that would not be identified using Frequ entist methods (Muthén & Asparouhov, 2012, Zyphur & Oswald, 2013). However, the ability to estimate more complex models via Bayesian estimation often requires the incorporation of priors. Zyph ur and Oswald (2013) suggest that priors should be obtained from 53 accumulated prior evidence, such as meta -analyses, which are not available for my hypothesized model at the within -person level of analyses. Even without model priors, Bayesian estimation provides greater flexibility in theory, but does not guarantee that all models will converge and be interpretable in practice (Muthén & Asparouhov, 2012, Zyphur & Oswald, 2013). Attempts at estimating my full hypothesized model w ere not successful using Mplus. Asparouhov (2018a) suggests that in cases of failure to attain model convergence, simplification of the model is the best approach, with the most impactful simplifications being the replacement of random effects with fixed effects. Further, Asparouhov (2018b) states that fixed effects modelled within the DSEM framewo rk should be nearly equivalent to the mean of random effects if those effects are specified solely at the within -person level of analysis. In addition to random effects at times being detrimental to model convergence in the DSEM framework, fixed and random effects produce meaningfully different computational costs (Asparaouhov, 2018a). For example, the entirety of the within -person structure of my hypothesized model including control variables (e.g. only omitting cross -level interactions) could reach conver gence within an hour using fixed effects. However, attempting to predict a single outcome with random effects from all studied antecedents, mediators, and control variables failed to converge after two days of processing with little evidence that convergen ce was near. I estimated multiple models to evaluate my hypotheses considering the relatively large costs of estimating random effects . Broadly, my model estimation strategy separates the estimation and evaluation of the within -person model I propose fro m the cross -level interaction effects. This approach is inspired by Rosen et al. (2018) who, in their study of the consequences of email demands on leadership behavior, evaluated within -person main effects in a primary model and between -person cross -level interactions in a secondary model. However, it should be noted that Rosen et al. (2018) 54 were able to estimate all within -person effects in their models as random and that the secondary model included all the within -person paths of their primary model. My m odel building approach uses a similar strategy of evaluating the within -person model initially, but relies on suggestions by Asparaouhov (2018a, b) to maintain sufficient model parsimony to achieve model convergence. Specifically, I estimated a primary mod el (Model 1) in which the entirety of my proposed model at the within -person level of analysis could be evaluated. However, to evaluate the model holistically I specified only fixed effects. My rationale is that fixed effects at the within -person level of analysis should be similar to the mean of random effects (Asparouhov, 2018b) and that investigating an overall within -person model is closer to the research goals I have here than estimating person -specific effects. Model 1 will be used to evaluate hypothe ses that are only at the within -person level of analysi s. The estimation of Model 1 is depicted in Figure 3. This model specified regressions of all outcomes (short -term and long -term goal progress, effort, forgetfulness) onto all mediators (state WMC, subjectively assessed ego depletion, performance -based ego depletion), antecedents (physical activity, experienced incivility, sleep quantity and quality, and state achievement motivation), and control variables (time effects, negative affect, caffein e consumption, and number of hours attended of class and work) 3. Time effects were estimated following the suggestions of Gabriel et al. (2018) of including a linear term that follows the days of the week, as well as sine and cosine functions for days of t he week to capture cyclical effects. Mediators were also regressed onto antecedents and control variables. Further, autoregressions were 3 An additional model was also estimated that was similar to Model 1 but included a squared term for daily variation in numerical updating task performance. In this additional model, regressions were specified from all outcomes onto this squared term. This was conducted to evaluate nonl inear relationships between state WMC and the outcomes studied here. No significant effects were found and, therefore, nonlinear relationships involving state WMC will not be discussed further. 55 Figure 3 . Graphical representation of Model 1 estimation. Variables grouped by their location in the theore tical model. An arrow from one group to another represents estimated paths between all variables in those groups. Autoregression boxes depict regression s of each variable in a group regressed onto i ts measurement from the prior day. 56 specifi ed for all outcome and mediator variables to control for levels of these variables from the prior day. Given that random effects appear to meaningfully increase model complexity, cross -level interactions will be evaluated in two separate models, one where the outcome is short -term academic goal progress and the other with long -term goal progress (Models 2 a nd 3, respectively) . These models contain a single outcome that is regressed onto all mediator, antecedent, and control variables listed above 4. Paths between antecedent and mediator variables were not estimated in these models. In addition to estimating p arsimonious models by focusing on a single outcome, each model included fixed paths for the same antecedents and control variables to ensure that the effects between mediators and outcomes correspond to those from the primary model (i.e. control for the sa me variables). Random effects were specified between mediator variables and the outcome to evaluate cross -level interactions. Further, only the autoregression of the outcome was included given that it was the only dependent variable in these models. The es timation approach of Models 2 and 3 is depicted in Figure 4. Given that no cross -level interactions were observed as hypothesized, no additional models were estimated to fully evaluate moderated mediation 5. 4 The same models were estimated without control variables. Re sults were consistent with and without control variables, thus only the results from models including control variables are presented. 5 An alternative model building strategy was also taken to evaluate my hypotheses. This alternative approach specified r andom effects for all regressions relevant to hypotheses, but omitted variables to achieve model parsimony. An initial model was estimated where the three mediators (state WMC, performance -based and subjectively assessed ego depletion) were regressed onto the five studied antecedents (physical activity, experienced incivility, sleep quantity and quality, and state achievement motivation). Four additional models were estimated, one for each outcome (short - and long -term academic goal progress, effort, forget fulness), where the outcome was regressed onto all three mediators and both within -person interaction terms. All random effects in these models were regressed onto the five between -person variables studied here (general mental ability, trait WMC, time -mana gement strategy use, trait self -control, and implicit theories of willpower). None of these models included control variables. However, conclusions about hypotheses drawn from this approach were the same as the model building strategy I focus on, thus this alternative approach is not discussed further. 57 Figure 4 . Graphical representation o f estimation of Models 2 and 3. The outcomes for Models 2 and 3 are Short -Term and Long -Term Academic Goal Progress, respectively. Variables grouped by their location in the theoretical model. An arrow from one group t o another represents estimated paths b etween all variables in those groups. Autoregression boxes depict regressions of each variable in a group regressed onto is measurement from the prior day. 58 All w ithin -person variables, except for state WMC and performance -based ego depletion, were person -mean centered so that effects observed at Level 1 represented within -person effects (Algina & Swaminathan, 2011). Isolation of within -person effects for state WMC and performance -based ego depletion was accomplished using the three -level models described previously. Variables used for investigating between -person effects were grand mean centered, including student™s WMC performance averaged across measurement occasions to represent trait levels of WMC. Following this approach, the estimation of these models required nearly all variables to be rescaled for model estimation to converge. Model convergence becomes difficult in Mplus when variables are measured on different scales resulting in meaningfully different variances (Muthén & Muthén, 2017 ). The scale of variables in the present study ranged from percentages reflecting accuracy on the numerical updating task to hundreds of minutes of sleep quantity . Following the recommendation by Muthén and Muthén (2017) each variable was mu ltiplied by a constant such that all variances ranged from 1 to 10. This approach facilitated model convergence but resulted in all variables being represented in different units than they were originally measured. As such, the potential transparency of un standardized results is obscured by the idiosyncratic scaling factors used for the variables studied here. To avoid issues in the interpretation of effects, all effects presented are standardized unless noted otherwise. The default standardization procedur es in Mplus were used, which primarily concern the method of standardizing within -person versus between -person effects. Following the rationale provided by Schurrman , Ferrer , de Boer -Sonnenschein , and Hamaker (2016), Mplus standardizes multilevel effects u sing the variance at the level of analysis of the effect itself. In other words, variable variance at the within -person level of analysis is used to standardize within -person effects, 59 whereas between -person variance is used to standardize between person ef fects ( Muthén & Muthén, 2017). 60 RESULTS Table 1 contains the means, standard deviations, and between -person intercorrelations of all studied variables. Intercorrelations at the within -person level of analysis are provided in Table 2. Of note, at the wit hin -person level of analysis state WMC was found to correlate with physical activity ( r = .11, p < .01), performance -based ego depletion ( r = -.07, p < .05), and short -term academic goal progress ( r = .06, p < .05). State WMC was not found to significantly correlate with subjectively assessed ego depletion ( r = -.03, ns). Correlations between state WMC and measures of ego depletion suggest that constructs are largely independent given the nonsignificant relationship with subjectively assessed ego depletion and weak correlation with performance -based depletion . Further, m any of the expected relationships with state WMC were not observed. Subjectively asses sed ego depletion correlated with several antecedents and outcom es, though did not correlate with performance -based ego depletion ( r = .01, ns). Performance -based ego depletion did not correlate with any antecedent or outcome that was related to subjectively assessed ego depletion. The pattern of relationships between both measures of ego depletion were largely distinct. Regarding outcome measures, Short - and Long -Term goal progress were highly correlated within -person ( r = .58, p < .01), suggesting some distinction but perhaps not as much as would be expected if these two outcomes were capturing a trade -off in the decision -making process. Table 3 contains the results of the three -level models for state WMC and performance -based ego depletion and null models for all other measures used to assess the percentage of varianc e at within - and between -levels of analysis. An important purpose of the three -level model s was to estimate the extent to which performance var ies systematically across days. 61 Table 1. Between -Person Descriptives and Intercorrelations among Studied Variables Variable M SD 1 2 3 4 5 6 1) Physical Activity 225.96 233.98 2) Experienced Incivility 1.40 .48 .34** 3) Sleep Quantity 449.61 62.98 -.07 -.04 4) Sleep Quality -10.42 11.46 -.01 .01 .06 5) Achievement Motivation 3.37 .75 .18* -.05 .20* .13 6) WMC .79 .21 -.11 -.36** -.16 -.09 -.07 7) Subjective Ego Depletion 2.38 .76 .22* .43** -.20* -.17 -.27** -.29** 8) Performance Ego Depletion 1642.44 235.64 .04 .07 .04 .07 .01 -.33** 9) Forgetfulness 1.62 .51 .11 .55** -.08 -.08 -.06 -.22* 10) Short -Term Goal Progress 3.43 .44 .03 -.19* .03 .03 .33** .15 11) Long -Term Goal Progress 3.14 .61 .14 -.08 .07 .11 .41** -.03 12) Effort 3.58 .55 .10 -.25** .00 .01 .42** .24** 13) Time Management 3.61 .49 .16 -.01 .19* -.02 .28** .00 14) Standardized Test Scores 25.54 3.95 .09 .05 -.16 -.20* -.11 .26** 15) Trait Self -Control 3.25 .68 .13 -.15 .15 .05 .30** .05 16) Implicit Theories of Willpower 2.95 .50 .04 -.13 .13 .03 .21* -.08 Note. Daily variables were aggregated to estimate between -person correlations. * p < .05, ** p < .01. After including regressions to control for practice effects, variation in task difficulty, and method of task completion, var iance components remain ed for test -halves (Level 1), days (Level 2), and individuals (Level 3). Schmeidek et al. (2013) suggest that the observed variance at Level 1 must be divided by the number of trials or blocks that are investigated at that level to be comp arable to the variance components at other levels, in this case the variance must be divided by two for each test half. 62 Table 1. (cont™d) Between -Person Descriptives and Intercorrelations among Studied Variables Variable 7 8 9 10 11 12 13 14 15 1) Physical Activity 2) Experienced Incivility 3) Sleep Quantity 4) Sleep Quality 5) Achievement Motivation 6) WMC 7) Subjective Ego Depletion 8) Performance Ego Depletion .19* 9) Forgetfulness .35** .17 10) Short -Term Goal Progress -.26** -.06 -.18* 11) Long -Term Goal Progress -.19* -.02 -.12 .59** 12) Effort -.32** -.12 -.21* .73** .48** 13) Time Management -.19* -.04 -.10 .43** .20* .38** 14) Standardized Test Scores .01 -.25** -.01 -.02 -.05 .00 .07 15) Trait Self -Control -.25** .06 -.31** .32** .20* .39** .62** -.02 16) Implicit Theories of Willpower -.22* .05 -.01 .17 .21* .24** .20* -.10 .39** Note. Daily variables were aggregated to estimate between -person correlations. * p < .05, ** p < .01. This approach suggests that 81% of the variance in state WMC and 68% of the variance in performance -based ego depletion was systematic. Overall, these analyses reveal that all measures examined here exhibited enough within -person variance to justify multil evel analyses. 63 Table 2. Within -Person Correlations among Daily Variables Variable 1 2 3 4 5 6 7 8 9 10 11 1) Physical Activity 2) Experienced Incivility .08** 3) Sleep Quantity -.04 -.04 4) Sleep Quality .03 -.05 .10** 5) Achievement Motivation .11** -.01 .05 .02 6) WMC .11** -.05 -.01 .00 .02 7) Subjective Ego Depletion .03 .14** -.05 -.05* -.14** -.03 8) Performance Ego Depletion .04 -.04 -.05 .02 -.01 -.07* .01 9) Forgetfulness .00 .10** -.04 -.01 .00 -.05 .14** .01 10) Short -Term Progress .11** -.03 .00 .10** .15** .06* -.07* -.01 -.03 11) Long -Term Progress .09** .01 -.01 .04 .14** .01 -.03 .01 -.03 .58** 12) Effort .09** -.01 -.03 .04 .21** .03 -.09** .03 -.05 .50** .43** Note. Within -person correlations are among person -centered daily variables except for WMC and performance ego depletion. WMC and Performance ego depletion are operationalized as the predicted daily deviations as described in the Method section for ea ch var iable. * p < .05, ** p < .01. Test of Hypotheses As described above, Model 1 estimated the hypothesized model at the within -person level of analysis and was used to evaluate hypotheses at this level. The results of this model are presented in Table 4. Hypothesis 1 was not supported, as neither sleep quality ( = -.01, ns) nor sleep quantity ( = -.01, ns) were significant predictors of daily variation in state WMC as expected. These results suggest that poor sleep on a given day has little impac t on state WMC. Further, performance -based ego depletion was also not predicted by either sleep quality ( = .02, ns) or quantity ( = -.05, ns).64 Table 3. Percentage of Variance Within -Person among Daily Variables Construct Within -Person Variance ( 2within ) Between -Person Variance ( 2between ) % of Within -Person Variance Physical Activity 41719.00 49117.00 46% Experienced Incivility .31 .20 61% Sleep Quantity 15646.00 1570.00 91% Sleep Quality 256.40 141.50 64% Achievement Motivation .47 .50 48% WMC .11 .18 38% Subjective Ego Depletion .79 .47 63% Performance Ego Depletion 51.00 163.80 24% Forgetfulness .23 .22 51% Short -Term Goal Progress .64 .13 84% Long -Term Goal Progress .51 .33 61% Effort .42 .25 62% Note . The percentage of variance within -individuals was calculated as 2within /(2within + 2between ). The within person variance component for all variables was statistically significant (p < .05). Within -person variance ( 2within ) for WMC and Performance ego depletion estimated using the three -level approach by Scmiedek et al. (2013). Using this approach , estimated error variance associated with daily measurement (WMC 2test -half = .51; Performance Ego Depletion 2test -half = .59) can be used to estimate the percent of variance in daily measurements that is systematic according to the following formula: 2within /(2within + 2test -half /2). According to this approach, the percentage of observed variance in daily variation in measurement that is systematic is 81% for WMC and 68% for Performance Ego Depletion. All variance components for these measures are estimated after controlling for practice effects and mobile versus computer responding (as well as difficulty for WMC). This pattern of results was the same for subjectively assessed ego depletion, as neither sleep quantity ( = -.02, ns) nor sleep quality ( = -.05, ns) were significant predictors. Thus, Hypothesis 2 was not supported. These resu lts suggest that nights of low quantity or quality of sleep have little impact on within -person variation in cognitive or self -regulatory resources. Hypothesis 3 prop osed that experienced incivility would negatively predict daily variation in state WMC. This hypothesis was not supported either as incivility did not significantly predict daily variation in state WMC ( = -.04, ns). 65 Table 4. Path Analytic Results from Model 1 Predictor State WMC Subjective Ego Depletion Performance Ego Depletion Short Term Goal Progress Long Term Goal Progress Effort Forgetfulness Antecedents Physical Activity .10* (.0 3) .04 (.0 3) .05* (.0 3) .06* (.0 3) .05 (.0 3) .03 (.0 3) -.01 (.03) Experienced Incivility -.04 (.03) .05* (.0 3) -.04 (.0 3) -.03 (.03) .01 (.0 3) -.02 (.0 3) .04 (.03) Sleep Quantity -.01 (.0 3) -.02 (.0 3) -.05 (.0 3) -.00 (.0 3) -.01 (.0 3) -.03 (.0 3) -.02 (.0 3) Sleep Quality -.01 (.0 3) -.05 (.0 3) .02 (.0 3) .08* (.0 3) .03 (.0 3) .03 (.0 3) -.00 (.0 3) Achievement Motivation .00 (.0 3) -.12* (.0 3) .01 (.0 3) .11* (.0 3) .11* (.0 3) .18* (.0 3) .03 (.0 3) Mediators State WMC .05 (.03) .00 (.03) .03 (.0 3) -.04 (.03) Subj . Ego Depletion -.04 (.0 3) -.01 (.0 3) -.07* (.0 3) .10* (.0 3) Perf . Ego Depletion .01 (.03) .03 (.0 3) .05 (.0 3) .01 (.03) Interactions WMC x Subj. Depletion -.02 (.0 3) -.02 (.0 3) -.01 (.0 3) .01 (.0 3) WMC x Perf. Depletion .01 (.0 3) .01 (.0 3) -.00 (.0 3) -.02 (.0 3) Controls Autoregression -.09* (.03) .06* (.0 3) -.11* (.03) .01 (.0 3) -.03 (.0 3) -.04 (.03) .02 (.03) Day .00 (.01) .01 (.0 2) -.01 (.01) -.03 (.0 2) -.03 (.0 2) -.04* (.01) -.00 (.0 2) Sine -.02 (.0 3) -.02 (.0 3) -.01 (.0 3) .17* (.0 3) .09* (.0 3) .16* (.0 3) .06* (.0 3) Cosine -.02 (.03) .02 (.03) .07* (.03) -.11* (.0 4) -.17* (.0 4) -.17* (.03) .04 (.0 4) Caffe ine Consumption .07* (.0 3) -.01 (.0 3) -.05 (.0 3) .04 (.0 3) .06* (.0 3) .04 (.0 3) -.03 (.0 3) Negative Affect -.04 (.0 3) .26* (.0 3) -.02 (.0 3) -.01 (.03) .00 (.03) .05 (.03) .15* (.0 3) Hours of Class -.05 (.03) -.04 (.03) -.06 (.03) .02 (.03) -.02 (.0 4) -.01 (.03) .01 (.0 4) Hours of Work -.02 (.0 3) -.01 (.0 3) .02 (.0 3) -.02 (.0 3) -.01 (.0 3) -.03 (.0 3) .01 (.0 3) Note . Estimates provided reflect standardized path estimates with Posterior S.D. ™s provided in parentheses. * p < .05. 66 Experienced incivility predicted subje ctively assessed ego depletion ( = .05, p = .05) but did not predict performance -based ego depletion ( = -.04, ns). Thus, Hypothesis 4 was supported. These results suggested that the experience of incivility is not related to daily fluctuation in the abi lity to store and process information as expected , but does lead to depletion . Engagement in physical activity was thought to promote state WMC, as described in Hypothesis 5. This hypothesis received support as engagement in physical activity was positiv ely related to state WMC ( = .10, p < .01). Physical activity was not related to subjectively assessed ego depletion ( = .0 4, ns) but was related to performance -based ego depletion ( = .0 5, p < .05). These results suggest that physical activity promotes greater momentary cognitive resources . It may also be the case that physical activity drains self -regulatory resources , but this relationship appears to be clearer for performance -based ego depletion than when it is subjectively assessed . Hypothesis 6 suggested that state achievement motivation would predict ego depletion . This hypothesis appears to be supported in that state achievement motivation negatively predicted subjectively assessed ego deple tion ( = -.12, p < .01). However, state achievement motivation did not predict performance -based ego depletion ( = .0 1, ns). As expected, state achievement motivation was also not a significant predictor of state WMC ( = .0 1, ns). These results suggest that state achievement motivation leads to a reduced degree of subjectively experienced depletion later in the day but has little impact on state WMC. Further , the relationship between state achievement motivation and ego depletion also appears to be depen dent on assessment method, as was the case for engagement in physical activity and experienced incivility . Hypothesis 7 suggested that daily variation in ego depletion would negatively predict daily variation in effort. Subjectively assessed ego depletion was negatively related to effort ( = 67 -.07, p < .05) whereas performance -based ego depletion was not significantly related to effort ( = .0 5, ns). Thus , Hypothesis 7 was supported , suggesting that students may exert less effort on their coursework if they feel like their self -regulatory capabilities are inhibited . When comparing methods of assessing ego depletion, it appears as though subjectively assessed ego depletion best captures this relationship. As expected, variation in state WMC did not appear to be related to the exertion of effort ( = .05, ns). Thus, the motivational outcome of effort is predicted by constructs related to self -regulatory resources more so than cognitive resources. Hypothesis 8 pro posed that daily variation i n state WMC should negatively predict daily variation in forgetfulness. Contrary to expectation, state WMC was not a significant predictor of forgetfulness ( = -.05, ns). Thus, Hypothesis 6 was not supported. Students ™ forgetful ness did not depend on thei r momentary ability to store information in mind . Ego depletion was not expected to predict forgetfulness, though subjectively assessed ego depletion was in fact found to be a significant predictor of this outcome ( = .10, p < .01). Performance -based ego depletion did not predict forgetfulness, which again appears to demonstrate that subjectively assessed ego depletion better captures the relationship between depletion and this outcome. Hypothes es 9a and 9b suggested that state WMC s hould predict daily variation in both short - and long -term academic goal progress , respectively . However, state WMC was not found to be a significant predictor of either outcome (short -term academic goal progress: = .05, ns; long -term academic goal progr ess: = .00, ns). Hypothesis 10 suggested that sleep quantity and quality, incivility, and physical activity would relate to (a) short - and (b) long -term academic goal progress indirectly through state WMC. Table 5 p resents the estimated indirect effects, though none of the hypothesized effects were significant. Thus, Hypothes es 10a and 10b were 68 not supported. No evidence was found for the idea that state WMC would help explain goal progress or determine the behavioral and experiential factors that could i mpact performance. Table 5. Indirect Effect Estimates from Model 1 Predictor Indirect Effect State WMC Subj. Ego Depletion Perf. Ego Depletion Short -Term Goal Progress Experienced Incivility -.00 (.00) -.00 (.00) .00 (.00) Physical Activity .01 (.00) -.00 (.00) .00 (.00) Sleep Quantity -.00 (.00) .00 (.00) .00 (.00) Sleep Quality .00 (.00) .00 (.00) .00 (.00) Achievement Motivation .00 (.00) .01 (.00) .00 (.00) Long -Term Goal Progress Experienced Incivility .00 (.00) -.00 (.00) -.00 (.00) Physical Activity .00 (.00) .00 (.00) .00 (.00) Sleep Quantity .00 (.00) .00 (.00) -.00 (.00) Sleep Quality .00 (.00) .00 (.00) .00 (.00) Achievement Motivation .00 (.00) .00 (.00) .00 (.00) Effort Experienced Incivility -.00 (.00) -.00 (.00) -.00 (.00) Physical Activity .00 (.00) -.00 (.00) .00 (.00) Sleep Quantity .00 (.00) .00 (.00) -.00 (.00) Sleep Quality .00 (.00) .00 (.00) .00 (.00) Achievement Motivation .00 (.00) .01* (.00) .00 (.00) Forgetfulness Experienced Incivility .00 (.00) .00 (.00) .00 (.00) Physical Activity -.00 (.00) .00 (.00) .00 (.00) Sleep Quantity .00 (.00) -.00 (.00) .00 (.00) Sleep Quality .00 (.00) -.00 (.00) .00 (.00) Achievement Motivation .00 (.00) -.01* (.00) .00 (.00) Note . Estimates are unstandardized indirect effect estimates with Posterior S.D. ™s provided in parentheses. Italicized effect estimates were hypothesized. Column headings under Indirect Effect represent estimates through a given mediator. Results presented below the headings of each outcome represent the indirect effect from each pred ictor through the mediator of a given column onto that outcome. * p < .05. Hypothesis 11 suggested that the relationship between state WMC and both (a) short - and (b) long -term academic goal progress would be stronger on days when students were less depleted. Interaction terms between subjectively assessed ego depletion and state WMC were not significant predictors of either short -term ( = -.02, ns) or long -term academ ic goal progress ( = 69 -.02, ns). The same was true for interaction terms computed using performance -based ego depletion and state WMC (short -term academic goal progress: = .01, ns; long -term academic goal progress: = .01, ns). Consequently, Hypothesis 11 was not supported. Thus, it does not appear as though the successful application of momentary cognitive resources depend s on presently available self -regulatory resources. G iven that the interaction terms between subjectively assessed ego depletion and state WMC were not significant, the moderated mediation effect s posed by hypothesis 12 were not assessed and this hypothesis was not supported. Hypothesis 13 pro posed that daily variation in ego depletion would be a stronger negative predictor of (a) long -term rather than (b) short -term academic goal progress. Subjectively assessed ego depletion was not found to be a significant predictor of either long -term ( = -.01, ns) or short -term -academic goal progress ( = -.04, ns). Performance -based ego depletion also did not predict either outcome ( long -term academic goal progress: = .0 3, ns; short -term academic goal progress = .0 1, ns). Given the lack of observable effects, a follow -up model was not estimated to directly test the relative strength of the relationship s between ego depletion and both long - and short -term academic goal progress . Hypothesis 13 was not supported. As such, evidence was not found for the pr oposition that the temporal characteristics of long -term goals would require greater self -regulatory resources to work toward. Cross -level interactions were estimated and used broadly to distinguish the within -person relationships between ego -depletion , state WMC , and short - and long -term goal progress. Hypotheses related to cross -level interactions were evaluated using Models 2 and 3, as described previously. The results from these models are presented in Table 6. Given that past work has demonstrated that individual differences in trait self -control and implicit theories of willpower 70 can moderate the within -person relationships ego depletion has with other outcomes (Johnson et al., 2017, Ma et al., 2016) , it was expected that these individ ual differences would moderate the relationship between ego depletion and both academic goal progress outcomes. This was not found to be the case for either cross -level moderator . Thus, Hypotheses 1 4 and 1 5, which suggested that implicit theories of willpo wer and trait self -control would moderate the within -person relationships between ego depletion and goal progress outcomes, respectively , were not supported . Further, I was not able to replicate effects suggesting that the within -person relationship betwee n ego depletion and subsequent behavior depend s on whether individuals have high trait self -control or believe that self -regulatory resources are limited. Regarding state WMC, Hypothesis 1 6 specified that the within -person relationship between state WMC and (a) short - and (b) long -term academic goal progress would be weaker for those with greater general mental ability. General mental ability did not moderate the relationship between state W MC and short -term ( = -.06, ns) or long -term goal progress ( = -.02, ns). Trait levels of WMC also did not moderate either relationship. Thus, Hypothesis 1 6 was not supported and evidence was not found for the notion that trait levels of cognitive resour ces could buffer against state decrements in resources. Hypothesis 17 also posed a cross -level interaction in which time management strategies would mitigate the negative effects of poor state WMC on both (a) short - and (b) long -term academic goal progress . Time management strategy use did not moderate the relationship between state WMC and short -term ( = .15, ns) or long -term academic goal progress ( = .14, ns). Thus, Hypothesis 17 was not supported . Further, it was expected that greater use of time management strategies would weaken the relationship between state WMC and goal progress, though these effects were in the opposite direction. 71 Table 6. Cross -Level Interaction Results from Models 2 and 3 Predictor Cross -Level Moderator Trait WMC General Mental Ability Time Management Theories of Willpower Trait Self -Control Short -Term Goal Progress State WMC .10 (.24) .07 (.26) .16 (.35) -.40 (.27) .01 (.32) Subj . Ego Depletion .05 (.18) -.15 (.18) -.09 (.22) -.05 (.18) .19 (.22) Perf . Ego Depletion -.02 (.22) -.10 (.24) .25 (.36) -.20 (.27) .20 (.36) Long -Term Goal Progress State WMC .06 (.0 5) -.02 (.05) .14 (.0 9) -.09 (.0 9) -.07 (.06) Subj . Ego Depletion -.01 (.01) -.01 (.01) -.02 (.0 3) -.00 (.02) .02 (.0 2) Perf . Ego Depletion .01 (.06) -.02 (.07) .05 (.11) -.05 (.1 1) -.06 (.0 9) Note . Results presented below Short -Term Goal Progress are the cross -level interaction effects estimated between the studied mediators and Short -Term Goal Progress in Model 2. The corresponding effects for Long -Term Goal Progress from Model 3 are presented below the Long -Term Goal Progress heading. Estimates provided are standardized path estimates with Posterior S.D. ™s provided in parentheses. Estimates are of the random slope between a given mediator and outcome regressed onto each cross -level moderator. None o f the presented effects are statistically significant. It may be the case that having a plan helps facilitate the pos itive impact state WMC can have on performance, but this notion is highly speculative given the observed results. In addition to an evaluation of hypotheses, observed direct effects of the antecedents and control variables are summarized her e as they may have theoretical and practical implications. These effects are from Model 1 which focused on the within -person level of analysis. 72 Broadly, state achievement motivation appeared to be an important determinant of productivity. In addition to pr edicting ego depletion as discussed above, state achievement motivation was a direct predictor of effort ( = .18, p < .01), short -term ( = .11, p < .01), and long -term academic goal progress ( = .11, p < .01). These results are notable given that they a re unique effects among several other variables in the model , as well as the fact that state achievement motivation was among the most temporally distant predictors of these outcomes. The use of negative affect as a control variable also appeared to be impactful, as negative affect was a predictor of subjectively assessed depletion ( = .26, p < .01) and forgetfulness ( = .15, p < .01). The inclusion of negative affect helps demonstrat e that the effects of incivility on ego -depletion and the effects of ego depletion on forgetfulness were self -regulatory rather than affective in nature (Baumeister et al. 1998) . Further, the fact that both depletion and negative affect predicted forgetfulness suggests motivational and affective mechanisms through which everyday memory functioning may be disrupted. Th ese findings contrast with the hypotheses proposed here that forg etfulness would primarily be determined by cognitive mechanisms. Additionally, negative affect was a relatively strong predictor of subjectively assessed ego depletion, but not performance -based ego depletion ( = -.02, ns), further distinguishing these measurement methods. Caffeine consumption predicted state WMC ( = .07, p < .01), which is useful to note as it provides additional evidence that variation in state WMC is systematic and predictable . Sleep quality also directly predicted short -term goal pro gress ( = .08, p < .01), suggesting that the benefits of sleep may not have been explained using the other measures in my study. Further, it should be noted that the outcomes investigated here all demonstrated evidence of weekly trends 73 and/or cyclical eff ects. Thus, controlling for these temporal effects was useful for the evaluation of hypothesized relationships and may be worth investigating directly in future research. 74 DISCUSSION The organizational sciences predominantly view mental ability as unidi mensional , static, and as a distal predictor of performance (Ployhart, 2012) . By examining a specific dimension of general mental ability, the present work corroborates evidence that working memory can vary systematically within -person (Sliwinski et al. 20 06; Brose et al., 2012, 2014; Schmeidek et al. 2013). Further, physical activity and caffeine consumption help explain why WMC can fluctuate daily. These effect s provide evidence that within -person variance in WMC is systematic , whereas prior investigation s of WMC have primarily occurred at the between -person level (Ludyga et al. 2016; Porath & Erez, 2007; Rafeali et al. 2012 ). The present study examined the consequences of within -person variation in WMC using a self -regulatory framework given the potential role cognitive resources can play in self -regulated behavior (Beal et al. 2005; Hoffman et al. 2012; Vancouver et al. 2015). Further, I hypothesized that state WMC would work in conjunction with within -person change in ego depletion, g eneral mental ability, and time management strategy use to predict these outcomes. However, hypotheses relating state WMC to self -regulatory outcomes were largely unsupported. In all , evidence for state WMC and an explanation for its variation was revealed but the consequences of this variation could not be demonstrated . These findings challenge long -standing static conceptualizations of mental ability held in the organizational sciences (e.g. Ployhart, 2012), but fall short of demonstrating how these views inadequately capture the relationship between mental ability and performance. Theoretical Implications Working Memory. Broadly, a key question of the present research was whether WMC varied within -person in a meaningful way. The multilevel modelling app roach proposed by Schmeidek et al. (2013) was used as an initial step in answering this question. Findings indicated 75 that although most of the variance in scores over time was represented by between -person differences, a meaningful amount of variance exist ed within -person. This is in line with findings of several recent examinations of within -person variance in working memory (Sliwinski et al. 2006; Brose et al., 2012, 2014; Schmeidek et al. 2013). However, the proportion of variance at the within -person le vel of analysis was smaller than other constructs such as incivility, depletion , and goal progress . Despite these findings , the within -person variance observed in the present study is potentially important for several reasons. First, the construct of general mental ability is a cornerstone in the study of individual differences and in the prediction of p erformance (Schmidt & Hunter, 1998; Ployhart, 2012 ). Recent work has also shown that measures of WMC are as effective as general mental ability in predicting performance when the two are directly compared using a relative weights approach (Bosco, Allen, & Singh, 2015). Further, multiple examples exist dem onstrating that WMC can help explain why experiences like incivility can lead to lower performance (Porath & Erez, 2007; Rafeali et al., 2012). Despite these reasons, state WMC yielded few relationships with other variables in the present study . These res ults point to the possibility that state WMC is of little real -world consequence. Studies that use WMC as a mediator between the experience of incivility and performance rely on laboratory manipulations and experimental performance tasks (Porath & Erez, 20 07; Rafeali et al., 2012). It may be the case that these findings do not reflect the dynamics of these constructs as they naturally occur. Additionally, work like that of Maltese et al. (2016) demonstrates that sleep deprivation among physicians hurts perf ormance on a battery of cognitive tasks, but these researchers only argue that these declines may be related to medical errors. The possibility remains that experiences like sleep deprivation merely impact the measurement qualities of cognitive tasks rathe r than meaningful changes in cognition that can impact performance. 76 Further, past examinations of daily variability in WMC only consider WMC as an outcome. Though these investigations find that within -person variability in stress, as well as both positive and negative affect systematically vary with WMC, it may be the case that variance in WMC is of little broader consequence (Sliwinski et al. 2007; Brose et al. 2012, 2014). Finally, the present findings show that state WMC exhibited less within -person vari ance than most other constructs studied here. This limited within -person variability may constrain the relationship between state WMC and subsequent outcomes. If it is in fact the case that within -person variation in WMC does not impact behavior , then futu re research on WMC within the organizational sciences should continue using an individual differences approach (e.g. Bosco, Allen, & Singh, 2015). However, other possibilities remain that could explain the null findings for state WMC in the present study . One explanation could be that daily measurement in state WMC may not adequately capture the within -person variation in this construct . It could be the case that meaningful within -person variation in WMC occurs during time periods that are shorte r than a day. In Model 1, the autoregressive effects for state WMC ( = -.09, p < .0 1) and performance -based ego depletion ( = -.13, p < .0 1) were both negative. In this model, subjectively assessed ego depletion demonstrated a positive autoregression ( = .06, p < .0 5), whereas all other outcome autoregressions were nonsignificant . Schuurman et al. (2016) summarize that positive autoregressive effects can be thought of as inertia, such that the construct resists the change process. As a result, a positive autoregression indicates that deviations from baseline levels will tend to return to baseline over multiple units of time. The weak, positive autoregression for subjective ego depletion suggests that change may weakly persist from one day to the next. Schu urman et al. (2016) suggest that negative autoregression indicates a change process that counteracts deviation. Thus, the negative autoregression estimates of state WMC and 77 performance -based ego depletion show evidence of these constructs returning to base line following a day of deviation, a daily change pattern that is distinct from the other measures used here. As discussed previously, most examinations of within -person variance in WMC take a daily approach (e.g. Brose et al. 2012, 2014; Sliwinski et al. 2007). However, a recent study by Dirk and Schmiedek (2016) had elementary school students (ages 8 to 11) complete working memory tasks three times per day. Following the multilevel variance decomposition approach by Schmiedek et al. (2013), Dirk and Schm iedek (2016) found that daily, intraday, and error variance components were of roughly equal size. These researchers interpreted this finding as evidence for meaningful variance in WMC both within and across days . Further, students who scored lower on math and reading scholastic achievement tests also demonstrated greater WMC variability at both within and across days . Day -to-day and intraday variance estimates correlated modestly across students, which Dirk and Schmiedek (2016) interpr eted as reflecting unique aspects of working memory variability. Though findings from elementary school children may not directly generalize to university students or working adults, Dirk and Schmiedek™s (2016) work provides some evidence for meaningful fl uctuation in working memory at intraday timescales. Both the negative autoregressive effect for state WMC in the present study and past evidence of meaningful variation in WMC over shorter timescales may indicate that daily measurement of WMC is not freque nt enough to detect the effects proposed in the present study . The psychometric qualities of state WMC may also have shaped the findings of the present study . Research by Brose et al. (2012, 2014) has focused on measurement and has yielded two major findi ngs that help shape the interpretation of the present study and inform future work. First, Brose et al. find that reaction time measures of WMC tend to be more 78 systematic within -person than accuracy -based measures. The present research used an accuracy mea sure of WMC and found a high percentage (81%) of within -person variance was systematic. However, the estimate I observed may be misleading given that multiple studies have found meaningfully less systematic within -person variance in accuracy based WMC meas urement (Brose et al. 2012, 2014; Schmiedek et al., 2013). Brose et al. (2012, 2014) also examine individual differences in the extent to which within -person variance in WMC appears systematic . Both studies examined data from the COGITO study, which had pa rticipants perform cognitive tasks for 100 days. From this amount of data, Brose et al. (2012, 2014) were able to estimate multilevel models specific to each person, in which a random error variance term could be estimated and assigned to that person. In f ollow up substantive models, they found that this random error variance term demonstrated significant cross -level moderation effects, whereby within -person relationships between WMC and affect or motivation were stronger among those whose daily WMC task pe rformance was more systematic. Future work that uncovers what traits or under what conditions individuals consistently provide systematic WMC measurements may facilitate future tests of substantive relationships within -person variance in WMC holds with oth er constructs. Ego Depletion. The present study also has implication s for the measurement of ego depletion . Rosen et al. (2016) apply the Stroop task as a performance -based measure of ego depletion, arguing that such a task has several advantages over su bjective reports of ego depletion. The advantages they discuss primarily concern avoiding the problems with self -reports , such as mood or social -desirability (Kruger & Dunning, 1999; Johnson, Rosen, & Djurdjevic, 2011). Further, the hypotheses they pose we re informed by ego depletion theory and were largely supported. However, in the present work , subjectively assessed and performance -79 based measures of ego depletion were compared and results indicate d that one measure may not be an improved version over the other as Rosen et al. (2016) suggest. First, the two measures were uncorrelated within -person ( r = .01, ns), implying that these measures may not be interchangeable measures of within -person variance in ego depletion. Further, state achievement motivat ion and incivility were both found to predict subjectively assessed ego depletion as expected but not performance -based ego depletion. Thus , the se measures exhibit differential relationships with other measures despite purportedly measuring the same constr uct. The previously mentioned autoregressive results from Model 1 also suggest that the constructs assessed by these measures vary across time in different ways . The negative autoregressive effect for the Stroop task indicates a process that counteracts change whereas the positive autoregression for subjectively assessed ego depletion suggests that deviations persist from day to day. Additionally, the negative autoregressive effect for state WMC suggest s that the temporal characteristics of performance -based tasks were more similar to subjectively assessed ego depletion than the Stroop task. Further, if the Stroop task reflects change in a construct that usually returns to baseline by the following day, this change process would be captured well in the des ign used by Rosen et al. (2016) investigating change within a day. Given the similarities between the variation in the Stroop task and state WMC, i t may be that change in the Stroop task reflects the executive function of inhibition more than the motivatio nal construct of ego depletion (Miyake et al., 2000, Baumeister et al., 1998). This would align with the fact that the Stroop task was originally developed as a measure of cognitive processing (MacLeod et al., 1991, Stroop, 1935 ). However, the limited evid ence here is far from establishing construct validity, but perhaps serves as support for Hofmann et al.™s (2012) argument that research in self -regulation and executive function should be integrated. 80 Such integration may help clarify whether the Stroop t ask can be used as a performance measure of ego depletion. Rosen et al. (2016) use the Stroop task in their study as this task has been used as an indication of depletion in laboratory examinations of depletion. For example, Webb and Sheeran (2003) found t hat depletion manipulations significantly worsened Stroop task performance . These researchers also found that having participants state their intentions to perform well, a common manipulation to buffer against depletion effects, resulted in participants still performing the Stroop task well even if they were depleted (Webb & Sheeran, 2003). Similarly, Inzlicht, McKay, and Aronson (2006) found that highlighting one™s membership to a stigmatized demographic group, an experience that can cause depletion, can also worsen Stroop task performance. These studies served as early examples of the connection between ego depletion and other domains that require control, such as the control of attention required by the Stroop task (MacLeod, 1991; Stroop, 1935). However , finding that depletion can lead to changes in Stroop task performance does not necessarily mean that changes in Stroop task performance reflect depletion. As a cognitive processing task, performance on the Stroop task ma y reflect other constructs that ar e more related to cognition than self -regulation (MacLeod, 1991; Stroop, 1935). The examination of the change in Stroop task performance reflecting constructs that are not self -regulatory in nature would reveal how appropriate the Stroop task is as a perfo rmance measure of ego depletion. Work such as this has already been conducted examining self -report measures of ego depletion, highlighting the need to distinguish self -reported ego depletion from negative affect (Baumeister et al. 1998). Based on this di stinction, studies of ego depletion often incorporate measures of negative affect to ensure that observed effects are attributable to self -regulation rather than changes in affect . Future work could look to models of executive function 81 to identify construc ts that could be used as controls when drawing conclusions about performance measures of ego depletion (e.g. Miyake et al. 2000). Though meaningful differences were observed across measures of ego depletion, little evidence was provided to suggest that s tate WMC can inform ego depletion theory. This lack of evidence stems from the fact that state WMC was not found to be a significant predictor of any outcome, either directly or indirectly . State WMC was not found to partially mediate relationships previou sly investigated using ego depletion as a solitary mediator (e.g. Barnes et al. 2015, Rosen et al. 2016). Further, the results of the present study do not illuminate the interdependence of cognitive and self -regulatory resources at the within -person level of analysis . Interdependence between cognition and self -regulation was suggested by several perspectives, including the spotlight analogy of attention, resource models of episodic performance, and the integration of self -regulation and executive function ( Beal et al. 2005, Hofmann et al. 2012, Muraven et al. 2017). However, these perspectives do not conceptualize cognitive resources as varying within -person. Should such interdependence exist at the within -person level of analysis yet was obscured due to som e decision made in the present investigation, an effective strategy moving forward may be to focus on establishing the predictive validity of state WMC before examining interactive effects. Antecedents of Performance. In the present study, multiple psych ological and experiential antecedents of daily variation in performance were examined . The only significant relationship observed in the present study for state WMC was with physical activity . This relationship has potential implications for both construct s. For WMC, the fact that physical activity (and caffeine consumption ) predicted variance in state WMC suggests that variance in this construct is systematic. Although general mental ability is conceptualized as static and 82 unchanging in the organizational sciences (e.g. Ployhart, 2012), evidence was found that WMC may fluctuate over time . This opens the door for future research to examine the causes and consequences of within -person variance in state WMC. As the results suggest , a cau se that can be incorporated into future work is physical activity. Focusing on physical activity, this behavior is conceptualized primarily as a health behavior in the organizational sciences (e.g. Barber et al. 2017; Johnson & Allen, 2013 ). However, the p resent work provides evidence that physical activity can also promote cognitive functioning. This is in line with past work demonstrating that individual bouts of exercise can impact executive function (Ludyga et al., 2016) . However, the present study demo nstrate s that these effects can be observed at the within -person level of analysis among students engaging in everyday behavior. This generalization from the lab to the field suggest s that future work examining the impact of physical activity on executive function may have broad benefits. Further, the current results suggest that the organizational sciences should consider the potential beneficial effects that physical activity may have on task -related behavior. The results observed here also broadly suppo rt two propositions by Elliot and Schatke (2018) which are that achievement motivation can be effectively conceptualized as a state and that this state is meaningfully related to goal setting (Locke and Latham, 1990). Measures of state achievement motivati on exhibited a meaningful amount of within -person variance and predicted several other outcomes in the present work, providing evidence for the utility of examining achievement motivation as a state. After accounting for several other factors, s tate achiev ement motivation was a significant predictor of both short -term and long -term academic goal progress. This suggests that those who start the day motivated to achieve tend to set goals and take the necessary action to make progress towards these goals (i.e. reduce the discrepancy 83 between set goals and present states, Locke and Latham, 1990). Further, it appears as though individuals who are motivated may expend more effort in pursuit of goals given that state achievement motivation also predicted daily expen diture of effort. These results also point to the need to embed state achievement motivation within the study of daily variation in ego depletion (e.g. Rosen et al. 2016). State achievement motivation predicted greater self -regulatory resources at midday w hereas the experience of incivility predicted greater depletion. Future work may seek to investigate how motivation to achieve may operate in parallel or buffer against experiences people have throughout the day. To date, incivility has primarily been inv estigated between -persons at long timescales in the university context (Caza & Cortina, 2007) . These findings demonstrate that incivility can be observed on a daily basis among university students and that the consequences of this experience may be immedia te in the form of ego depletion. The study of incivility in professional settings has begun to focus on time scales that are more closely aligned to the experience itself (Cole & Shipp, 2016; Rosen et al. 2016). These findings suggest that the investigatio n of incivility in the university context should take a similar approach. Further, the link between incivility and ego depletion broadens the array of outcomes worth investigating among university students. At present, incivility experienced among students has primarily focused on outcomes related to their membership in the university as well as negative psychological experiences (e.g. detachment and distress, Caza & Cortina, 2007, Jensen et al. 2016). However, depletion has been related to negative health outcomes such as overeating and alcohol consumption , the experience of burnout, as well as general failure of self -control (Baumeister et al., 1998; Christiansen, Cole, & Field, 2012; Vohs & Heatherton, 200 0). Considering the consequences of depletion may help 84 direct future work demonstrating the consequences of experiencing incivility among university students. Surprisingly, sleep quantity and quality had a much narrower impact than expected in the present work. Measurement may be the primary issue why th is was the case. Despite past work arguing for the effectiveness of the approach taken here (e.g. Barnes et al. 2011; 2015), it may be the case that design limitations hurt response quality. These limitations are discussed further below. However, if measur ement is not the primary cause for the predominantly null effects of variation in sleep, it could be that sleep behavior and its outcomes vary systematically across university students and samples of working adults , which have been examined in previous research (e.g. Barnes et al., 2015). Though sleep quality directly predicted short -term goal progress , the consequences of poor sleep may be offset by choices students can make. For example, flexibility in scheduling and attendance may afford stud ents the opportunity to sleep longer when needed or work at a different time than originally planned. Such flexibility may allow students to adapt where as working adults are more constrained. Among students, f urther consideration of the role of caffeine co nsumption may also b e useful, as caffeine consumption may be used by university students to offset the performance -related costs of sleep deprivation. If this is the case, daily variance in goal progress may not be the most apparent consequence of sleep be havior . As was the case with the consequences of state WMC, reconsidering the choice of outcomes may help reveal the consequences of sleep deprivation. Practical Implications Many of the practical implications of this work relate to the antecedents of per formance that were investigated. A key practical implication is that p hysical activity may not be solely a buffer against negative health outcomes. Evidence was found that physical activity predicted 85 higher levels of state WMC. To the extent that future re search reveals the utility of state WMC, the current findings can broaden the purpose of organizational wellness programs related to physical activity . Typically, wellness programs are implemented to promote the health of organizational members and indirec tly improve organizational productivity through reduced absenteeism (Hutchinson & Wilson, 2011) . However, the findings here potentially uncover an important missing piece when evaluating the utility of wellness; the benefits gained from improved cognitive function . If it is the case that improved state WMC is beneficial, physical activity can be used to as a daily intervention to make sure levels of state WMC are high. Here, the benefits of physical activity in the morning were shown to have an impact by th e afternoon. Ludyga et al. (2016) note that most experimental studies demonstrate a positive effect on executive function 20 to 60 minutes after exercise. Thus, exercise right before the workday starts or incorporated into work activities may be beneficial . Further, Lugdyga et al. ™s (2016) meta -analysis focused specifically on moderate exercise, suggesting that the workday need not be broken up by strenuous exercise to achieve benefits. Future research clarifying the benefits of physical activity on executive function ing and exploring both t he timing and types of useful physical activity may help to improve the application of the present findings. The present work also suggests that the benefits of state achievement motivation at the start of the day continue throughout the day . However, a relatively small number of studies has been done on state achievement motivation. Locke and Shattke (2018) suggest that state achievement motivation should be r elated to goal setting (Locke and Latham, 1990). Thus, goal setting may be used to help bring about a state of achievement motivation. For example, c ourse goals that are set by either the student or instructor could be used as a motivational tool. Further, course goals can extend beyond deadlines for completed assignments, such as specifying 86 learning goals or skills to improve ( Masuda, Locke, & Williams, 2015 ). With goal -set ting , taking time to reflect on goals could further improve their utility. Morisano, Hirsh, Peterson, Pihl, and Shore (2010) found that course performance improved when university students reflected on life goals. Combining this research with the results of the current study, taking some time to reflect on goals at the start of the day ma y improve performance throughout the day through engendering a state of achievement motivation. Further research should examine this proposition directly to improve student performance by considering motivation on a daily basis. The findings presented he re also imply that the experience of incivility impacts students in a university context on a daily basis. Further, incivility was found to primarily predict ego depletion, which may have a host of negative consequences beyond what was investigated here (e.g. Baumeister et al., 1998 ). Rather than wait until the experience of incivility manifests into negative outcomes like psychological distress and disengagement from the university (e.g. Caza & Cortina, 2007, Jensen et al. 2016 ), the findings here imply th at intervention s may be more effective on a shorter time scale. Given that incivility is a phenomenon that occurs between multiple people, intervention may need to also be framed through different perspectives. For students who may be the target of incivil ity, affirmation that their experiences are valid and that consequences may include a temporary unwillingness to exert self -control may be helpful. Training in strategies to combat the negative consequences of incivility, such as effective copin g, may equip students to more effectively deal with being the target of incivility. Additionally, replenishment activities revealed by ego depletion research, such as sleep or experiences that bring about positive affect (Barnes et al., 2015; Tice, Baumeister, Sh mueli, & Muraven, 2007), may help targets of incivility minimize motivational consequences . 87 For perpetrators, many may not realize that their actions have immediate consequences for their targets. Given that incivility is low -intensity, rude behavior, per petrators may erroneously conclude that their actions have little effect on others ( Andersson & Pearson, 1999 ). Knowledge about the benefits of civility, both for the people around perpetrators as well as organizations, may help reduce the prevalence of in civility (Porath & Pearson, 2013). Additionally, knowledge about how the effects of incivility are thought to unfold and highlighting what a perpetr ator can do to rectify the situation may be helpful. Given that the negative consequences of incivility are thought to manifest as a result of rumination and reappraisal of an action with unclear intent (Ander sson & Pearson, 1999) , perpetrators may help bring clarity by apologizing in an authentic way. Clarity, along with an apology, may reduce the subsequent ne gative processing of the experience that the target may engage in and reduce the likelihood of further negative consequences. Though the negative consequences of incivility found here may appear minor, daily effects spread across a large student body sugge st a large real -world impact. Limitations and Future Directions Likely the greatest limitation of the current work was the design and assessment of the outcome measures. Problems with this part of the model could have inhibited the observation of within -person relationships for these outcomes and negatively affected the ability to observe cross -level interactions. For all outcomes, a within -person approach helped to protect against individual differences in responding influencing results (Bryk & Raudenbush , 2002). Further, a retrospective assessment of these outcomes helped create temporal spacing to mitigate common method variance associated with measurement timing (Johnson et al., 2011; Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). However, it is unclear how accurately participants were 88 able to report their standing on each outcome measure by the following morning. Several factors related to criterion measurement may have inhibited observation of hypothesized effects. Operationalization, measurement timin g, and inappropriateness of measures for a student sample may have all contributed to ineffective criterion measurement. Regarding forgetfulness, by definition , forgetful participants would not have been aware of the things they had forgotten . Alternativel y, even for event s that occurred, participants may not have had a clear memory of the event by the following day. Thus, operationalizing forgetfulness as students™ subjectively recalled experience of forgetfulness may not have accurately captured their act ual behavior. In addition to accuracy in memory influencing measurement, goal progress and effort outcomes are somewhat dependent on students having the need and desire to pursue goals or expend effort. The measurement of these outcomes followed previous r esearch by asking participants to reflect on how their goal progress or effort compared to their desired levels (Carver, 2004; Koopman et al. 2016, Rosen et al. 2018). It is possible that this approach suits working adults but not university students. Work ing adults may have more consistent goals (or at least amount of work) to complete each day, providing a consistent referent to compare their progress. Students, on the other hand, may vary from day to day both in terms of goal progress, but also in goals held each day. Thus, when participants did not endorse statements of goal progress, it could be because they did not actually make progress towards the goals that they set or because they did not have goals for that day. The same can be said for effort, where students were asked if they completed work even if they did not want to, questions that may assume that students feel the need to do work each day. Both goal progress and effort outcomes may have also been subject to inaccurate recall, whereby a students™ evaluation of what happened deviated from what truly occurred by the following morning. These issues in outcome measurement likely produced construct 89 contamination, po ssible resulting in outcomes that reflected students™ attitudes related to goal progress or effort rather than actual behavior. As a result, the hypothesized effects based on changes in behavior may have been harder to detect. To avoid these pitfalls, fu ture research should consider alternative measurement schedules and outcome measures. Outcome measure s administered at the end of the day rather than the following morning (e.g. Rosen et al. 2018) may help ensure the accuracy of responses. Assessment at th e end of the day may yield enough temporal spacing yet avoid overreliance on participant recall (Johnson et al., 2011; Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). Considering objective measures would also help to avoid potential problems with subjective measures. Although objective outcomes are common in many jobs (e.g., Rafeali et al., 2015), these outcomes may also become increasing ly available among students. As more of the student s™ educational experience is delivered through E -Learning methods, trac king interaction s with such tools may serve as a useful objective outcome ( Bell & Federman, 2013 ). Using both objective and subjective criterion measurement may also help address hypotheses while promoting accuracy in participant responding. In the present study, the distinction between short - and long -term academic goal progress may have been unclear for participants, resulting in similar responses across outcomes. A similar distinction could have been explored using different measurement approaches, such as objective measurement of task -related behavior as a short -term outcome and subjective measurement of future -oriented behavior, such as employee voice behaviors (Van Dyne & LePine, 1998). Future research should consider alternative designs bearing in min d the burden on participant s. Beyond outcom es, the assessment of time management strategy use presented another limitation related to construct measurement. Time management was assessed using a measure 90 developed for students ( Drzakowski et al. 2005 ). How ever, the present work found that this measure exhibited poor reliability. Measurement unreliability may have meaningfully obscured the hypothesized cross -level interaction between time management and the within -person relationship between state WMC and go al progress. Given that there did not appear to be a clear reason for the unreliability of this measure, additional research is needed to explore the measure ment of time management strategy use. Another limitation of the current work is that an excessive burden may have been placed on participants. The observed time that participants took on morning and afternoon surveys was near the expected times of five and ten minutes, respectively. Though the time of each survey was relatively short , and several modi fications were made to self -report scales to help mitigate the burden (Gabriel et al. 2018) , even these durations may have been too demanding for participants after 14 measurement occasions. Beyond mere survey length, having two performance measures may have compounded issues related to participant burden. Time spent completing performance measures may be perceived as more burdensome than responding to self -report items. If the design employed here was excessively burdensome, this may have negatively impacted the quality of WMC measurement as well as response effort in general (Meade & Craig, 2012). Therefore, the theoretical model tested here should also be examined across multiple studies usi ng different designs . The benefits of assessing the entirety of the model among the same participants may not outweigh the costs of participant burden. Finally, the use of a survey platform may have been a limitation and should be considered in future re search for accuracy in obtaining reaction time data. The method for obtaining reaction time data for the Stroop task required the survey platform to take the difference in time between stimulus presentation and participant response. Though further examinat ion of how these data 91 were produced is not possible, the survey platform provides output for the timing of stimuli displayed for a fixed amount of time. For example, the numerical updating task was designed to have stimuli presented on screen for two secon ds. After data are collected, the survey platform provides a record of the time stimuli were presented. Examination of the recorded durations for one stimulus revealed that the standard deviation of stimulus presentation times was 46.59 ms. Though this may seem insignificant , such variation in timing appears detrimental when compared to variation in average Stroop task performance ( SD = 235.64 ms). Additionally, the three -level modelling approach used here (i.e. Schmiedek et al. 2013) revealed that the accu racy -based numerical updating task exhibited more systematic daily variance than the Stroop task, whereas past work seems to suggest that reaction time measures (vs. accuracy measures) should capture daily variation more systematically , at least for workin g memory tasks (Brose et al. 2012; 2014). Future work should consider alternative survey platforms specifically designed for accuracy in collecting reaction time data. Conclusion The findings presented here challenge the predominant v iew of general menta l ability in the organizational sciences as static (Ployhart, 2012) . This view provides a fallible sense of security that mental ability is impervious to the dynamics of everyday life. By examining working memory, a specific dimension of cognitive ability, the current work corroborates evidence that aspects of cognition can vary systematically within -person (Sliwinski et al. 2006; Brose et al., 2012, 2014; Schmeidek et al. 2013). However, the results found here do not indicate that this variation had an impact on performance. The present study examined the consequences of within -person variation in WMC using a self -regulatory framework given the role cognitive resources have play in previous th eory (Beal et al. 2005; Hoffman et al. 2012; Vancouver et al. 92 2015). At this point it is difficult to say whether such consequences were not found due to a failure in theory or method. However, what was gained w as evidence that further reconsideration of mental ability with respect to time may be a worthwhile pursuit. 93 APPENDIX 94 APPENDIX Self -Report Measures and Associated Instructions Initial Survey Cognitive Ability Please report the highest overall ACT composite score you received prior to admission at MSU. (Numerical Response, or fiDid not take ACTfl) (If they did not take ACT) Please report the highest overall SAT composite score you received prior to admission at M SU. (Numerical Response) Time Management Strategies ( Drzakowski et al. 2005) Use the following scale to respond to the next set of items. Indicate how frequently each was true during this semester. 1=Never 2=On a few occasions 3=Sometimes 4=Often 5=Very often 1. How often were you late for class? 2. How often do you start working on major school projects early (e.g., a final paper)? 3. How often do you study intensely right before an exam (i.e., cram)? 4. How often have you turned in class assignmen ts late? 5. How often do you make or use schedules (daily, monthly, or for a school term)? 6. How often do you tend to get distracted when you study or work on a school paper (e.g., find other things to do, daydream)? 7. How often do you get too emotional or upset to go to class or get school work done? 95 Implicit Theories of Willpower (Job, Dweck, & Walton 2010) This questionnaire has been designed to investigate your ideas about willpower. Willpower is what you use to resist temptations, to stick to your intentions, and to remain in strenuous mental activity. There are no right or wrong answers. We are intereste d in your ideas. Using the scale below, please indicate how much you agree or disagree with each of the following statements. 1 Œ Strongly Disagree, 2 - Disagree, 3 Œ Neither agree nor disagree, 4- Agree, 5 - Strongly Agree Strenuous Mental Activity 1. Str enuous mental activity exhausts your resources, which you need to refuel afterwards (e.g. through taking breaks, doing nothing, watching television, eating snacks) (Reversed) 2. After a strenuous mental activity, your energy is depleted and you must rest t o get it refuelled again. (Reversed) 3. When you have been working on a strenuous mental task, you feel energized and you are able to immediately start with another demanding activity 4. Your mental stamina fuels itself. Even after strenuous mental exerti on, you can continue doing more of it. 5. When you have completed a strenuous mental activity, you cannot start another activity immediately with the same concentration because you have to recover your mental energy again. (Reversed) 6. After a strenuous m ental activity, you feel energized for further challenging activities. Resisting Temptations 96 1. Resisting temptations makes you feel more vulnerable to the next temptations that come along. (Reversed) 2. When situations accumulate that challenge you with t emptations, it gets more and more difficult to resist the temptations. (Reversed) 3. If you have just resisted a strong temptation, you feel strengthened and you can withstand any new temptations. 4. It is particularly difficult to resist a temptation afte r resisting another temptation right before. (Reversed) 5. Resisting temptations activates your willpower and you become even better able to face new upcoming temptations. 6. Your capacity to resist temptations is not limited. Even after you have resisted a strong temptation you can control yourself right afterwards. Trait Self -Control (Tangney et al. 2004) Using the scale provided, please indicate how much each of the following statements reflects how you typically are. 1 Œ Not at all like me 2 Œ Somewhat unlike me 3 Œ Neutral 4 Œ Somewhat like me 5 Œ Very much like me 1. I am good at resisting temptation 2. I have a hard time breaking bad habits (Reversed) 3. I am lazy (Reversed) 4. I say inappropriate things (Reversed) 5. I do certain things that are bad for me, if they are fun (Reversed) 97 6. I refuse things that are bad for me 7. I wish I had more self -discipline (Reversed) 8. People would say that I have iron self -discipline 9. Pleasure and fun sometimes keep me from getting work done (Reversed) 10. I have trouble concentrating (Reversed) 11. I am able to work effectively toward long -term goals 12. Sometimes I can™t stop myself from doing something, even if I know it is wrong 13. I often act without thinking through all the alternative s Morning Survey Are you completing this survey on a computer (e.g. desktop, laptop) or mobile device (e.g. phone, tablet)? - Computer - Mobile Device Pittsburgh Sleep Diary (Monk et al. 1994) Please answer the following questions about your sleep las t night: What time did you go to bed last night? [Time] How many minutes did it take you to fall asleep? [Number of Minutes] What time did you wake up? [Time] 98 After falling asleep, how many minutes were you awake? For example, if you were asleep until 1 a.m., woke at 1 a.m. and fell back asleep at 1:20 a.m. for the rest of the night, your answer would be 20 minutes. [Number of Minutes] Class Attendance Yesterday, how many hours of class (lectures, lab sections, meetings, etc.) did you have scheduled? (Numerical Response) Of the hours of class you had scheduled, how many did you attend? (Numerical Response) Work Attendance Yesterday, how many hours of work did you have scheduled? (Numerical Response, Option to indicate no employment) Of the hours of work you had scheduled, how many did you attend? (Numerical Response, Option to indicate no employment) Cognitive Failures Questionnaire, Forgetfulness Dimension (Broadbent et al. 1982) The following questions are about minor mistakes whic h everyone makes from time to time. Please report how often you made these mistakes yesterday using the following scale: 1 Œ Never 2 Œ Once 3 Œ A Few Times 4 Œ Quite Often 5 Œ Very Often 1. Did you find you couldn™t quite remember something although i t's "on the tip of your tongue"? 2. Did you forget where you put something like your phone or keys? 99 3. Did you read something and find you haven't been thinking about it and must read it again? Short - and Long -term Academic Goal Progress ( Koopman, Lanaj, & Scott , 2016) Most classes have activities or assignments that must be completed regularly throughout the course with short -term deadlines. Examples include completing homework assignments, studying for quizzes or mid -term exams, and completing course rea dings. Rate the extent to which you agree to the following statements about the progress you made toward short -term coursework using the following scale: 1 Œ Strongly disagree, 2 - Disagree, 3 Œ Neither agree nor disagree, 4 - Agree, 5 - Strongly Agree 1. I made good progress toward my short -term course goals 2. Things did not go well with my short -term course goals 3. I had a productive day yesterday in relation to my short -term coursework Most classes also have long -term activities or assignments to be completed by the end of the semester. Examples include studying for final exams, writing term papers, and completing end of semester projects or presentations. Rate the extent to which you agree to the following statements about the progress you made towa rd this end of semester coursework using the following scale: 1 Œ Strongly disagree, 2 - Disagree, 3 Œ Neither agree nor disagree, 4 - Agree, 5 - Strongly Agree 100 1. I made good progress toward my end of semester course goals 2. Things did not go well with my end of semester course goals 3. I had a productive day yesterday in relation to my end of semester coursework Effort (Pintrich & DeGroot, 1990) Consider the coursework you completed yesterday for your classes (e.g. completing assignments, studying, et c.). Please rate how well the following statements describe your work using the scale below: 1 2 3 4 5 6 7 Not at all Very true true of me of me 1. I worked hard to do well in my classes even if I didn™t like what we were doing 2. When class work was difficult, I gave up or only studied the easy parts 3. Even when course materials were dull and uninteresting, I managed to keep working until I finished 4. I often felt so lazy or bored when studying for my classes that I quit before I finished what I planned to do State Need for Achievement (Lang & Fries, 2006) 101 Use the following scale to rate how much you agree each statement reflects how you feel about today. 1 Œ Strongly Disagree, 2 - Disagree, 3 Œ Neither agree nor disagree , 4- Agree, 5 - Strongly Agree 1. Today, I am eager for situations that allow me to test my abilities 2. I look forward to tasks that let me test my abilities 3. I look forward to situations today in which I can find out how capable I am Afternoon Survey Are you completing this survey on a computer (e.g. desktop, laptop) or mobile device (e.g. phone, tablet)? - Computer - Mobile Device Subjective Ego Depletion ( Twenge, Muraven, & Tice , 2004) Please read the following statements and indicate to what extent they represent how you feel at this moment using the following scale: 1- Very slightly or not at all 2 Œ A little 3 Œ Moderately 4 Œ Quite a bit 5 - Extremely 1. I feel drained 2. Right now, it would take a lot of effort for me to concentrate on something 102 3. I feel mentally exhausted 4. I feel like my willpower is gone Negative Affect (Watson, Clark, & Tellegen, 1988) This scale consists of words that describe different feelings an d emotions. Read each word and then use the following scale to indicate the extent to which you have felt this way today. 1- Very slightly or not at all 2 Œ A little 3 Œ Moderately 4 Œ Quite a bit 5 - Extremely 1. Afraid 2. Upset 3. Nervous Experi enced Incivility (Lim & Cortina, 2005) Consider the interactions you have had with others since you woke up this morning. Please read the following statements and rate the extent to which each of these experiences had a negative effect on you using the f ollowing scale: 0 Œ Did not Occur 1 Œ No effect 2- Minor effect 3 Œ Moderate effect 4- Major Effect 1. I interacted with someone put me down or was condescending to me 2. Someone doubted my judgement on a matter over which I have responsibility 3. Someone paid little attention to my statements or showed little interest in my opinion 4. Someone made demeaning or derogatory remarks about me 103 Physical Activity (Godin & Shepard , 1985) Consider the physical activity you engaged in today. Report the number of minutes that you engaged in each activity. Strenuous Activity (heart beats rapidly) [Number of Minutes] Examples: running, jogging, weightlifting, vigorous swimming, vigorous cycling, basketball, soccer, football Moderate Exercise (not exhausting) [Number of Minutes] Examples: fast walking, easy bicycling, baseball, tennis, dancing Mild Exercise (minimal effort) [Number of Minutes] Examples: light yoga, bowling, golf, easy walking Caffeine Intake Prior to taking this survey, how many caffeinated drinks have you had today? Consider one caffeinated drink to be the equivalent of one 8 oz. cup of coffee or one small (8.4 oz) can of Red Bull Energy Drink. (Numerical Response) 104 REFERENCES 105 REFERENCES Ackerman, P. L., Beier, M. E., & Boyle, M. O. (2005). Working memory and intelligence: The same or different constructs?. Psychological bulletin, 131 (1), 30. Algina, J., & Swaminathan, H. (2011). Centering in two -level nested designs. Handbook of Advanced Multilevel Analysis , 285-312. Alt, D., & Itzkovich, Y. (2016). Adjustment to college and perceptions of faculty incivility. Current Psychology , 35(4), 657 -666. Andersson. L. M., & Pearson, C. M. 1999. Tit for tat? The spiraling effect of incivility in the workplace. Academy of Management Review , 24, 452-471. Asparouhov, T. (2018a , May 30). Fixed vs random coefficient? [Msg 2]. Message posted to http://www.statmodel.com/discussion/messages/24588/25686.html?1536605125 Asparouhov, T. (2018b, September 18). PROCESSORS/CHAINS in DSEM with Bayes [Msg 2]. Message posted to http://www.statmodel.com/discussion/messages/24588/25686.html?1536605125 Baddele y, A. (1996). Exploring the central executive. The Quarterly Journal of Experimental Psychology Section A , 49(1), 5 -28. Baddeley, A. (2003). Working memory: looking back and looking forward. Nature Reviews Neuroscience , 4(10), 829 -839. Baddeley, A. D., & H itch, G. (1974). Working memory. In Psychology of Learning and Motivation (Vol. 8, pp. 47 -89). Academic press. Banks, S. , & Dinges, D. F. (2007). Behavioral and physiological consequences of sleep restriction. Journal of Clinical Sleep Medicine , 3(5), 519 -528. Barber, L. K., Taylor, S. G., Burton, J. P., & Bailey, S. F. (2017). A self -regulatory perspective of work -to-home undermining spillover/crossover: Examining the roles of sleep and exercise. Journal of Applied Psychology, 102 (5), 753 -763. Barnes, C. M. (2012). Working in our sleep: Sleep and self -regulation in organizations. Organizational Psychology Review , 2(3), 234 -257. Barnes, C. M., Lucianetti, L., Bhave, D. P., & Christian, M. S. (2015). fiYou wouldn™t like me when I™m sleepyfl: Leaders ™ sleep, daily abusive supervision, and work unit engagement. Academy of Management Journal , 58(5), 1419 -1437. Barnes, C. M., Schaubroeck, J. M., Huth, M., & Ghumman, S. 2011. Lack of sleep and unethical behavior. Organizational Behavior and Human Decision Processes , 115, 169 Œ180. Baumeister, R. F., Bratslavsky, E., & Muraven, M. (2018). Ego depletion: Is the active self a limited resource?. Journal of Personality and Social Psychology, 74 (5), 1252 -1265. 106 Baumeister, R. F., & Heatherton, T. F. (1996). Self -regulation failure: An overview. Psychological inquiry, 7 (1), 1 -15. Baumeister, R. F., Muraven, M., & Tice, D. M. (2000). Ego depletion: A resource model of volition, self -regulation, and controlled processing. Social Cognition , 18(2), 130 -150. Beal, D. J. , Weiss, H. M., Barros, E., & MacDermid, S. M. (2005). An episodic process model of affective influences on performance. Journal of Applied Psychology , 90(6), 1054 -1068. Blumberg, E. J., Foroughi, C. K., Scheldrup, M. R., Peterson, M. S., Boehm -Davis, D. A ., & Parasuraman, R. (2015). Reducing the disruptive effects of interruptions with noninvasive brain stimulation. Human Factors , 57(6), 1051 -1062. Bosco, F., Allen, D. G., & Singh, K. (2015). Executive attention: an alternative perspective on general mental ability, performance, and subgroup differences. Personnel Psychology , 68(4), 859 -898. Broadbent, D. E., Cooper, P. F., FitzGerald, P., & Parkes, K. R. (1982). The cognitive failures questionnaire (CFQ) and its correlates. British Journal of Clinical Psychology, 21 (1), 1 -16. Brose, A., Lövdén, M., & Schmiedek, F. (2014). Daily fluctuations in positive affect positively co-vary with working memory pe rformance. Emotion , 14(1), 1 -6. Brose, A., Schmiedek, F., Lövdén, M., & Lindenberger, U. (2012). Daily variability in working memory is coupled with negative affect: the role of attention and motivation. Emotion , 12(3), 605 -617. Caldwell Jr, J. A., Caldwel l, J. L., Brown, D. L., & Smith, J. K. (2004). The effects of 37 hours of continuous wakefulness on the physiological arousal, cognitive performance, self -reported mood, and simulator flight performance of F -117A pilots. Military Psychology , 16(3), 163 -181. Caldwell, J. A., Caldwell, J. L., Smyth, N. K., & Hall, K. K. (2000). A double -blind, placebo -controlled investigation of the efficacy of modafinil for sustaining the alertness and performance of aviators: a helicopter simulator study. Psychopharmacology , 150(3), 272 -282. Carter, E. C., Kofler, L. M., Forster, D. E., & McCullough, M. E. (2015). A series of meta -analytic tests of the depletion effect: self -control does not seem to rely on a limited resource. Journal of Experimental Psychology: General , 144(4), 796 -816. Caza, B. B., & Cortina, L. M. (2007). From insult to injury: Explaining the impact of incivility. Basic and Applied Social Psychology , 29(4), 335 -350. Chai, W. J., Abd Hamid, A. I., & Abdullah, J. M. (2018). Working Memory from the Psychological and Neurosciences Perspectives: A Review. Frontiers in psychology , 9, 401-417. Christian, M. S., & Ellis, A. P. (2011). Examining the effects of sleep deprivation on workplace deviance: A self -regulatory perspective. Academy of Managemen t Journal , 54(5), 913 -934. 107 Clarke, P. (2008). When can group level clustering be ignored? Multilevel models versus single -level models with sparse data. Journal of Epidemiology and Community Health , 62, 752-758. College Board (2018). Guide to the 2018 ACT®/SAT® Concordance . The Coll ege Board, ACT Inc. Cole, M. S., Shipp, A. J., & Taylor, S. G. (2016). Viewing the interpersonal mistreatment literature through a temporal lens. Organizational Psychology Review , 6(3), 273 -302. Credé, M., & Phillips, L. A. (2011). A meta -analytic review of the Motivated Strategies for Learning Questionnaire. Learning and Individual Differences, 21 (4), 337 -346. Diamond, A. (2013). Executive functions. Annual Review of Psychology , 64, 135-168. Doran, S. M., Van Dongen, H. P. A., & Dinges, D. F. (2001). Sust ained attention performance during sleep deprivation: evidence of state instability. Archives Italiennes de Biologie , 139(3), 253 -267. Drzakowski, S. M., Friede, A. J., Imus, A. L., Kim, B. H., & Shivpuri, S. (2005). Non -cognitive criteria for college stud ent success: Satisfaction and persistence . Report submitted to College Board East Lansing, MI: Department of Psychology, Michigan State University. Edwards, J. R., & Lambert, L. S. (2007). Methods for integrating moderation and mediation: a general analyti cal framework using moderated path analysis. Psychological Methods , 12(1), 1 -22. Engle, R. W. (2002). Working memory capacity as executive attention. Current Directions in Psychological Science , 11(1), 19 -23. Engle, R. W., Tuholski , S. W., Laughlin, J. E., & Conway, A. R. (1999). Working memory, short -term memory, and general fluid intelligence: a latent -variable approach. Journal of Experimental Psychology: General , 128(3), 309 -331. Fang, H. (2006). A Monte Carlo study of power ana lysis of hierarchical linear model and repeated measures approaches to longitudinal data analysis (Doctoral dissertation). Retrieved from ProQuest Dissertations and Theses. (UMI No. 3230529) Farrell, A. H., Provenzano, D. A., Spadafora, N., Marini, Z. A., & Volk, A. A. (2016). Measuring adolescent attitudes toward classroom incivility: Exploring differences between intentional and unintentional incivility. Journal of Psychoeducational Assessment , 34(6), 577 -588. Ferrara, M., & De Gennaro, L. (2001). How muc h sleep do we need?. Sleep Medicine Reviews , 5(2), 155 -179. Gabriel, A. S., Podsakoff, N. P., Beal, D. J., Scott, B. A., Sonnentag, S., Trougakos, J. P., & Butts, M. M. (2018). Experience sampling methods: a discussion of critical trends and considerations for scholarly advancement. Organizational Research Methods , DOI: 1094428118802626. 108 Gailliot, M. T., Schmeichel, B. J., & Baumeister, R. F. (2006). Self -regulatory processes defend against the threat of death: Effects of self -control depletion and trait se lf-control on thoughts and fears of dying. Journal of Personality and Social Psychology , 91(1), 49. Godin, G. (2011). The Godin -Shephard leisure -time physical activity questionnaire. The Health & Fitness Journal of Canada , 4(1), 18 -22. Godin, G., & Shephard, R. J. (1985). A simple method to assess exercise behavior in the community. Canadian Journal of A pplied Sport Science , 10(3), 141 -146. Goel, N., Rao, H., Durmer, J. S., & Dinges, D. F. (2009). Neurocognitive consequences of sleep deprivation. In Seminars in Neurology (Vol. 29, No. 4, p. 320). NIH Public Access. Hafner, M., Stepanek, M., Taylor, J., Tr oxel, W. M., & Van Stolk, C. (2016). Why sleep matters Œthe economic costs of insufficient sleep . Europe: RAND Corporation. Hagger, M. S., Wood, C., Stiff, C., & Chatzisarantis, N. L. (2010). Ego depletion and the strength model of self -control: a meta -anal ysis. Psychological Bulletin , 136(4), 495 -525. Hales, B. M., & Pronovost, P. J. (2006). The checklist Ša tool for error management and performance improvement. Journal of Critical Care, 21 (3), 231 -235. Hendy, K. C., Liao, J., & Milgram, P. (1997). Combining time and intensity effects in assessing operator information -processing load. Human Factors, 39 (1), 30 -47. Hofmann, W., Gschwendner, T., Friese, M., Wiers, R. W., & Schmitt, M. (2008). Working memory capacity and self -regulatory behavior: toward an indivi dual differences perspective on behavior determination by automatic versus controlled processes. Journal of Personality and Social Psychology , 95(4), 962 -977. Hofmann, W., Schmeichel, B. J., & Baddeley, A. D. (2012). Executive functions and self -regulation. Trends in Cognitive Sciences , 16(3), 174 -180. Inzlicht, M., McKay, L., & Aronson, J. (2006). Stigma as ego depletion: How being the target of prejudice affec ts self -control. Psychological Science, 17 (3), 262 -269. Iqbal, S. T., & Horvitz, E. (2007, April). Disruption and recovery of computing tasks: Field study, analysis, and directions . In Proceedings of the SIGCHI conference on human factors in computing syst ems (pp. 677 Œ686). New York, NY: Association for Computing Machinery. Jensen, J. M., Ahmad, A. S., King, E. B., & Lee, J. (2016). A Cross -Cultural Investigation of the Effects of Incivility on Occupational Aspirations. Journal of College Student Developmen t, 57(3), 233 -247. Jewsbury , P. A., Bowden, S. C., & Strauss, M. E. (2016). Integrating the switching, inhibition, and updating model of executive function with the Cattell ŠHorn ŠCarroll model. Journal of Experimental Psychology: General, 145 (2), 220 -245. Job, V., Dweck, C. S., & Wal ton, G. M. (2010). Ego depletion ŠIs it all in your head? Implicit theories about willpower affect self -regulation. Psychological Science , 21(11), 1686 -1693. 109 Job, V., Walton, G. M., Bernecker, K., & Dweck, C. S. (2013). Beliefs about willpower determine the impact of glucose on self -control. Proceedings of the National Academy of Sciences, 110 (37), 14837 -14842. Johnson, R. C., & Allen, T. D. (2013). Examining the links between employed mothers' work characteristics, physical activity, and child health. Journal of Applied Psychology, 98 (1), 148-157. Johnson, R. E., Chang, C. H., & Lord, R. G. (2006). Moving from c ognition to behavior: What the research says. Psychological Bulletin , 132(3), 381 -415. Johnson, R. E., Rosen, C. C., & Djurdjevic, E. (2011). Assessing the impact of common method variance on higher order multidimensional constructs. Journal of Applied Psy chology, 96(4), 744. Johnson, R. E., Lin, S. H., & Lee, H. W. (2018). Self -Control as the Fuel for Effective Self -Regulation at Work: Antecedents, Consequences, and Boundary Conditions of Employee Self -Control. In Advances in Motivation Science (Vol. 5, pp . 87-128). Elsevier. Jones, F., O' Connor, D. B., Conner, M., McMillan, B., & Ferguson, E. (2007). Impact of daily mood, work hours, and iso -strain variables on self -reported health behaviors. Journal of Applied Psychology, 92 (6), 1731 -1740. Kane, M. J., & Engle, R. W. (2003). Working -memory capacity and the control of attention: the contributions of goal neglect, response competition, and task set to Stroop interference. Journal of Experimental Psychology: General , 132 (1), 47 -70. Kane, M. J., Bleckley, M. K ., Conway, A. R., & Engle, R. W. (2001). A controlled -attention view of working -memory capacity . Journal of Experimental Psychology: General, 130(2), 169 -183. Kanfer, R., & Ackerman, P. L. (1989). Motivation and cognitive abilities: An integrative/aptitude -treatment interaction approach to skill acquisition. Journal of Applied Psychology , 74(4), 657 -690 . Kanfer , R. (1990). Motivation theory and industrial and organizational psychology. Handbook of Industrial and Organizational Psychology, 1 (2), 75 -130. Klein, K. J., & Kozlowski, S. W. J. (Eds.). (2000). Multilevel theory, research, and methods in organizations: Foundations, extensions, and new directions . San Francisco, CA, US: Jossey -Bass. Koopman, J., Lanaj, K., & Scott, B. A. (2016). Integrating the bright and dark sides of OCB: A daily investigation of the benefits and costs of helping others. Academy of Mana gement Journal , 59(2), 414 -435. Kruger, J., & Dunning, D. (1999). Unskilled and unaware of it: how difficulties in recognizing one's own incompetence lead to inflated self -assessments. Journal of Personality and Social Psychology, 77 (6), 1121. 110 Kruschke, J. K., Aguinis, H., & Joo, H. (2012). The time has come: Bayesian methods for data analysis in the organizational sciences. Organizational Research Methods, 15 (4), 722 -752. Lanaj, K., Johnson, R. E., & Barnes, C. M. (2014). Beginning the workday yet already depleted? Consequences of late -night smartphone use and sleep. Organizational Behavior and Human Decision Processes , 124(1), 11 -23. Lang, J. W., & Fries, S. (2006). A revised 10 -item version of the Achievement Motives Scale. European Journal of Psychologic al Assessment , 22(3), 216 -224. Lim, J., & Dinges, D. F. (2008). Sleep deprivation and vigilant attention. Annals of the New York Academy of Sciences , 1129 (1), 305 -322. Lim, S., & Cortina, L. M. (2005). Interpersonal mistreatment in the workplace: the interface and impact of general incivility and sexual harassment. Journal of Applied Psychology , 90(3), 483-496. Locke, E. A., & Latham, G. P. (1990). A theory of goal setti ng & task performance . Prentice -Hall, Inc. Locke, E. A., & Schattke, K. (2018). Intrinsic and extrinsic motivation: Time for expansion and clarification. Motivation Science. Advance online publication. Lord, R. G., Diefendorff, J. M., Schmidt, A. M., & Hal l, R. J. (2010). Self -regulation at work. Annual Review of Psychology , 61, 543-568. moderate aerobic exercise on specific aspects of executive function in diffe rent age and Psychophysiology, 53 (11), 1611 -1626. Ma, J., Lin, S. -H., Johnson, R. E., & Chang, C. -H. (2016). Commute stressor and effective leader behaviors: An ego depletion perspective. Paper presented at the 31st annual society for indus - trial and organizational psychology conference, Anaheim, California. MacLeod, C. M. (1991). Half a century of research on the Stroop effect: an integrative review. Psychological Bulletin , 109(2), 163 -203. Maltese, F., Adda, M., Bablon, A., Hraeich, S., Guervilly, C., Lehingue, S., ... & Thirion, X. (2016). Night shift decreases cognitive performance of ICU physicians . Intensive Care Medicine, 42(3), 393 -400. Maranges, H. M., Schmeichel, B. J., & Baumeister, R. F. (2017). Comparing cognitive load and self -regulatory depletion: Effects on emotions and cognitions . Learning and Instruction, 51, 74-84. McClelland, D. C., Atkinson, J. W., & Clark, R. A. (4). R., & Lowell, E.(1953). The achievement motive. McGrew, K. S. (2009). C HC theory and the human cognitive abilities project: Standing on the shoulders of the giants of psychometric intelligence research. Intelligence. 37 (1), 1 -10. 111 McVay, J. C., & Kane, M. J. (2012). Why does working memory capacity predict variation in reading comprehension? On the influence of mind wandering and executive attention. Journal of Experimental Psychology: General, 141 (2), 302 -320. Miller, H. C., DeWall, C. N., Pattison, K., Molet, M., & Zentall, T. R. (2012). Too dog tired to avoid danger: Self -control depletion in canines increases behavioral approach toward an aggressive threat. Psychonomic Bulletin & Review, 19 (3), 535 -540. Miyake, A., & Shah, P. (Eds.). (1999). Models of working memory: Mechanisms of active maintenance and executive control . Ca mbridge University Press. Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., Howerter, A., & Wager, T. D. (2000). The unity and diversity of executive functions and their contributions to complex fifrontal lobefl tasks: A latent variable analysis. Cognitive Psychology, 41 (1), 49 -100. Monk, T. H., Reynolds III, C. F., Kupfer, D. J., Buysse , D. J., Coble, P. A., Hayes, A. J., ... & Ritenour, A. M. (1994). The Pittsburgh sleep diary. Journal of Sleep Research, 3 (2), 111 -120. Muraven, M., & Baumeister, R. F. (2000). Self -regulation and depletion of limited resources: Does self -control resemble a muscle?. Psychological Bulletin, 126 (2), 247 -259. 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 (3), 774 -789. Muthén, B., & Asparo uhov, T. (2012). Bayesian structural equation modeling: a more flexible representation of substantive theory. Psychological Methods, 17 (3), 313. Muthén, L. K., & Muthén, B. O. (201 7). Mplus statistical modeling software: Release 8.0. Los Angeles, CA: Muthé n & Muthén. Neal, A., Ballard, T., & Vancouver, J. B. (2017). Dynamic self -regulation and multiple -goal pursuit. Annual Review of Organizational Psychology and Organizational Behavior, 4 , 401-423. Neal, A., Hannah, S., Sanderson, P., Bolland, S., Mooij, M. , & Murphy, S. (2014). Development and validation of a multilevel model for predicting workload under routine and nonroutine conditions in an air traffic management center. Human Factors, 56 (2), 287 -305. Nye, C. D., & Drasgow, F. (2011). Effect size indice s for analyses of measurement equivalence: Understanding the practical importance of differences between groups. Journal of Applied Psychology, 96 (5), 966 -980. exercise. British Journal of Health Psychology, 11 (4), 717 -733. Oswald, F. L., Hambrick, D. Z., & Jones, L. A. (2007). Keeping all the plates spinning: Understanding and predicting multitasking performance. In D. H. Jonassen (Ed.), Learning to solve complex scientific problems (pp. 77 Œ97). Mahwah, NJ: Erlbaum. 112 Pintrich, P. R., & DeGroot, E. (1990, April). Quantitative and qualitative perspectives on student motivation al beliefs and self -regulated learning . In Annual Meeting of the American Educational Research Association, Boston, MA (Vol. 128). Ployhart, R. E. (2012). Personnel selection: Ensuring sustainable organizational effectiveness through the acquisition of hum an capital. The Oxford handbook of organizational psychology, 1, 221-246. Porath, C. L., & Erez, A. (2007). Does rudeness really matter? The effects of rudeness on task performance and helpfulness. Academy of Management Journal, 50 (5), 1181 -1197. Porath, C ., & Pearson, C. (2013). The price of incivility. Harvard Business Review, 91 (1-2), 114 -121. Porath, C., MacInnis, D., & Folkes, V. (2010). Witnessing incivility among employees: Effects on consumer anger and negative inferences about companies. Journal of Consumer Research, 37 (2), 292 -303. Preacher, K. J., Curran, P. J., & Bauer, D. J. (2006). Computational tools for probing interactions in multiple linear regression, multilevel modeling, and latent curve analysis. Journal of Educational and Behavioral Statistics, 31 (4), 437 -448. Preacher, K. J., Zyphur, M. J., & Zhang, Z. (2010). A general multilevel SEM framework for assessing multilevel mediation. Psychological Methods, 15 (3), 209 -233. Rafaeli, A., Erez, A., Ravid, S., Derfler -Rozin, R., Treister, D. E., & Scheyer, R. (2012). When customers exhibit verbal aggression, employees pay cognitive costs. Journal of Applied Psychology, 97 (5), 931 -950 . Randall, J. G., Oswald, F. L., & Beier , M. E. (2014). Mind -wandering, cognition, and performance: A theory -driven meta -analysis of attention regulation. Psychological Bulletin, 140 (6), 1411 -1432 . Rast, P., Zimprich, D., Van Boxtel, M., & Jolles, J. (2009). Factor structure and measurement inva riance of the cognitive failures questionnaire across the adult life span. Assessment , 16(2), 145 -158. Redick, T. S., Calvo, A., Gay, C. E., & Engle, R. W. (2011). Working memory capacity and go/no -go task performance: Selective effects of updating, maintenance, and inhibition. Journal of Experimental Psychology: Learning, Memory, and Cognition, 37 (2), 308 -324. Redick, T. S., Shipstead, Z., Meier, M. E., Montroy, J. J., Hicks, K. L., Unsworth, N., ... & Engle, R. W. (2016). Cognitive predictors of a common multitasking ability: Contributions from working memory, attention control, and fluid intelligence . Jou rnal of Experimental Psychology: General, 145 (11), 1473 -1492. Robbins, S. B., Lauver, K., Le, H., Davis, D., Langley, R., & Carlstrom, A. (2004). Do psychosocial and study skill factors predict college outcomes? A meta -analysis. Psychological bulletin, 130 (2), 261. 113 Rosen, C. C., Koopman, J., Gabriel, A. S., & Johnson, R. E. (2016). Who strikes back? A daily investigation of when and why incivility begets incivility. Journal of Applied Psychology, 101(11), 1620 -1634. Rosen, C. C., Simon, L. S., Gajendran, R. S., Johnson, R. E., Lee, H. W., & Lin, S. -H. (J.). (2019). Boxed in by your inbox: Implications of daily e -mail demands for managers™ leadership behaviors. Journal of Applied Psychology, 104 (1), 19 -33. Sallis, J. F., & Saelens, B. E. (2000). Assessment of physical activity by self -report: status, limitations, and future directions. Research Quarterly for Exercise and Sport , 71(2), 1 -14. Saper, C. B., Scammell, T. E., & Lu, J. (2005). Hypothalamic regulation of sleep and circadian rhythms. Nature, 437 (7063), 1257 -1263. Scherbaum, C. A., & Ferreter, J. M. (2009). Estimating statistical power and required sample sizes for organizational research using multilevel modeling. Organizational Research Methods, 12 (2), 347 -367. Schermelleh -Engel, K., Moosbrugger , H., & Müller, H. (2003). Evaluating the fit of structural equation models: Tests of significance and descriptive goodness -of-fit measures. Methods of Psychological Research Online , 8(2), 23 -74. Schmidt, F. L., & Hunter, J. E. (1998). The validity and uti lity of selection methods in personnel psychology: Practical and theoretical implications of 85 years of research findings. Psychological Bulletin , 124(2), 262. Schmiedek, F., Lövdén, M., & Lindenberger, U. (2013). Keeping it steady: older adults perform more consistently on cognitive tasks than younger adults. Psychological Science , 24(9), 1747-1754. Schmeichel, B. J., Volokhov, R. N., & Demaree, H. A. (2008). Working memory capacity and the self -regulation of emotional expression and experience. Journal o f Personality and Social Psychology, 95 (6), 1526 -1540. Schuurman, N. K., Ferrer, E., de Boer -Sonnenschein, M., & Hamaker, E. L. (2016). How to compare cross -lagged associations in a multilevel autoregressive model. Psychological Methods, 21(2), 206 -221. Selig, J. P., & Preacher, K. J. (2008). Monte Carlo method for assessing mediation: An interactive tool for creating confidence intervals for indirect effects [Computer software]. Singh, R. K., & Göritz, A. S. (2018). Ego Depletion Does Not Interfere With Wo rking Memory Performance. Frontiers in Psychology, 9 , 538-548. Sliwinski, M. J., Smyth, J. M., Hofer, S. M., & Stawski, R. S. (2006). Intraindividual coupling of daily stress and cognition. Psychology and Aging, 21 (3), 545 -557. Smallwood, J., & Schooler, J. W. (2006). The restless mind. Psychological bulletin, 132 (6), 946 -958. Snijders, T. A., & Bosker, R. J. (1994). Modeled variance in two -level models. Sociological Methods & Research, 22 (3), 342 -363. 114 Steel, P., & König, C. J. ( 2006). Integrating theories of motivation. Academy of Management Review, 31 (4), 889 -913. Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental Psychology, 18 (6), 643 -662 . Spearman, C. (1904). " General Intelligen ce," objectively determined and measured. The American Journal of Psychology, 15 (2), 201 -292. Sweller, J., Van Merrienboer, J. J., & Paas, F. G. (1998). Cognitive architecture and instructional design. Educational Psychology Review, 10 (3), 251 -296. Tangney adjustment, less pathology, better grades, and interpersonal success. Journal of Personality, 72 (2), 271 -324. Toker, S., & Biron, M. (2012). Job burnout and depression: unra veling their temporal relationship and considering the role of physical activity. Journal of Applied Psychology, 97(3), 699 -710. Twenge, J., Muraven, M., & Tice, D. (2004). Measuring state self -control: Reliability, validity, and correlations with physical and psychological stress, Unpublished manuscript. San Diego State University. Van Dongen, H., Maislin, G., Mullington, J. M., & Dinges, D. F. (2003). The cumulative cost of additional wakefulness: dose -response effects on neurobehavioral functions and sleep physiology from chronic sleep restriction and total sleep deprivation. Sleep, 26 (2), 117 - 126. Van Dyne, L., & LePine, J. A. (1998). Helping and voice extra -role behaviors: Evidence of construct and predictive validity. Academy of Management jour nal, 41 (1), 108 -119. Vancouver, J. B., & Kendall, L. N. (2006). When self -efficacy negatively relates to motivation and performance in a learning context. Journal of Applied Psychology, 91 (5), 1146 -1153. Vrugt, A., & Oort, F. J. (2008). Metacognition, achievement goals, study strategies and academic achievement: pathways to achievement. Metacognition and Learning, 3 (2), 123 -146. Wagner, D. T., Barnes, C. M., Lim, V., & Ferris, D. L. (2012). Lost sleep and cyberloafing: Evidence from the laboratory and a Daylight -Saving Time qua si-experiment. Journal of Applied Psychology, 97 (5) , 1068Œ1076. Wanberg, C. R., Kanfer, R., & Banas, J. T. (2000). Predictors and outcomes of networking intensity among unemployed job seekers. Journal of Applied Psychology, 85(4) , 491-503. Watkins, E. R. ( 2008). Constructive and unconstructive repetitive thought. Psychological Bulletin, 134 (2), 163 -206. 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, 54 (6), 1063 -1070. 115 Webb, T. L., & Sheeran, P. (2003). Can implementation intentions help to overcome e go-depletion?. Journal of Experimental Social Psychology, 39 (3), 279 -286. Wilhelm, O., Hildebrandt, A. H., & Oberauer, K. (2013). What is working memory capacity, and how can we measure it?. Frontiers in Psychology, 4 , 433-455. Yuan, Y., & MacKinnon, D. P. (2009). Bayesian mediation analysis. Psychological Methods, 14(4), 301 -322. Zyphur, M. J., & Oswald, F. L. (2013). Bayesian probability and statistics in management research: A new horizon. Journal of Management, 39 (1), 5 -13.