Abstract
Reattachment to work (i.e., rebuilding a mental connection to work) before actually starting work is important for work engagement during the day. Building on motivated action theory, this study examines anticipated task focus, positive affect, and job resources (job control and social support) as mediators that translate reattachment in the morning into work engagement during the day. We collected daily-survey data from 151 employees (total of 620 days) and analyzed these data with a multilevel path model. We found that day-level reattachment to work in the morning predicted anticipated task focus, positive affect, social support, and job control through goal activation and that anticipated task focus, positive affect, social support, and job control predicted work engagement during the day. This study points to the important role of reattachment to work in employee experiences and behaviors throughout the workday and specifically highlights the benefits of such initial mental boundary crossing between life domains for employee engagement at work.
In today’s work organizations, employee work engagement is crucial (Schneider, Yost, Kropp, Kind, & Lam, 2018). This means that employees need to be energetic, dedicated, and absorbed into their daily work (Bakker, 2017). Work engagement—defined as “a positive motivational state directed toward deriving meaningfulness at work” (Byrne, Peters, & Weston, 2016: 1222)—is positively associated with job performance (Christian, Garza, & Slaughter, 2011) and with indicators of employee health and well-being (Crawford, LePine, & Rich, 2010; Leijten, van den Heuvel, van der Beek, Ybema, Robroek, & Burdorf, 2015). During the past two decades, research has identified a plethora of job characteristics and individual differences (for overviews, see Bakker, Demerouti, & Sanz-Vergel, 2014; Byrne et al., 2016) as well as day-specific on-the-job events and experiences (Xanthopoulou, Bakker, Demerouti, & Schaufeli, 2009) as predictors of work engagement. Because most of these factors are rooted in the work environment or employees’ personality, they are often beyond employees’ control. Only recently, research began to address the question concerning what employees themselves can do in order to increase work engagement during the day (Demerouti, Bakker, & Halbesleben, 2015; Venz, Pundt, & Sonnentag, 2018).
In line with the idea that mornings are particularly important for setting the tone for the day (Rothbard & Wilk, 2011), Sonnentag and Kühnel (2016) have argued that when employees actively reattach to work in the morning, they are more engaged during the day at work. Reattachment refers to rebuilding a mental connection to work, for instance, by thinking about specific tasks or by mentally simulating the workday before actually starting work. Reattachment typically occurs while transitioning from the nonwork to the work role (Hall & Richter, 1988) and is largely under employees’ control. In a daily-survey study, Sonnentag and Kühnel provided the first empirical evidence that mentally reattaching to work in the morning predicted work engagement during the day. This finding not only adds to the understanding of work engagement and role transitions between life domains but also provides suggestions for easy-to-implement strategies that can help boost work engagement on a daily basis.
Despite this first empirical evidence that reattachment in the morning predicts daily work engagement, the mechanisms that link reattachment to work engagement are still poorly understood. Virtually nothing is known about the processes that are stimulated when reattaching oneself to work and that, in turn, help to be engaged at work. To arrive at a better understanding of why reattachment matters for work engagement, it is crucial to address the processes underlying the association between reattachment and work engagement. Moreover, also with respect to practical implications, gaining more insights into the mechanisms that link reattachment to work engagement will point to the specific ways in which managers and organizations may facilitate employee engagement.
In order to advance theory building on reattachment and to better explain how the transition from nonwork life into the workday unfolds, we examine mediators of the relationship between reattachment and work engagement. Building on the emergent literature on reattachment to work that has suggested redirection of attention to work, energy mobilization, and allocation of job resources as the core purposes of reattachment (Sonnentag & Kühnel, 2016), we examine anticipated task focus, activated positive affect, and job resources as mediators. These specific mediators are not only important in linking reattachment to work engagement but also crucial on-the-job experiences in themselves. For example, anticipated task focus and positive affect are relevant for job performance (Beal, Weiss, Barros, & MacDermid, 2005; Bindl & Parker, 2012), and social support as well as job control help in mitigating the effects of job stressors on strain indicators (Karasek & Theorell, 1990). By addressing anticipated task focus, activated positive affect, and job resources as more immediate consequences of reattachment, our study highlights the importance of reattachment—even beyond its relationship with work engagement.
Research indicates that reattachment as well as work engagement fluctuate substantially from day to day (Sonnentag & Kühnel, 2016; Xanthopoulou, Bakker, & Ilies, 2012). Thus, the degree to which an employee reattaches to work in the morning of a specific day may translate into work engagement experienced during that day. To address these day-to-day dynamics of reattachment and work engagement, we develop our hypotheses at the day level and test these hypotheses via a daily-survey study.
Our study contributes to the literature in several ways. First, it adds to the emerging literature on reattachment to work by specifying and empirically testing the pathways that link reattachment to work in the morning to work engagement experienced during the day. By examining multiple mediators, our study provides a deeper understanding of the potentially complex processes triggered by reattachment. Accordingly, it helps clarify how thoughts about the upcoming workday translate into a performance-enhancing state of mind experienced throughout the workday. Second, our study adds to research on work engagement, particularly to research on day-level fluctuations of work engagement that has described day-level job resources as predictors of day-level fluctuations in work engagement (Kühnel, Sonnentag, & Bledow, 2012; Xanthopoulou, Bakker, Heuven, Demerouti, & Schaufeli, 2008). Our study extends the still sparse research on antecedents of day-level fluctuations in job resources (Breevaart, Bakker, Hetland, Demerouti, Olsen, & Espevik, 2014) by examining reattachment and goal activation as antecedents of day-specific job resources. Finally, our study adds to the literature on boundary transitions between various life domains (Allen, Cho, & Meier, 2014) by examining reattachment as a boundary-crossing activity (Ashforth, Kreiner, & Fugate, 2000). Specifically, we propose that the way employees enter the work domain in the morning sets the tone for the entire workday and relates to their immediate affective and attentional states as well as work engagement later during the day. Importantly, we focus on reattachment to work as an activity that employees can use as self-regulatory behavior to stay engaged at work during the day (Breevaart, Bakker, & Demerouti, 2014).
Reattachment to Work
Sonnentag and Kühnel (2016: 380) described reattachment to work as rebuilding a mental connection to one’s work after a nonwork period (e.g., a free evening or a weekend). Reattachment means thinking about the upcoming workday and mentally preparing for it. It is a form of prospection (Gilbert & Wilson, 2007) and occurs when psychologically crossing the border between the nonwork and the work domain (Clark, 2000), that is, transitioning from the nonwork to the work role (Ashforth et al., 2000; Hall & Richter, 1988). On the basis of in-depth interviews, Hall and Richter suggested that employees go through such daily transitions from home to work, and vice versa, and they found that “people use an anticipatory transition style in the move from home to work” (1988: 215), starting to think about work long before leaving home. In essence, during reattachment, one tunes into one’s work role and brings work into one’s attentional focus again. Planning activities (Gollwitzer, Fujita, & Oettingen, 2004) or mental simulation (Szpunar, 2010) are possible but not necessary activities during reattachment. In contrast to other daily experiences that also reflect a strong connection to one’s work (e.g., attention and absorption, Rothbard, 2001; intrinsic motivation, Deci, Olafsen, & Ryan, 2017), reattachment refers to an experience before actually starting work—as opposed to experiences occurring while performing one’s work.
We build on motivated action theory (DeShon & Gillespie, 2005) to explain the link between reattachment to work and work engagement. According to motivated action theory (DeShon & Gillespie, 2005) and other theoretical approaches describing goal pursuit (Austin & Vancouver, 1996; Frese & Zapf, 1994), human behavior is regulated by hierarchically organized goals. These goals become activated by perceived discrepancies between a current and a desired state and are subsequently pursued in a process comprising elements such as problem exploration, planning, resource allocation, and feedback processing. In terms of this theoretical framework, reattachment activates work-related goals. Compared to nonwork goals that might be in the attentional focus during nonwork time, work-related goals become more salient while reattaching to work. Even when not actually starting work on specific tasks, early phases of goal pursuit may be initiated, for instance, by planning the upcoming day.
Reattachment to work may already start when still at home (e.g., when thinking about specific tasks that need to be done or when discussing the upcoming workday over breakfast with one’s spouse). It can occur during the commute (e.g., when mentally simulating a difficult conversation with one’s supervisor while sitting on the train) or upon arrival at work before actually starting to work (e.g., when mentally running through one’s to-do list while standing in line for a coffee). Usually, reattachment takes place before logging into a formal time-tracking system or before starting billable hours.
Development of Hypotheses
Building on earlier research (Sonnentag & Kühnel, 2016), we argue that reattachment to work in the morning predicts work engagement during the day. Work engagement refers to “the relationship between people and their work” (Schneider et al., 2018: 464) and has been conceptualized in various ways (Rich, LePine, & Crawford, 2010; Rothbard, 2001; Schaufeli & Bakker, 2004). Our research is based on Schaufeli and Bakker’s conceptualization of work engagement as “a positive, fulfilling, work-related state of mind” (2004: 295). According to Schaufeli and Bakker, engaged employees are vigorous (i.e., they feel energetic, persistent, and are willing to invest effort into their work), dedicated (i.e., they are enthusiastic, feel inspired, and perceive meaning in their work), and absorbed (i.e., they are fully immersed into their work).
Within the motivated action theory framework (DeShon & Gillespie, 2005), reattachment activates work-related goals. Because goals are highly interconnected within goal structures (Shah, Kruglanski, & Friedman, 2003), activation of work-related goals during reattachment triggers the activation of response tendencies (i.e., thoughts, affective states, actions) that are in line with the currently activated work-related goals (DeShon & Gillespie, 2005: 1111). Thus, response tendencies that potentially facilitate goal attainment become activated, whereas response tendencies that hinder goal attainment become inhibited. Although the specific response tendencies might vary to a large extent (DeShon & Gillespie, 2005: 1109), activation of work-related goals will stimulate response tendencies that help to pursue the goals by being in an optimal cognitive and affective state as well as response tendencies that mobilize resources from the environment.
With respect to a cognitive state that is crucial for goal pursuit, we examine anticipated task focus because directing one’s attention to the task is crucial for being engaged at work (Rich et al., 2010). With respect to an affective state, we focus on activated positive affect—a state that has been shown to be beneficial for goal pursuit (Bindl & Parker, 2012) and to facilitate work engagement (Bledow, Schmitt, Frese, & Kühnel, 2011). With respect to resource mobilization, we address social support and job control, two important job resources (Daniels, Beesley, Wimalasiri, & Cheyne, 2013) that are crucial for work engagement (Christian et al., 2011). Accordingly, we propose that anticipated task focus, positive activated affect, and job resources mediate the relationship between reattachment and associated goal activation on one hand and work engagement on the other hand. Figure 1 shows our conceptual model.
Conceptual Model
Reattachment as Predictor of Anticipated Task Focus, Positive Activated Affect, and Job Resources
Anticipated task focus
Reattachment to work in the morning will help draw employees’ attention to their work tasks so that they will anticipate a strong task focus. Task focus is “a problem-oriented strategy consisting of efforts to maintain concentration on the steps needed to fulfill task requirements” (Brown, Westbrook, & Challagalla, 2005: 793). Accordingly, anticipated task focus is the expectation that one will be able to keep attention on work in order to accomplish one’s tasks. By reattaching to work, work-related goals become activated and upcoming work tasks become more salient. This goal activation and mental proximity of the upcoming tasks will trigger the expectation that it is feasible to keep one’s attention on the tasks because the tasks are already mentally present. Specifically, we suggest that when employees reattach to work, they will anticipate a stronger task focus.
One core prerequisite of successful goal pursuit is to focus one’s attention on goal-relevant cues (Gollwitzer, Martiny-Huenger, & Oettingen, 2014). In typical work environments, multiple cues compete for employees’ attention (Gardner, Dunham, Cummings, & Pierce, 1989), both from the work and the nonwork domain (Smit, Maloney, Maertz, & Montag-Smit, 2016). According to motivated action theory and related theoretical approaches (DeShon & Gillespie, 2005), it is crucial to direct attention to the cues that are most important in the very moment. To succeed in goal pursuit, employees need to maintain on-task focus (Kanfer & Ackerman, 1989) by directing attention to relevant tasks and task cues and by ignoring off-task attentional demands (Beal et al., 2005). Thus, during work hours, employees need to allocate their attention to work tasks and need to refrain from thinking about nonwork issues. Regulation of attention is facilitated by the “attentional pull” (Beal et al., 2005: 1059) of the task itself, resulting—among others—from the perceived importance of the task and the presence of goals associated with this task. Thus, tasks that employees regard as meaningful in the very moment and that are linked to desirable goals receive more attention than tasks that employees cannot connect to any relevant goal. In other words, when having work-related goals activated for which the task is relevant, allocation of attention to the task is enabled.
By thinking about the tasks of the upcoming workday during reattachment and by activating work-related goals, task attentional pull becomes more likely, facilitating the direction of attention to the tasks and increasing anticipated task focus. Thus, during the process of reattachment to work, work content mentally moves into the “foreground,” whereas nonwork topics—although being highly central in most people’s lives—temporarily move into the background.
In addition, while reattaching to work and thinking about the upcoming workday, employees may deliberately decide to focus on high-priority work tasks (Tripoli, 1998), for instance, working on a report that has a tight deadline. The intention to address such high-priority and important work tasks, in turn, should increase attentional pull and anticipated task focus (Beal et al., 2005). In contrast, when employees are distracted by nonwork issues while being at work, concentration and task focus suffer (Nohe, Michel, & Sonntag, 2014). Thus, reattachment helps to mentally disconnect from nonwork issues and to place work into the focus of one’s attention. Accordingly, we propose:
Hypothesis 1a: Day-specific reattachment to work in the morning is positively associated with day-specific anticipated task focus through the activation of work-related goals.
Positive activated affect
We further propose that reattachment to work in the morning will be associated with experiencing positive activated affect at the start of the workday. Positive activated affect is a state of positive valence (i.e., pleasantness) and high arousal (Yik, Russell, & Steiger, 2011). It is experienced as excitement, energy, alertness, and determination (Watson, 1988) and has been shown to be important in the goal-striving process (Custers & Aarts, 2005). Reattachment to work triggers work-related goals and during reattachment one thinks about what will happen during the specific day. In the context of motivated action theory (DeShon & Gilllespie, 2005), employees will select actions that help to attain these goals, for example, by planning how to allocate time on various tasks or how to address a difficult work-related problem. Mentally preparing for work before actually starting work increases one’s confidence that one will be able to handle the challenges of the upcoming workday. Being more confident that one can complete the tasks that need to be accomplished will be associated with a higher level of positive affect (Fisher, Minbashian, Beckmann, & Wood, 2013), for instance, by triggering positive self-evaluations (Zuroff, Sadikaj, Kelly, & Leybman, 2016). Importantly, thinking about the workday and the tasks that have to be accomplished helps attain one’s work goals and might be perceived as a first step towards goal achievement, resulting in perceived goal progress. Experiencing goal-enhancing events (Zohar, Tzischinski, & Epstein, 2003) and perceived goal progress are associated with positive affective states (Scott, Colquitt, Paddock, & Judge, 2010).
Of course, not all work-related thoughts occurring during reattachment will be positive. Reattachment to work, however, may function as an antecedent-focused emotion regulation strategy that may help address anticipated negative events (Gross, 1998; Mauss, Bunge, & Gross, 2007). Accordingly, by thinking about an upcoming negative event and mentally simulating ways to master it, employees can mobilize positive affect that will be instrumental in addressing the event (Taylor, Pham, Rivkin, & Armor, 1998). For instance, experimental research showed that participants who mentally simulated a stressful event and the associated emotions, compared to participants in two control conditions, experienced increased positive activated states such as increased feelings of energy (Rivkin & Taylor, 1999). Thus, thinking about the upcoming workday and mentally simulating what might happen “may stir the emotions necessary to motivate an individual toward enacting problem-solving behaviors” (Szpunar, 2010: 154). Employees may mentally prepare for high-stake events by furnishing themselves with positive thoughts and by increasing their energy level. Accordingly, we propose that reattachment facilitates the experience of positive activated affect.
Hypothesis 1b: Day-specific reattachment to work in the morning is positively associated with day-specific positive activated affect in the morning through the activation of work-related goals.
Job resources
When reattaching to work and activating work-related goals, employees will also think about job resources needed to pursue these goals (DeShon & Gillespie, 2005). Among the job resources most often discussed in the organizational literature are social support and job control (Daniels et al., 2013; Karasek & Theorell, 1990). Because social support and job control foster intrinsic motivation (Schaufeli & Bakker, 2004), these resources are highly relevant for work engagement (Bakker et al., 2014). Meta-analytical evidence emphasizes the importance of social support and job control for work engagement, particularly for day-level work engagement (Christian et al., 2011).
Following Daniels, we conceptualize social support and job control as “emergent and dynamic characteristics of the job” (2006: 276) that result when employees actively interpret and shape their jobs (Wrzesniewski & Dutton, 2001) and that help in solving problems in order to achieve desired goals (Daniels et al., 2013). Thus, social support and job control may vary from day to day because employees perceive their jobs differently and—most importantly—act differently from day to day. In essence, reattachment to work should facilitate the receipt of social support and the experience of job control in a favorable way.
Social support can be defined as “the provision of emotional or instrumental help” (Parker, 2014: 665). In the job context, this help typically comes from coworkers or supervisors. In our study, we focus on social support from coworkers, given that social support from the supervisor may not be available on a daily basis (Barnes, Lucianetti, Bhave, & Christian, 2015).
Mentally reattaching to work typically implies thinking about the tasks that have to be completed during the day. When anticipating the task-accomplishment process, one might also consider potential difficulties and barriers during goal pursuit. Thus, one might realize that help from others will be needed to make progress towards one’s work goals. As a consequence, one might proactively seek out help or advice (Tims, Bakker, & Derks, 2012). For instance, one might send a short message to a coworker early during the workday so that the coworker can plan accordingly and will provide support later on. Even when not explicitly asking others for help, having thought about the need for help has activated the associated goal (DeShon & Gillespie, 2005). When running into a coworker by chance during the day, one will be better prepared to ask for help because the relevant goal (“ask for support”) is salient. These processes may eventually result in receiving some support.
Hypothesis 1c: Day-specific reattachment to work in the morning is positively associated with day-specific social support through the activation of work-related goals.
Reattachment to work will also facilitate job control during the workday. Job control is defined as the extent to which employees are “capable of controlling their tasks and general work activity” (Häusser, Mojzisch, Niesel, & Schulz-Hardt, 2010: 2). Important aspects of job control refer to an employee’s autonomy regarding when and how to perform one’s work activities (Parker, 2014). Reattachment to work in the morning will help employees experience more job control during the day at work. More specifically, when reattaching to work, employees may think about when and how they will accomplish their goals that day. By planning and mentally simulating the upcoming workday, it becomes salient to employees that various options are available regarding when and how they may complete their tasks. Thus, reattachment will increase awareness of one’s job control. In addition, by deliberately allocating tasks to specific time slots or choosing specific approaches to task accomplishment, employees may actually enact the control they have. Accordingly, research on time management indicates that setting goals and priorities—processes that might occur during reattachment—is positively associated with perceived control over time (Macan, 1994). After having mentally reattached to work and realized that they need more control for successful goal pursuit, employees may even ask their supervisors to grant them more discretion about how to accomplish work during the specific day (Rousseau, Ho, & Greenberg, 2006).
Hypothesis 1d: Day-specific reattachment to work in the morning is positively associated with day-specific job control through the activation of work-related goals.
Anticipated Task Focus, Positive Activated Affect, and Job Resources as Predictors of Work Engagement
Positive states and on-the-job experiences help employees stay engaged in work during the day (Bakker, 2014; Venz et al., 2018). We argue that anticipated task focus, positive activated affect, and job resources resulting from reattachment and goal activation in the morning are such states and experiences that facilitate work engagement during the day.
Anticipated task focus
We propose that anticipating a strong task focus at the start of the workday will predict high work engagement during the day. Expecting and anticipating that one will be able to perform a specific behavior helps substantially in performing this behavior (Bandura, 1997). Accordingly, anticipating that one will be able to maintain a strong task focus during the day will make actual task focus more likely. The belief that one will be able to concentrate and be able to direct one’s attention to the work tasks will assist in allocating regulatory resources to the tasks, thereby helping maintain task focus, even when distractions occur (Beal et al., 2005). A strong task focus is important for feeling vigorous and vital at work (Niessen, Sonnentag, & Sach, 2012; Spreitzer, Sutcliffe, Dutton, Sonenshein, & Grant, 2005). Moreover, directing one’s attention to one’s work is strongly related to the experience of feeling absorbed (Rothbard, 2001). On the basis of his qualitative research, Kahn described “maintaining a focus on tasks” (1990: 716) as an important predictor of engagement. Because anticipating a strong task focus should foster actual task focus, we propose that anticipated task focus facilitates work engagement during the workday.
Hypothesis 2a: Day-specific anticipated task focus is positively associated with day-specific work engagement during the day.
Positive activated affect
We further propose that positive activated affect in the morning will predict a high level of work engagement during the day. Positive activated affect sets a positive affective tone that will help employees approach work in a positive way—a key feature of work engagement (Schaufeli & Bakker, 2004: 205). More specifically, positive activated affect facilitates goal striving by increasing self-efficacy (Seo & Ilies, 2009), fostering goal setting (Ilies & Judge, 2005), and stimulating effort investment (Seo, Bartunek, & Barrett, 2010). Successful goal striving, in turn, should increase work engagement (Bakker, 2011). Initial empirical evidence indicates that morning positive affect is indeed positively related to work engagement later in the day (Bledow et al., 2011; McGrath, Cooper-Thomas, Garrosa, Sanz-Vergel, & Cheung, 2017). Thus, we hypothesize that positive activated affect at the beginning of a workday will be associated with higher work engagement during the workday.
Hypothesis 2b: Day-specific morning positive activated affect is positively associated with day-specific work engagement during the day.
Job resources
We further suggest that social support and job control will help individuals experience high levels of work engagement during the workday. Social support and job control are typical job resources that have been identified as core predictors of work engagement (Bakker et al., 2014; Schaufeli & Bakker, 2004). Schaufeli and Bakker (2004) described two pathways through which job resources such as social support and job control should predict work engagement. First, social support and job control satisfy basic human needs (Ryan & Deci, 2000) and stimulate motivational processes by enhancing meaningfulness and psychological safety (Kahn, 1990). Thus, by feeling supported and being aware of one’s job control, one will be more intrinsically motivated to immerse oneself into one’s work, to become absorbed, and to feel energetic while doing one’s work. Second, by facilitating goal achievement, social support and job control are instrumental in getting one’s work done (Schaufeli & Bakker, 2004). More specifically, social support fosters confidence in one’s ability to overcome difficulties in goal pursuit (Xanthopoulou et al., 2008), and job control allows one to choose the procedures and tools that make goal achievement most likely (Parker, 2014). As a result, employees will be more willing to invest effort into their work and dedicate themselves fully to their work—thus, they will experience higher work engagement. Accordingly, past research indicates that employees who experience more social support and job control report higher levels of work engagement (Christian et al., 2011; Crawford et al., 2010). Moreover, studies examining work engagement at the day level found that on days when employees feel more supported (Simbula, 2010; Xanthopoulou et al., 2008) and on days when they enjoy a higher level of job control (Breevaart et al., 2014; Kühnel et al., 2012), they are more engaged at work. In our study, we build on these earlier findings by examining the day-level relationships between social support and job control on one hand and work engagement on the other hand.
Hypothesis 2c: Day-specific social support is positively associated with day-specific work engagement during the day.
Hypothesis 2d: Day-specific job control is positively associated with day-specific work engagement during the day.
Indirect Effects of Reattachment on Work Engagement
Building on the motivated action theory framework (DeShon & Gillespie, 2005), we have argued that reattachment to work in the morning is linked to higher anticipated task focus, activated positive affect, job control, and social support via the activation of work-related goals because thinking about the upcoming workday during reattachment makes work-related goals highly salient. Anticipated task focus, positive activated affect, job control, and social support, in turn, should be associated with work engagement during the workday. Thus, reattachment to work in the morning and the associated activation of work-related goals contributes to beneficial cognitive and affective states (i.e., anticipated task focus and activated positive affect) that, in turn, should facilitate work engagement. Accordingly, when employees reattach to work in the morning, they should be more engaged during the day because they have already anticipated high task focus (Rich et al., 2010) and because they experience activated positive affect (Bledow et al., 2011). In addition, reattachment to work in the morning and the associated activation of work-related goals draws employees’ attention to job resources they potentially need to accomplish their goals. As a consequence, when employees reattach to work in the morning, they are more engaged during the day because they may enact job resources they need for pursuing work-related goals. Thus, reattachment prepares employees to be in an optimal state and to have resources available so that they can be engaged at work. Resources such as social support and job control are particularly important for work engagement (Christian et al., 2011).
Hypothesis 3: Day-specific reattachment to work in the morning shows an indirect relationship with day-specific work engagement during the day through the activation of work-related goals and (a) via anticipated task focus, (b) via morning positive activated affect, (c) via social support, and (d) via job control.
Method
Procedure and Sample
To recruit participants, we distributed study information via professional online networks and social media (e.g., xing.com). We described our study as research on “starting the workday.” To be eligible to participate, employees had to work 5 days per week and have access to the Internet at their workplace. Participants were invited to take part in a lottery in which they could win gift cards (€20) from an online retailer.
At the onset of data collection, participants provided demographic and other background information (e.g., trait affectivity) in an initial online survey. Participants were then asked to complete two short online surveys per day, over the period of 1 workweek (5 days from Monday to Friday). Specifically, participants were sent e-mail invitations including a link to the online survey twice a day, one for a morning survey and one for an afternoon survey. The morning survey had to be completed upon arrival at the workplace; the afternoon survey had to be completed at the end of the workday.
A total of 172 employees expressed interest in participating in the study by providing demographic and background information in the initial survey. With respect to the daily surveys, 170 participants provided responses in the morning, on a total of 769 days (4.52 morning surveys per person), and 167 participants provided responses in the afternoon on a total of 740 days (4.43 afternoon surveys per person). Time-stamp information collected with the daily-survey data indicated that some of the daily surveys were completed at wrong times (e.g., the morning survey was completed at the same time as the afternoon survey or a survey was completed on the next day). These data were removed from the data set, resulting in 713 morning responses from 163 participants (4.37 morning surveys per person), and 682 afternoon responses from 164 participants (4.16 afternoon surveys per person). Matching morning surveys with afternoon surveys on the respective days resulted in a total of 627 day-level data sets from a total of 158 participants (3.97 days per person). To capture within-person variation in our study variables (Nezlek, 2011), we excluded 7 study participants who had valid data on just 1 day, resulting in our final data set of 620 day-level data sets from 151 participants (4.11 days per person).
About half of this final sample of 151 persons was female (55.5%). Mean age was 38.1 years (SD = 10.8), mean organizational tenure was 8.9 years (SD = 9.2), and mean contract working time was 40.0 hours (SD = 6.8). On average, participants were highly educated, with 57.5% holding a university degree. Participants worked in a broad range of industries, with most participants working in the private service sector (e.g., finance industry; 20.5%), energy sector (13.0%), public administration (13.0%), information and communication sector (11.0%), or health sector (9.6%). Among all participants, 18.5% held a leadership position.
Measures
We assessed our study variables in two surveys per day. All items were in German. Table 1 shows descriptive information and zero-order correlations for the study variables.
Zero-Order Correlations, Means, Standard Deviations, and Intraclass Coefficients
Note: Means and standard deviations at the between-person level are displayed in Columns 8 and 9; means and standard deviations at the within-person level are displayed in Columns 10 and 11. Correlations below the diagonal are between-person correlations (N = 151), with correlations of |r| ≥ .21 being significant at p < .01 and correlations of |r| ≥ .16 being significant at p < .05. Correlations above the diagonal are within-person correlations (620 days), with correlations of |r| ≥ .11 being significant at p < .01 and correlations of |r| ≥ .08 being significant at p < .05. Within-person correlations do not take into account the nested data structure. ICC = percentage of variance between persons (ICC = variance between persons / (variance between persons + variance within)).
We assessed day-specific reattachment in the morning survey with five items developed by Sonnentag and Kühnel (2016); answers were provided on a 5-point scale (1 = I fully disagree; 5 = I fully agree). These items capture morning experiences that help people mentally reconnect with work before starting the actual workday. 1 Participants responded to the items with respect to “Before I started my work . . .” and answered these specific items: “. . . this morning, I mentally prepared for it,” “. . . this morning, I mentally tuned into my work,” “. . . this morning, I gave some thought to the upcoming workday,” “. . . this morning, I thought about what I wanted to achieve at work today,” and “. . . this morning, I thought about what I will encounter at my work today.” Cronbach’s alpha computed separately for the 5 days ranged between .92 and .96 (mean alpha = .94).
To assess goal activation in the morning, we used three items from Snyder, Sympson, Ybasco, Borders, Babyak, and Higgings (1996). We slightly adjusted the items so that they addressed “work goals” instead of “goals” in general. A sample item was “At the present moment, I am energetically pursuing my work goals.” Participants responded on a 4-point scale (1 = definitely false; 4 = definitely true). Cronbach’s alpha ranged from .72 to .85 (mean alpha = .77).
We measured day-specific anticipated task focus in the morning by adapting three items from the measure used by Nohe et al. (2014). Whereas the Nohe et al. original items captured a retrospective report of task focus, we aimed at assessing anticipated task focus. Accordingly, our items were “I think it will be easy for me to keep my mind on work today,” “I think I will fully concentrate on my work today,” and “I think that my thoughts will wander during work today” (reverse coded). Participants provided their answers on a 5-point scale (1 = I fully disagree; 5 = I fully agree). Cronbach’s alpha ranged between .87 and .91 (mean alpha = .89).
We assessed day-specific positive activated affect in the morning with positive-affect items from the Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988). To keep the daily survey reasonably short, we presented a subset of six items used in earlier research (Sonnentag, Binnewies, & Mojza, 2008), specifically “active,” “interested,” “excited,” “strong,” “inspired,” and “alert.” Participants were instructed to report how they felt at the moment on a 5-point scale (1 = not at all; 5 = very much). Cronbach’s alpha ranged between .81 and .91 (mean alpha = .88).
We measured day-specific social support from coworkers in the afternoon, using two items from the German version of the Copenhagen Psychosocial Questionnaire (Kristensen, Hannerz, Høgh, & Borg, 2005; Nübling, Stößel, Hasselhorn, Michaelis, & Hofmann, 2011). We adapted the items for day-specific assessment on a 5-point scale (1 = not at all; 5 = very much). The items were “To what extent did you receive help and support from your colleagues today?” and “To what extent were your colleagues willing to listen to your work-related problems today?” The day-level correlation (r) between the two items ranged from .66 to .76 (mean r = .71).
We assessed job control in the afternoon, using three items from the Job Content Questionnaire (Karasek, Brisson, Kawakami, Houtman, Bongers, & Amick, 1998) adapted for day-level assessment. The items were “During work today, I could make many decisions myself,” “During work today, I had a lot of say about how things were done,” and “During work today, I had little decision latitude about how I do my work” (reverse coded). We used a 4-point response scale (1 = I fully disagree; 4 = I fully agree). Cronbach’s alpha ranged between .82 and .85 (mean alpha = .83).
We assessed our outcome variable work engagement in the afternoon with eight items from the Utrecht Work Engagement Scale (UWES; Schaufeli, Bakker, & Salanova, 2006) adapted for day-level assessment. The UWES is a frequently used work-engagement measure (Christian et al., 2011) and has been widely applied in day-level assessments (Lanaj, Johnson, & Barnes, 2014; Uy, Lin, & Ilies, 2017). We deliberately omitted one of the original nine items from the analysis (“When I got up in the morning, I felt like going to work”) because it captures an experience before work instead of during work. 2 Sample items were “During my work, I felt strong and vigorous today,” “Today, I was enthusiastic about my job,” and “Today, while I was working, I forgot everything else around me.” Participants provided their answers on a 5-point scale (1 = I fully disagree; 5 = I fully agree). Cronbach’s alpha ranged between .92 and .95 (mean alpha = .93).
To examine the construct validity of our study measures, we performed a set of multilevel confirmatory factor analyses. A seven-factor model with all factors modeled at the between-person as well as the within-person level and with all items loading on their respective factors showed an acceptable fit—χ2 = 1,401.94, df = 769, comparative fit index (CFI) = .93, Tucker-Lewis index (TLI) = .92, root mean square error of approximation (RMSEA) = .04—and fit the data better than alternative models (see Table 2).
Findings From Confirmatory Factor Analysis
Note: The residual variance of one item assessing anticipated task focus was set to 0 at the between-person level in all models. Model 2a: Goal activation and anticipated task focus loading on one common factor. Model 2b: Goal activation and work engagement loading on one common factor. Model 2c: Social support and job control loading on one common factor. Model 2d: Reattachment and goal activation loading on one common factor. Model 2e: Positive activated affect and work engagement loading on one common factor. Model 2f: Anticipated task focus and work engagement loading on one common factor. Model 2g: Reattachment and anticipated task focus loading on one common factor. Model 2h: Reattachment and work engagement loading on one common factor. Model 2i: Goal activation and activated positive affect loading on one common factor; model did not converge. Model 3: All items assessed in the afternoon (social support, job control, work engagement) loading on one common factor. Model 4: All items assessed in the morning (reattachment, goal activation, anticipated task focus, positive activated affect) loading on one common factor; model did not converge. Model 5: All items assessed in the morning (reattachment, goal activation, anticipated task focus, positive activated affect) loading on a first common factor and all items assessed in the afternoon (social support, job control, work engagement) loading on a second common factor; model did not converge. Model 6: All indicators loading on one common factor. CFI = comparative fit index; TLI = Tucker-Lewis index; AIC = Akaike information criterion; RMSEA = root mean square error of approximation; SCF = scaling correction factor; S-B χ2 = Satorra-Bentler χ2, referring to the comparison with the seven-factor model.
Statistical Analyses
Because our daily-survey design resulted in a hierarchically structured data set (days nested within persons), we tested our hypotheses with a multilevel approach. Specifically, we used a path-analytic approach following the multilevel structural equation modeling framework (Preacher, Zyphur, & Zhang, 2010), using Mplus, Version 7.4 (Muthén & Muthén, 1998-2015). To arrive at unbiased estimates for the indirect relationships, we followed the recommendation of Preacher et al. (2010) and modeled paths at the within-person (Level 1) and the between-person (Level 2) level. Because variables assessed at Level 1 are implicitly partitioned into within-person and between-person variance components (Preacher et al., 2010: 215), we refrained from centering our variables. We tested fixed effects at the within-person level and allowed the residuals of anticipated task focus, positive activated affect, social support, and job control to correlate at the within-person and the between-person level. Our main analysis focused on processes happening within 1 day.
Results
Variance Decomposition
We examined intraclass correlations (ICC1s) to gain insight into the variance components at the within-person and between-person level. ICC1s ranged between .48 and .64 (see Table 1), indicating that between 52% and 36% of the total variance of our study variables was within-person variance. Thus, the ICC1s demonstrate that there is sufficient within-person and between-person variance in all study variables, allowing for a proper test of our hypotheses.
Hypothesis Tests
We tested our hypotheses in an overall model in which we specified paths at the within-person and between-person level. This model had a good fit (χ2 = 10.745, df = 8, CFI = .996, TLI = .981, RMSEA = .024).
We hypothesized that reattachment to work in the morning is associated with anticipated task focus, activated positive affect, and job resources (social support, job control) via goal activation and that anticipated task focus, activated positive affect, and job resources, in turn, are associated with work engagement during the day. We first examined whether reattachment predicted goal activation. This was indeed the case (see Table 3). Table 4 displays the unstandardized path coefficients for the prediction of the four distal mediator variables (anticipated task focus, positive activated affect, social support, and job control). Relevant for our hypothesis tests are the estimates at the within-person level. Results indicate that goal activation was positively related to anticipated task focus, positive activated affect, social support, and job control at the within-person level. To examine the indirect effects of reattachment to anticipated task focus, positive activated affect, social support, and job control via goal activation, we computed confidence intervals (CIs) using the Monte Carlo method (Preacher & Selig, 2010) with 20,000 repetitions. The indirect effects from reattachment to the distal mediators via goal activation were significant for anticipated task focus (indirect effect = 0.093, 95% CI = [0.0499, 0.1459]), positive activated affect (indirect effect = 0.058, 95% CI = [0.0277, 0.0939]), social support (indirect effect = 0.036, 95% CI = [0.0023, 0.0730]), and job control (indirect effect = 0.030, 95% CI = [0.0032, 0.0577]), providing support for Hypotheses 1a, 1b, 1c, and 1d. As shown in Table 4, the direct relationship between reattachment and social support was negative, rendering the total effect from reattachment to social support nonsignificant (unstandardized estimate = −0.055, SE = 0.045, z = −1.218, p = .223).
Unstandardized Coefficients From Multilevel Path Model Predicting Goal Activation
Note: Estimates are unstandardized, resulting from one overall analysis including the prediction of goal activation, of the distal mediators, and of work engagement in one model.
Unstandardized Coefficients From Multilevel Path Model Predicting the Distal Mediators
Note: Estimates are unstandardized, resulting from one overall analysis including the prediction of goal activation, of the distal mediators, and of work engagement in one model.
Table 5 shows the unstandardized path coefficients for the prediction of work engagement. Anticipated task focus, positive activated affect, social support, and job control were positively related to work engagement at the within-person level, supporting Hypotheses 2a, 2b, 2c, and 2d. We tested the serial indirect effects from reattachment via goal activation to work engagement by using the Monte Carlo method (Preacher & Selig, 2010) with 20,000 repetitions. In line with Hypotheses 3a, 3b, 3c, and 3d, the serial indirect effects of reattachment via goal activation on work engagement were significant for anticipated task focus (indirect effect = 0.008, 95% CI = [0.0010, 0.0173]), positive activated affect (indirect effect = 0.021, 95% CI = [0.0088, 0.0350]), social support (indirect effect = 0.005, 95% CI = [0.0002, 0.0117]), and job control (indirect effect = 0.009, 95% CI = [0.0008. 0.0191]). 3
Unstandardized Coefficients From Multilevel Path Model Predicting Work Engagement
Note: Estimates are unstandardized, resulting from one overall analysis including the prediction of goal activation, of the distal mediators, and of work engagement in one model.
Supplementary Analyses
Individual differences play a role in goal-related processes (DeShon & Gillespie, 2005). In people’s everyday work situations, individual differences in affectivity are particularly important (Judge & Ilies, 2004) because they shape individuals’ perceptions of themselves and their environments as well as their reactions to these perceptions (Kaplan, Bradley, Luchman, & Haynes, 2009). To explore whether individual differences in affectivity matter for the processes tested in our study, we conducted a set of supplementary analyses examining trait negative affectivity and trait positive affectivity as moderators of the relationships between (a) morning reattachment to work and goal activation and (b) goal activation on one hand and anticipated task focus, activated positive affect, social support, and job control on the other hand. Trait negative affectivity refers to a person’s “tendency to experience aversive emotional states, such as fear, guilt, sadness, and anger” (Watson, Wiese, Vaidya, & Tellegen, 1999: 829), whereas trait positive affectivity refers to a person’s tendency to experience “positive states, such as enthusiasm, confidence, and cheerfulness” (Watson et al., 1999: 829).
One could expect that when employees high on trait negative affectivity think about their work before actually starting to work, they will think more about possible negative features of their work and might worry more about the upcoming workday (Watson & Clark, 1984). By thinking more about negative features associated with work, it will be less likely that work-related goals become activated. In addition, because they may have lower expectations of goal attainment (Seo, Barrett, & Bartunek, 2004), they will be less likely to set in motion the processes that will facilitate goal attainment, namely, anticipated task focus, positive affect, and of job resources. Employees high on trait positive affectivity, however, will be more likely to reattach to work in a positive affective state that, in turn, will help to activate positively toned work-related goals (Bindl, Parker, Totterdell, & Hagger-Johnson, 2012; Seo & Ilies, 2009). They will have higher expectations to attain their goals, implying that they will put more effort in the goal-striving process (Seo et al., 2004), which should be reflected in higher anticipated task focus, higher positive activated affect, and a higher level of job resources.
We assessed trait negative affectivity and trait positive affectivity as background information in the initial survey, instructing participants to report how they feel “in general.” Specifically, we used six PANAS items from the negative-affect scale to assess trait negative affectivity (Watson et al., 1988; Cronbach’s alpha = .80) and six PANAS items from the positive-affect scale to assess positive affectivity (Watson et al., 1988; Cronbach’s alpha = .86). We examined the role of trait negative and positive affectivity for the relationship between reattachment and the mediators via goal activation by testing cross-level interaction effects of trait negative affectivity and trait positive affectivity. To achieve convergence of these relatively complex multilevel models with random slopes on the day level, we tested unconflated multilevel models (UMMs; i.e., modeling the between-person part of the model by using person-mean scores of the day-level variables and by using person-mean centered predictor and mediator variables at the day level; 4 Preacher et al., 2010) and ran separate models for trait negative affectivity and trait positive affectivity as cross-level moderators. Trait negative affectivity moderated the relationship between reattachment and goal activation (unstandardized estimate of trait negative affectivity predicting the random slope between reattachment and goal activation = −0.086, SE = 0.043, z = −2.004, p = .045). The simple slope was weaker for high trait negative affectivity (+1 SD; unstandardized estimate = 0.122, SE = 0.038, z = 3.171, p = .002) than for low trait negative affectivity (−1 SD; unstandardized estimate = 0.229, SE = 0.038, z = 5.958, p < .001), and the difference between the simple slopes was significant (unstandardized estimate = −0.107, SE = 0.053, z = −2.004, p = .045). We tested conditional indirect effects between reattachment and the distal mediators via goal activation at high (+1 SD) versus low (−1 SD) levels of trait negative affectivity. Most of the conditional indirect effects were significant and did not differ between high versus low levels of trait negative affectivity. Specifically, at high levels of trait negative affectivity, the indirect effect was significant for anticipated task focus (indirect effect = 0.060, 95% CI = [0.0228, 0.1027]), positive activated affect (indirect effect = 0.039, 95% CI = [0.0124, 0.0727]), social support (indirect effect = 0.024, 95% CI = [0.0019, 0.0546]), and job control (indirect effect = 0.020, 95% CI = [0.0028, 0.0423]). At low levels of trait negative affectivity, the indirect effect was significant for anticipated task focus (indirect effect = 0.113, 95% CI = [0.0649, 0.1694]), positive activated affect (indirect effect = 0.073, 95% CI = [0.0323, 0.1233]), social support (indirect effect = 0.045, 95% CI = [0.0039, 0.0924]), and job control (indirect effect = 0.038, 95% CI = [0.0060, 0.0734]). Although the estimates were higher for high trait negative affectivity than for low trait negative affectivity, the differences between the respective indirect effects at high versus low levels of trait negative affectivity did not reach the conventional significance level for anticipated task focus (unstandardized estimate = −0.053, SE = 0.029, z = −1.841, p = .066), positive activated affect (unstandardized estimate = −0.034, SE = 0.020, z = −1.720, p = .085), social support (unstandardized estimate = −0.021, SE = 0.015, z = −1.426, p = .154), or job control (unstandardized estimate = −0.018, SE = 0.013, z = −1.358, p = .175).
Similarly, we calculated conditional serial indirect effects between reattachment and work engagement via goal activation and the distal mediators at high (+1 SD) versus low (−1 SD) levels of trait negative affectivity. At high levels of trait negative affectivity, the indirect effect was significant for anticipated task focus (indirect effect = 0.005, 95% CI = [0.0005, 0.0125]), positive activated affect (indirect effect = 0.014, 95% CI = [0.0031, 0.0269]), and job control (indirect effect = 0.006, 95% CI = [0.0003, 0.0128]) but not for social support (indirect effect = 0.004, 95% CI = [−0.0002, 0.0085]). At low levels of trait negative affectivity, the indirect effect was significant for anticipated task focus (indirect effect = 0.010, 95% CI = [0.0012, 0.0212]), positive activated affect (indirect effect = 0.026, 95% CI = [0.0096, 0.0457]), social support (indirect effect = 0.007, 95% CI = [0.0001, 0.0155]), and job control (indirect effect = 0.012, 95% CI = [0.0005, 0.0239]). Again, the differences between the respective serial indirect effects at high versus low levels of trait negative affectivity did not reach the conventional significance level for anticipated task focus (unstandardized estimate = −0.005, SE = 0.003, z = −1.452, p = .146), positive activated affect (unstandardized estimate = −0.012, SE = 0.007, z = −1.702, p = .089), social support (unstandardized estimate = −0.003, SE = 0.002, z = −1.370, p = .171), or job control (unstandardized estimate = −0.005, SE = 0.004, z = −1.330, p = .184).
Trait negative affectivity did not moderate any of the relationships between goal activation and the mediators (anticipated task focus, positive activated affect, social support, job control). Trait positive affectivity was not a cross-level moderator in either the relationship between reattachment and goal activation or the relationships between goal activation and the mediators.
In an additional analysis, we examined whether work engagement predicts reattachment on the next day. We specified a model in which we added within-person and between-person paths from work engagement to next-morning reattachment to our main model (as summarized in Figure 1). This new model had a poor fit (χ2 = 270.045, df = 20, CFI = .786, TLI = .400, RMSEA = .142), and the within-person path from work engagement to next-morning reattachment was not significant (unstandardized estimate = −0.085, SE = 0.238, z = −0.356, p = .721).
Discussion
During the past decade, research has paid substantial attention to the question of how employees can gain mental distance from their work during nonwork hours and how this distance from work relates to employee outcomes (Bennett, Bakker, & Field, 2018; Sonnentag & Fritz, 2015). Our study adds to more recent research that emphasizes the complementary experience of mentally reconnecting with work (Sonnentag & Kühnel, 2016). Specifically, the present study showed that on days on which employees better mentally reattach to work in the morning, they anticipate a higher task focus, experience higher activated positive affect, receive more social support, and experience more job control due to early activation of work-related goals. In turn, anticipated task focus, activated positive affect, social support, and job control are positively associated with work engagement during the day.
Theoretical Contributions
Our study is the first to examine the mechanisms underlying the relationship between reattachment to work in the morning and work engagement later in the day. Therefore, our study helps explain why and how reattachment to work in the morning relates to work engagement the same day. Drawing on motivated action theory (DeShon & Gillespie, 2005), reattachment can be seen as building a mental bridge to work early during the day, thereby activating work-related goals before actually starting work. This early activation of work-related goals, in turn, helps to prepare for the upcoming day by stimulating subsequent cognitive (i.e., anticipated task focus) and affective (i.e., activated positive affect) processes, as well as processes related to the mobilization of job resources (i.e., social support and job control). Our study adds to motivated action theory by highlighting the specific processes operating at a relatively low level of a goal hierarchy (i.e., at the level of action-plan goals as opposed to achievement or principle goals). It is important to note not only that cognitive and behavioral processes play a role but also that affective processes seem to help in translating reattachment into work engagement.
Our results further contribute to research on boundary crossing between life domains (Ashforth et al., 2000). Our study builds on early research on employees’ transition styles (Hall & Richter, 1988) by showing that employees’ degree of reattachment to work in the morning (i.e., enactment of their individual transition style) fluctuates from day to day. In the context of boundary theory (Ashforth et al., 2000), this finding indicates that employees’ mental boundary crossing between home and work is not a stable factor but undergoes substantive within-person fluctuation from day to day. Importantly, the extent to which employees reconnect to work in the morning is related to subsequent experiences during the workday. The strong within-person fluctuation of reattachment to work (also known as mental boundary crossing) suggests that the boundaries between life domains are highly volatile and are subject to day-to-day adjustments.
Our supplementary analyses showed that the association between reattachment and goal activation was weaker for employees high in trait negative affectivity. Possibly, these persons might not only think about their goals when reattaching to work but also dwell on negative aspects of their work. Interestingly, however, although the association between reattachment and goal activation was weaker when trait negative affectivity was high, neither the indirect effects from reattachment to anticipated task focus, activated positive affect, and job resources nor the serial indirect effects from reattachment to work engagement differed significantly between persons high versus low in trait negative affectivity. Possibly, the association between reattachment and goal activation is still sufficiently strong for persons high in trait negative affectivity so that they benefit from reattaching to work as well.
Reattachment to work in the morning showed a positive indirect relationship with social support via goal activation but a negative direct relationship. Thus, it might be that those aspects of reattachment that are not channeled into goal activation reduce an employee’s likelihood of receiving social support. In addition, on days when employees do not reattach well to work in the morning, they will need to seek social support more spontaneously in order to address emerging problems.
One might argue that reattaching to work before actually starting work at the beginning of a workday is a form of worry about how the workday may unfold (Brosschot, Gerin, & Thayer, 2006). Thus, employees might primarily think about problems that could occur during the day. Accordingly, such negative thoughts would have more negative affective consequences. Although we did not assess the specific thought content occurring during reattachment, in our view, it is unlikely that reattachment is dominated by worry. If reattachment was mainly characterized by a negative tone, it might dampen positive affect. Reattachment, however, was strongly associated with activated positive affect.
Limitations and Directions for Future Research
Despite its contributions, this study is not without limitations. First, we assessed all our data with self-report measures, which might raise concerns about common method bias. However, because we tested our research model simultaneously at the within-person and the between-person level, we can rule out the possibility that interindividual differences (e.g., social desirability) have influenced the within-person findings. Nevertheless, there still might be unmeasured within-person third variables that played a role, such as recovery during the previous evening. Sonnentag and Kühnel (2016), however, found that day-specific reattachment was a significant predictor of day-specific work engagement, even when controlling for psychological detachment from work during the previous night. Future research might want to reduce common method bias by including data from other sources, such as assessments of work engagement through coworkers (Venz & Sonnentag, 2015).
Second, we collected data only twice per day. We measured goal activation and two of our mediator variables together with the predictor variable reattachment and two of our mediator variables together with the outcome variable work engagement. It would have been preferable to assess the mediator variables at a third measurement occasion during the workday. Despite this potential limitation, we are confident that measuring the mediator variables together with the predictor and the outcome variable, respectively, has not driven our major research findings. Specifically, reattachment measured in the morning was related to job control measured in the afternoon, and anticipated task focus and activated positive affect measured in the morning were related to work engagement measured in the afternoon.
Third, our data do not provide information about exactly where and when reattachment occurred and whether the specific place and specific time of reattachment matters for how employees experience and shape the upcoming workday. For instance, it might be important to reattach early in the morning (e.g., when still at home), or it might be that the time when physically arriving at work is most important. Therefore, future research should address the timing of reattachment in more detail and examine when reattachment is most beneficial.
Our study relied on Schaufeli and Bakker’s (2004) conceptualization of work engagement, thereby neglecting the differentiation between physical, cognitive, and emotional engagement (Rich et al., 2010). Future research should include other conceptualizations of work engagement to examine whether reattachment and the mediators uniformly relate to the distinct facets of engagement. Possibly, anticipated task focus may matter more for cognitive engagement while positive affect may matter more for emotional engagement.
Future research should also explore the actual experience of reattaching to work as well as its predictors and the specific goals that are activated. With respect to the reattachment experience, place and time of reattachment should be considered as well as thought content and affective tone. In addition, it would be helpful to explore the role of online technology in the reattachment experience. More and more employees are using online technology to stay connected to their work even when physically away from their work (Braukmann, Schmitt, Ďuranová, & Ohly, 2017; Lanaj et al., 2014). Checking online messages and online planning tools early in the morning might be part of reattachment and may even occur without much conscious thought (van Deursen, Bolle, Hegner, & Kommers, 2015) before deliberately thinking about the upcoming workday.
Future studies might want to use latent-growth modeling and explore how reattachment develops over the course of a workweek. Such an approach will make it particularly important to examine measurement invariance of the reattachment construct over time (Hertzog & Nesselroade, 2003).
In addition to trait negative and trait positive affectivity, future research should examine other cross-level moderators. On the basis of motivated action theory, one could argue that those goals that are most salient for a person become most easily activated during reattachment. For instance, for an employee high in mastery-goal orientation, day-specific goals related to learning will become more easily activated than day-specific goals mainly relevant for impression management.
Future research should also consider cross-level moderators beyond the framework of motivated action theory. For instance, preference for integration versus segmentation of life domains (Kreiner, 2006) or self-complexity (Linville, 1987) might play a role in the relationship between reattachment and work engagement. Possibly, employees preferring segmentation and employees with higher self-complexity need to engage in a more elaborate reattachment process in order to trigger the cognitive, affective, and resource-related processes needed for work engagement, whereas employees preferring integration and employees with lower self-complexity (and high salience of work within their self-concept) may not need explicit reattachment to feel highly engaged at work because their work-related goals may remain activated even when they are away from the workplace.
As suggested by Sonnentag and Kühnel (2016) future research should explore predictors of reattachment to work. In terms of person-level predictors, studies should examine both job characteristics (e.g., job complexity, predictability of job tasks) and individual factors (e.g., job involvement, conscientiousness). At the within-person level, research might look at the day of the week (e.g., more reattachment might be needed on Mondays than on others days of the week) or at anticipated work demands (Casper, Sonnentag, & Tremmel, 2017). In addition, the location and timing of reattachment should be examined more explicitly. For instance, some employees might prefer to develop a habit of reattaching to work during the commute via public transport, while others might choose to develop the routine of mentally reattaching to work while walking from the parking lot to the office building or while standing in line in the onsite coffee shop.
Finally, research should include a broader set of outcome variables beyond work engagement. Importantly, it remains to be seen whether reattachment to work is beneficial for task performance and other performance indicators such as proactive work behavior. In addition, future research may explore the impact of reattachment on well-being and strain processes occurring during the day. For instance, one could argue that reattachment helps to better deal with the demands emerging during the day, which, in turn, might reduce the occurrence of strain symptoms.
Practical Implications
In terms of practical implications, our study suggests that employees’ work engagement benefits from deliberately reattaching to work in the morning. Accordingly, employees should be encouraged to actively reattach to work before starting work. Managers could support their employees’ reattachment by allowing for a few quiet minutes after employees arrive at work or by initiating a short planning conversation about the upcoming workday. It is particularly important that work-related goals are activated early during the workday. For instance, employees may take a close look at their to-do lists and decide about the most important goals that need to be achieved during the day. Also, managers may help with goal activation by encouraging employees to prioritize their most important goals. Organizations might even think about explicitly stimulating employees’ tuning into work by providing short checklists that employees could use when reattaching to work.
More specifically, during reattachment and goal activation at the start of the workday, employees should anticipate high task focus, maybe by thinking about temporal and situational features that enable high task focus, for instance, by identifying quiet periods during the upcoming day or by negotiating an uninterrupted hour (König, Kleinmann, & Höhmann, 2013). Reattachment and goal activation should help employees to enact confidence about what they can accomplish during the day (Fisher et al., 2013), for instance, by thinking about what they have accomplished in the past (Seo et al., 2004). Social support and job control should follow from reattachment and goal activation. Specifically, when activating work-related goals at the onset of the workday, employees may want to take into account that during the day, they might need help from others and, accordingly, should plan when and how to request social support. In addition, through reattachment and goal activation, employees may become aware of multiple approaches to task accomplishment, and they may want to decide about the most suitable approach. To be able to follow through on the most suitable approach, they need to use the autonomy the job provides or negotiate more autonomy for specific tasks (Daniels, 2006).
Conclusion
Taken together, this study brought to light the multiple pathways by which reattachment to work in the morning relates to work engagement during the day. Results suggest that reattaching to work before actually starting work sets the tone for the workday through anticipating high task focus, mobilizing positive activated affect, and recognizing available job resources. Task focus, positive activated affect, and the availability of job resources, in turn, enable high work engagement. Reattachment to work in the morning is a low-effort strategy that employees can use in order to start their workday in an optimal way. Organizations may develop norms and routines that help employees reattach to work and that support them in smoothly transitioning into the workday.
Supplemental Material
ONLINE_SUPPLEMENT – Supplemental material for Morning Reattachment to Work and Work Engagement During the Day: A Look at Day-Level Mediators
Supplemental material, ONLINE_SUPPLEMENT for Morning Reattachment to Work and Work Engagement During the Day: A Look at Day-Level Mediators by Sabine Sonnentag, Kathrin Eck, Charlotte Fritz and Jana Kühnel in Journal of Management
Footnotes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
