Advertisement

Occupational Health Science

, Volume 2, Issue 4, pp 337–363 | Cite as

Workplace Telepressure and Worker Well-Being: The Intervening Role of Psychological Detachment

  • Alecia M. Santuzzi
  • Larissa K. Barber
Major Empirical Contribution

Abstract

Workplace telepressure—an employee’s preoccupation and urge to respond quickly to work-related messages via information and communication technologies (ICTs)—may be associated with negative well-being outcomes for workers. The present study expands upon past work on ICT-related stressors and worker well-being with an examination of the presumed role of lower psychological detachment from work in the relationships between workplace telepressure and negative worker outcomes. A three-wave web-based survey with 234 employed adults confirmed between-person associations between workplace telepressure and lower psychological detachment from work, higher levels of exhaustion (physical and cognitive), and more sleep problems. Moreover, results supported the predicted indirect effect of workplace telepressure to physical exhaustion and sleep problems through psychological detachment at the between-person level. Results also showed a negative indirect effect of workplace telepressure through psychological detachment on within-person variation in work engagement, despite the positive bivariate association between workplace telepressure and engagement (absorption). Finally, exploratory analyses suggested that workplace telepressure might be a stronger predictor of exhaustion when ICT connection demands at work are low. We discuss implications for workplace telepressure in terms of both health impairment and motivational processes with respect to work recovery.

Keywords

Telepressure Psychological detachment Work recovery Technology Employee well-being 

Message-based information and communications technologies (ICTs) such as email and text messages may increase flexibility and convenience in responding to work-related requests (Harpaz 2002) and facilitate team collaboration across geographic and other accessibility barriers (Gilson et al. 2015). However, workers who use ICTs at work also feel the need to be continuously connected to work, which can negatively affect worker health and well-being in a variety of ways (Day et al. 2010; Mazmanian et al. 2013; Olson-Buchanan and Boswell 2006). For example, using ICT for work purposes while at home has been linked to higher reports of work-life conflict, burnout, and poorer sleep quality (e.g., Barber and Jenkins 2014; Derks and Bakker 2014; Derks et al. 2014; Ferguson et al. 2016; Olson-Buchanan and Boswell 2006).

Such findings suggest that the flexibility of ICTs for work may encourage behaviors that are detrimental to employee well-being. Being able to connect to work from anywhere, at any time, may lead employees to believe that they are expected to do so. Importantly, employees might feel pressure to stay technologically connected to work and ready to respond during times when they may need a break from work demands. Researchers have proposed that high levels of preoccupation and urge to respond quickly to ICT messages (i.e., workplace telepressure; Barber and Santuzzi 2015) could interfere with recovery time from work demands. Yet, this proposition has not empirically tested with respect to the role of recovery processes in linking workers’ concerns about responding to ICT messages to stress and well-being.

Guided by recovery theories in occupational health psychology (Bakker and Demerouti 2007; Sonnentag and Fritz 2015) and previous work on ICT use and employee recovery processes (Derks and Bakker 2014; Park et al. 2011), the current study aims to make two key contributions to the ICT and employee well-being literature. First, we explored whether workplace telepressure indirectly and negatively relates to worker well-being through interference with a key recovery process: psychological detachment. Toward this end, we estimated and tested indirect effects of workplace telepressure through psychological detachment on burnout, poor sleep quality and sleep inconsistency, and work engagement. These well-being outcomes have been shown to be associated with telepressure in previous research (Barber and Santuzzi 2015; Grawitch et al. 2017), and thus serve as the best outcomes for a study to explain why those outcomes were associated with workplace telepressure in past research. Second, we contributed to the continued development of the workplace telepressure construct by examining the stability of telepressure and its relationships to well-being outcomes across three waves of survey data (with one-month time lags). Although our hypotheses are specified at the between-person level of analysis (i.e., testing differences between individuals), the repeated measurements afforded us the opportunity to report the stability of the workplace telepressure construct and examine relationships with well-being outcomes at both the between-person and within-person (variation across measurements) levels of analysis.

Defining Workplace Telepressure

Research has identified a wide variety of sources of ICT-related demands at work that could have a negative impact on employee stress and well-being (Day et al. 2010, 2012). Examples include technological hassles (e.g., computer freezes and breakdowns), expectations to be available or contacted during non-work hours (e.g., on-call positions), information overload, and employee monitoring (e.g., recording keystrokes and performance monitoring). Of particular interest to this study, immediate response expectations for electronic communications (e.g., via email, instant messenger, or phone) are also considered to be a key ICT-related work demand.

However, the mere presence of ICT demands does not have a strong association with worker stress and impairments to employee well-being. Day et al. (2012) found that reports of high response expectations were not predictive of stress and burnout, especially when accounting for other types of ICT demands. This may be because external expectations for immediate responses are not always internalized by workers as important for their jobs. Instead, it is more useful to identify the internal psychological state that keeps workers connected to work-related technology.

Barber and Santuzzi (2015) used the term “workplace telepressure” to describe the preoccupation and urge to respond to message-based ICTs for work purposes. Workplace telepressure captures the internalization of response expectation norms that are anticipated to be more directly linked to heightened worker stress and lower well-being. It has shown to be unique from other work-related constructs related to ICTs and work connection, such as frequency of ICT use at work and home, ICT demands, creation of ICT boundaries at home, workload, and workaholism, and personality traits (e.g., conscientiousness, extraversion, and public self-consciousness; Barber and Santuzzi 2015). Instead, workplace telepressure was associated with perceived norms for response expectations, which may be driven by individual interpretation. A recent examination of the workplace telepressure construct demonstrated significant associations between telepressure and workaholism, work overload, emotional exhaustion, less psychological detachment from work, and lower work-life balance satisfaction (Grawitch et al. 2017). Other validation work with telepressure outside of the workplace (i.e., general telepressure experienced by college students) has found that telepressure is also distinct from the fear of missing out, self-control, and social media engagement (Barber and Santuzzi 2017).

Some key aspects of the workplace telepressure construct are currently unclear. First, although workplace telepressure is conceptualized as a response to work demands, we do not know the extent to which it varies across time and situation. If driven by workplace demands, one might expect reported levels of telepressure to vary month to month, similar to findings from other studies examining the role of work stressors in worker experiences of strain (de Lange et al. 2003; Zapf et al. 1996). Research on general telepressure from work and non-work sources showed some variability in telepressure reports between two time points over a one-month period (Barber and Santuzzi 2015). The present study involved a larger data collection at three time points with a one-month lag to provide an initial exploration of stability in the workplace telepressure construct and its relationship to well-being outcomes. The repeated measurements allowed us to explore possible within-person variability in telepressure that might inform the definition of the construct which, like many organizational constructs, has been limited to only between-person examination (Beal 2015).

Secondly, the reason why workplace telepressure is associated with negative worker well-being has been assumed to be due to interference with work recovery processes, consistent with other research on ICT behaviors and work-home experiences (Barber and Jenkins 2014; Derks and Bakker 2014; Derks et al. 2014; Park et al. 2011). Yet, this explanation has not yet been formally tested and confirmed. The present study examines psychological detachment as a key recovery process that may be negatively affected by workplace telepressure, which should predict more burnout, poor sleep quality, and lower levels of work engagement.

In the following sections, we review past findings for the associations between workplace telepressure and well-being outcomes. We then apply occupational health psychology theories regarding work recovery processes to guide our prediction that workplace telepressure will indirectly relate to those outcomes through lower psychological detachment from work.

Well-Being Consequences of Workplace Telepressure

Workplace telepressure is expected to be problematic in organizations because of its contribution to worker burnout, poor sleep, and (if it interferes with recovery), lower work engagement. According the job demands-resources (JD-R) model (e.g., Bakker and Demerouti 2007), work demands require sustained effort and attention. If workers fail to get a break from these demands, they can experience work exhaustion (i.e., physical, cognitive, and emotional; Shirom and Melamed 2006) or other negative health and well-being outcomes. Workplace telepressure reflects a worker’s psychological response to perceived work demands, namely, demands to respond quickly to work-related ICT messages. Although workers are not necessarily completing effortful work when checking and responding to ICT demands, workplace telepressure can divert employees’ attention to work-related thoughts, which have been shown to increase work exhaustion and lead to poor sleep quality (Querstret and Cropley 2012).

Some empirical support for the negative relationship between workplace telepressure and employee health and well-being currently exists in the literature. Results from initial cross-sectional studies showed that workplace telepressure was associated with higher levels of physical and cognitive exhaustion, but not emotional exhaustion (Barber and Santuzzi 2015). Additionally, these studies found that workplace telepressure was associated with poor sleep quality, but not sleep quantity or consistency. The first goal of the present study is to replicate the association between workplace telepressure and the previously examined negative well-being outcomes: exhaustion and sleep quality. Consistent with past findings, we expect the following associations to emerge.
  • Hypothesis 1: Workplace telepressure predicts a) higher levels of exhaustion and b) poor sleep quality.

A competing argument for workplace telepressure is that the urge to respond quickly may actually encourage more engagement in work tasks, generally viewed as a positive outcome. Work engagement describes the psychological state of feeling vigorous, dedicated, and absorbed in one’s work tasks, which is associated with higher ratings of job performance (Bakker and Bal 2010; Christian et al. 2011).

A positive relationship between workplace telepressure and work engagement at first seems contradictory to past research on ICT demands and workplace telepressure highlighting negative well-being outcomes.

Whether workplace telepressure positively or negatively predicts work engagement may depend on the extent to which telepressure reflects a reaction to response demands evaluated as challenges (promote work mastery or work goal attainment) or hindrances (undermine growth or work goal attainment) by employees (Cavanaugh et al. 2000). Challenges include demands such as workload or time pressure, whereas hindrances include demands such as negative social interactions and bureaucratic red tape. Challenge demands tend to have positive relationships with both work burnout and work engagement, whereas hindrance demands are positively associated with burnout but negatively associated with engagement (Crawford et al. 2010; Van den Broeck et al. 2010). Rising to the challenge of staying connecting and ready to respond to work-related ICT messages should be associated with higher levels of work engagement, but also feelings of exhaustion.

Following this logic, we expect workplace telepressure to have a positive association with work engagement because, on average, it likely qualifies as a reaction to a challenge demand along the lines of workload and time pressure. Employees often report that email is a source of stress because it represents their current workload, with a “clean inbox” signifying work accomplishment (Barley et al. 2011). Consistent with past research on challenge demands and work engagement (Crawford et al. 2010; Van den Broeck et al. 2010), Barber and Santuzzi (2015) found a small positive association between workplace telepressure and work engagement in a cross-sectional survey. Thus, we expected a similar finding in the current study.
  • Hypothesis 2: Workplace telepressure is associated with higher work engagement.

The Intervening Role of Psychological Detachment

In addition to replicating previously detected associations between telepressure and well-being, we also aim to understand why workplace telepressure may be associated with negative well-being outcomes. The negative well-being consequences of workplace telepressure are expected to emerge to the extent that telepressure interferes with employee recovery processes. The rationale is that telepressure is a response to internalized demands to maintain communication and reinforces reminders of what needs to be done at work (Day et al. 2010). Consistent with effort-recovery theory (Meijman and Mulder 1998), this prolonged response to work demands might interfere with important recovery processes such as psychological detachment from work; that is, not thinking about work and work-related events (Sonnentag and Fritz 2007). Workplace telepressure can transform message-based ICT use into “inescapable work” which can decrease psychological detachment from work, thus impairing the recovery process. Indeed, workplace telepressure has been linked with less psychological detachment (Barber and Santuzzi 2015), but the presumed mediating role of detachment on employee well-being has not been examined.

The stressor-detachment model (Sonnentag and Fritz 2015) extends the effort-recovery model by suggesting that psychological detachment is particularly important to positive worker well-being. Psychological detachment is associated with less emotional exhaustion (Sonnentag et al. 2010a, b), less burnout (Sonnentag and Fritz 2007) fewer psychosomatic complaints (Sonnentag et al. 2010a; Sonnentag and Fritz 2007), better sleep (Sonnentag and Fritz 2007), and more life satisfaction (Fritz et al. 2010; Sonnentag and Fritz 2007). Moreover, past research demonstrates that detachment partially mediates the relationship between job stressors and exhaustion, such that lower detachment as a function of job stressors predicts more exhaustion (Sonnentag et al. 2010b). Therefore, less psychological detachment should serve a similar intervening role in the relationship between workplace telepressure and negative well-being outcomes (exhaustion and poor sleep quality).

Less detachment also has been shown to impair an employee’s capacity to sustain work engagement and productivity over the long run (Fritz and Sonnentag 2006; Kühnel et al. 2009; Sonnentag 2003). A recent review noted that, with only one exception, the research to date shows no positive effects on employee productivity and work engagement for employees who experience low psychological detachment (Sonnentag and Fritz 2015). Moreover, Kühnel et al. (2009) found that even when job involvement had a positive association with work engagement, there was a negative indirect effect through lower psychological detachment. In other words, high job involvement might backfire and negatively affect work engagement to the extent that involvement interferes with recovery. This may create a hindrance rather than a challenge experience (Cavanaugh et al. 2000), leading to negative rather than positive effects on work engagement. Similarly, workplace telepressure could negatively predict work engagement to the extent that it is associated with impaired recovery via lower psychological detachment.
  • Hypothesis 3: Workplace telepressure predicts lower levels of psychological detachment from work.

  • Hypothesis 4a: Workplace telepressure will indirectly relate to negative worker outcomes (exhaustion and poor sleep quality) through lower psychological detachment.

  • Hypothesis 4b: Workplace telepressure will indirectly and relate to lower work engagement through lower levels of psychological detachment (despite a positive bivariate relationship between telepressure and engagement).

To date, the construct of workplace telepressure has not been examined for stability across time. Thus, we conducted preliminary analyses to explore the stability of workplace telepressure and proposed well-being outcomes at one-month and two-month time lags. All hypotheses were proposed and tested at the between-person level of analysis. However, the findings for stability informed whether within-person variation in workplace telepressure also should be accounted for in the analyses.

Method

Participants

We recruited participants from Amazon’s Mechanical Turk (MTurk), which is an online crowd-sourcing platform that is becoming more widely adopted organizational researchers due to the quality and diversity of participants (see Landers and Behrend 2015) and quality of data with the use of appropriate attention checks (Goodman et al. 2013). This data source was chosen to access employees from a wide variety of organizations and occupations. To select employees most relevant to the study, we used an initial prescreening survey ($0.10 payment) that asked about demographic information (including employment and frequency of ICT use at work), various work demands (general, ICT-related, and task interdependence), and interest in completing the longitudinal portion of the study by providing an email address for follow-up surveys. The prescreening approach helped to reduce the likelihood of unemployed respondents completing the study and provided temporal separation between measures of ICT demands and other measures in the study to reduce the impact of common method variance. Out of the 425 people who completed the prescreening survey, 355 (83.5%) were employed and provided an email address for further contact.

A final sample of 252 employed adults completed at least two waves of data collection and passed at least two of three data quality check items (i.e., “When you get to this question, please select ‘strongly agree’”). Given that previous research uses multiple items to reliably detect careless responding (e.g., Huang et al. 2015; Meade and Craig 2012), missing only one attention check item was deemed to be insufficient for exclusion because it could be a result of random error. Data were collected in three repeated surveys across 3 months (1 month lags). Participants received $2 for completing Time 1, an additional $5 for completing Time 2, and an additional $10 for completing Time 3. Forty-four percent reported using ICTs in their jobs every day, with the remainder reporting ICT use as often (17.5%), sometimes (17.5%), rarely (14%), or never (7%). After excluding participants who do not use ICT tools for work purposes, the final sample included 234 adults (45% male; 55% female) with varying levels of reported frequency of ICT use (M = 3.99, SD = 1.13). The average age of the sample was 32.83 years (SD = 9.58) and the majority reported being White (80%) with others reporting African American (9.4%), Hispanic (6.4%), Asian (6.8%), and Other (1.3%). One person chose to not provide ethnicity information. Only 13% of participants identified as virtual workers or telecommuters, 19% reported part-time (< 35 h per week), and 36.8% held managerial or supervisory roles. On average, participants held the job for 4.76 years (SD = 4.79). The sample showed a broad representation of occupations, including management, business and financial operations, computer and mathematical, architecture and engineering, life and related sciences, community and social service, legal, healthcare, arts and sports, protective services, food services, personal care, sales, office and administrative, farming, construction, maintenance and repair, production, and transportation occupations.

Missing data in the final sample occurred to a small degree at Time 2 (5%) and Time 3 (8%). Missing data were not significantly related to any of the variables in the dataset. Given the small amount of data presumed to be missing completely at random and adequate statistical power, missing cases were not imputed. The small amount of missing data was not expected to bias results from the maximum likelihood estimation in the analyses (Peugh and Enders 2004).

Measures and Procedure

During the prescreening survey, participants reported on the frequency of ICT use (described above) and ICT connection demands. Participants then completed previously validated measures of workplace telepressure, psychological detachment, and psychological work and sleep impairments at three time points with a one-month lag. Each of the following measures used for this study’s analyses was included in a larger collection of measures in the survey.

Workplace Telepressure

Participants reported their general thought about how they use technology to communicate with people in the workplace on a validated a six-item measure of workplace telepressure (Barber and Santuzzi 2015). Example items are “I can’t stop thinking about a message until I’ve responded” and “I feel a strong need to respond to others immediately.” Each item was accompanied by a Likert response scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree). Reliability was .92, .89, and .94 at each of the three time periods, respectively.

Psychological Detachment from Work

Psychological detachment was measured using the four detachment items from the Recovery Experience Questionnaire (Sonnentag and Fritz 2007). Participants rated their agreement with four statements about their non-work time during the past month (e.g., “I don’t think about work at all”). Response options ranged from 1 (Strongly Disagree) to 5 (Strongly Agree). Reliability estimates were .84, .86, and .84 for the three time points, respectively.

Work Exhaustion

Psychological work impairment during the past month was assessed using the Shirom-Melamed Burnout Measure (Shirom and Melamed 2006), The measure includes 14 items divided among three subscales for physical fatigue (e.g., “I feel tired”), cognitive weariness (e.g., “My thinking process is slow”), and emotional exhaustion (e.g., “I feel I’m unable to be sensitive to the needs of coworkers or customers”). The response scale for each item ranged from 1 (Never or Almost Never) to 7 (Always or Almost Always). Reliability was high for each subscale at each time period. The reliabilities for physical fatigue at each respective time point were.95, .95, and .94. For cognitive weariness, the reliabilities were .94, .96, and .95. Emotional exhaustion showed reliabilities of .93, .96, and .95, respectively.

Work Engagement

Work engagement during the past month was measured with nine items representing three subscales: vigor, dedication, and absorption (Schaufeli et al. 2006). All items were accompanied by a seven-point response scale ranging from 0 (Never) to 6 (Always). Each subscale comprised three items and was examined as separate outcome variables in this study. Vigor had reliabilities of .87, .88, and .89 at each time period, respectively. Dedication showed similar reliabilities of .87, .85, and .88. Absorption also showed acceptable reliabilities of .79, .77, and .79.

Sleep Quality and Consistency

A measure of chronic insomnia (Jenkins et al. 1988) was used to assess poor sleep quality during the past month. Participants reported their experiences with four symptoms over the past month (viz., trouble falling asleep, trouble staying asleep, waking up several times during the night, and waking up after one’s usual amount of sleep feeling tired and worn out). Response options ranged from 0 (Never) to 5 (Nearly every night), with higher scores indicating lower sleep quality. Reliabilities were .81, .82, and .80 for each time period, respectively.

Sleep inconsistency during the past month was measured with three items from the Sleep Hygiene Index (Mastin et al. 2006), referring to going to bed at different times, getting out of bed at different times, and staying in bed longer than one should (i.e., sleeping in) day to day. Response options ranged from 1 (never) to 5 (always); higher scores represented more sleep inconsistency. Reliabilities were .79, .74, and .72 for the three time periods.

ICT Connection Demands (Prescreening Survey Only)

In addition to frequency of ICT use at work, we also measured ICT-related demands (α = .83) that were specifically related to staying connected. ICT connection demands were measured with two items for response expectations (e.g., “I am expected to respond to email messages immediately”), four items for availability (e.g., “I’m expected to check e-mail and/or voicemail when I’m out of the office”), and three items for workload (e.g., “Technology creates more work for me”) subscales from Day et al. (2012). Response options ranged from 1 (Never) to 5 (Always). Participants showed variability in reported ICT demands (M = 3.45, SD = 0.78), with scores ranging between 1.67 and 5.00.

Analysis

One goal of this study was to empirically identify the extent to which workplace telepressure changes from month to month, which could inform expectations for within-person relationships with outcome variables. We first conducted preliminary cross-lagged panel analyses (CLPA; Finkel 1995; Kenny 1975) using Mplus v.8.3 to test the stability of telepressure and well-being outcomes after a one-month and two-month lag, as well as identify any evidence for directional relationships between the variables above and beyond the stability effects. These models specified the stability of each variable, concurrent relationships (stationary), cross-lagged effects, and reciprocal effects between variables. For example, the relationship between telepressure and psychological detachment across a one-month lag would involve regressing the Time 2 measurement of each variable on its own Time 1 measurement (stability), correlating telepressure and detachment within each time period (stationary), regressing psychological detachment at Time 2 on telepressure at Time 1 (cross-lagged effect), and regressing telepressure at Time 2 on psychological detachment at Time 1 (reciprocal effect). All effects were tested simultaneously in path analysis models for each outcome variable as predicted by workplace telepressure.

In addition, the intraclass correlation (ICC) for workplace telepressure was computed to identify the extent to which variability in telepressure was likely driven by stable individual differences, as compared to within-person variability across the three time periods. Finally, repeated measures analysis of variance (ANOVA) tested whether there were significant differences in workplace telepressure across the three time periods, providing additional information about the likelihood of meaningful within-person change.

As described below, the preliminary stability analyses suggested that workplace telepressure scores might have included variability at the individual level as well as within-person across time. Thus, hypotheses were tested with two separate components of telepressure: a between-person component (individual means) and a within-person component (within individual mean-deviation scores). See Table 1 for all variable means, standard deviations, bivariate correlations, and intraclass correlations.
Table 1

Descriptive statistics and correlations among study variables

 

Variable

1

2

3

4

5

6

7

8

9

10

1

Workplace telepressure

(0.92)

−0.11*

0.17*

0.09*

0.17*

0.11*

0.05

0.04

0.04

0.15*

2

Psychological detachment

−0.20*

(0.85)

−0.28*

−0.07

−0.20*

−0.18*

−0.10*

0.14*

0.09*

0.03

3

Poor sleep quality

0.23*

−0.33*

(0.81)

0.42*

0.57*

0.45*

0.36*

−0.27*

−0.17*

−0.12*

4

Sleep inconsistency

0.12

−0.10

0.47*

(0.75)

0.50*

0.45*

0.30*

−0.36*

−0.28*

−0.23*

5

Physical fatigue

0.21*

−0.24*

0.62*

0.58*

(0.95)

0.72*

0.61*

−0.62*

−0.51*

−0.36*

6

Cognitive weariness

0.14*

−0.19*

0.48*

0.51*

0.76*

(0.95)

0.59*

−0.49*

−0.44*

−0.35*

7

Emotional exhaustion

0.09

−0.13*

0.43*

0.36*

0.68*

0.66*

(0.95)

−0.42*

−0.41*

−0.32*

8

Vigor

0.02

0.10

−0.32*

−0.44*

−0.71*

−0.59*

−0.50*

(0.88)

0.82*

0.73*

9

Dedication

0.03

0.06

−0.22*

−0.37*

−0.61*

−0.54*

−0.49*

0.87*

(0.87)

0.76*

10

Absorption

0.15*

−0.03

−0.16*

−0.29*

−0.46*

−0.44*

−0.37*

0.76*

0.82*

(0.78)

 

M

3.47

3.33

2.45

2.74

3.78

3.18

2.92

3.50

3.90

3.93

 

SD

0.81

0.79

1.52

0.78

1.37

1.18

1.24

1.00

1.00

0.86

 

ICC

0.53

0.45

0.72

0.60

0/74

0/68

0/59

0.69

0.67

0.62

Means and standard deviations are based on averaged individual level scores for all variables (N = 234). Average reliability for each variable appears on the main diagonal. The lower triangle provides individual level correlations. The upper triangle displays correlations based on data at the measurement period level

*p < .05

To account for the clustered nature of the data (time points nested within individual), we used multilevel structural equation models (MSEM; Preacher et al. 2010, 2015) with maximum likelihood estimation (Mplus v.8.3). All models to test hypotheses specified relationships at the between-person and within-person levels (although hypotheses are stated only for between-person relationships). The two components for workplace telepressure were included as predictors in a bivariate model predicting burnout (latent composite of three subscales), poor sleep quality, sleep inconsistency, and work engagement (latent composite of three subscales). In subsequent models, psychological detachment was added as an intervening variable to examine its relationships with outcomes with workplace telepressure accounted for in the model. In the latter models, indirect effects of workplace telepressure to all outcomes through psychological detachment were computed and tested for statistical significance.

As both ICT behaviors and expectations have been implicated in the literature as predictors of well-being impairments, ICT frequency of use and ICT connection demands were tested as potential alternative explanations for telepressure in exploratory analyses. These variables were tested as predictors of well-being outcomes and as potential moderators of the relationships between workplace telepressure and well-being outcomes. Workplace telepressure (between-person), ICT frequency of use, and ICT connection demands were grand mean-centered at the individual level of analysis prior to computing interaction terms involving those variables. Significant interactions were probed for simple effects according to current practices in multilevel analysis for psychological research (Bauer and Curran 2005).

Results

Stability of Workplace Telepressure and Well-Being

The CLPA analyses showed strong stability effects for telepressure as well as outcome variables over the three-month period (see Table 2). After controlling for workplace telepressure and detachment scores across time, there was some evidence for lagged effects with psychological detachment 1 and 2 months later. Specifically, more workplace telepressure predicted lower psychological detachment 1 and 2 months later. A reciprocal effect emerged at 2 months, but not 1 month later. This suggests that lower psychological detachment at Time 1 predicted higher reported levels of workplace telepressure 2 months later (but not 1 month later).
Table 2

Preliminary cross-lagged models for each worker well-being outcome at a) 1 month and b) 2 month lags

One-month lag

Outcome

Telepressure stability

Outcome stability

Stationary (T1)

Stationary (T2)

Telepressure predicting outcome

Reciprocal effect

 Psychological detachment

0.41*

0.41*

−0.05

−0.11*

−0.14*

−0.04

 Physical fatigue

0.41*

0.75*

0.33*

0.08

0.004

0.004

 Cognitive weariness

0.42*

0.73*

0.19*

0.09

−0.01

−0.004

 Emotional exhaustion

0.42*

0.62*

0.14

−0.03

0.07

−0.01

 Poor sleep quality

0.40*

0.75*

0.24*

0.10

0.07

0.05

 Sleep inconsistency

0.41*

0.54*

0.11

0.02

0.02

0.04

 Vigor

0.41*

0.73*

0.09

0.04

−0.11*

0.02

 Dedication

0.42*

0.72*

0.08

0.07

−0.07

0.001

 Absorption

0.42*

0.65*

0.22*

0.07

−0.08

0.002

Two-month lag

 

Telepressure stability

Outcome stability

Stationary (T1)

Stationary (T3)

Telepressure predicting outcome

Reciprocal effect

 Psychological detachment

0.49*

0.35*

−0.05

0.04

−0.14*

−0.14*

 Physical fatigue

0.49*

0.68*

0.33*

0.05

0.04

0.03

 Cognitive weariness

0.49*

0.69*

0.19*

−0.02

0.01

0.07

 Emotional exhaustion

0.50*

0.57*

0.14

0.01

0.03

−0.004

 Poor sleep quality

0.49*

0.64*

0.24

0.03

0.16

0.05

 Sleep inconsistency

0.50*

0.47*

0.11

0.02

−0.01

0.05

 Vigor

0.50*

0.70*

0.09

−0.01

−0.07

−0.01

 Dedication

0.50*

0.61*

0.08

−0.02

−0.04

−0.02

 Absorption

0.50*

0.62*

0.22*

0.04

0.01

0.002

Unstandardized estimates from the full cross-lagged models for each outcome are presented

*p < .05

After controlling for workplace telepressure and well-being outcomes across time points, most of the lagged relationships between telepressure and well-being 1 month later were not significant. The only exception was a significant negative relationship between workplace telepressure and vigor (work engagement), suggesting that telepressure at the first measurement predicted less vigor 1 month later. Also, none of the reciprocal relationships (i.e., where well-being outcomes were tested as predictors of workplace telepressure) were significant. Overall, the CLPA analyses provided little evidence of lagged relationships between workplace telepressure and well-being outcomes over a two-month period after the stability of variables was accounted for in the analyses.

However, further diagnostics suggest some potential variability in workplace telepressure across the time periods that should be considered. Workplace telepressure showed a significant and large ICC, indicating a substantial amount of individual-level variance (ICC = .53, p < .001), but almost half of the variance remained as within-person variability. Moreover, a repeated-measures ANOVA showed workplace telepressure scores to differ significantly (although not linearly) across the three time points, F(2, 404) = 6.56, p = .002. To account for both the between-person and within-person variance in telepressure, hypotheses were tested with both components of workplace telepressure as separate predictor variables to test whether both sources of variability can predict well-being outcomes.

Hypothesis Testing: Bivariate Relationships

See Table 3 for fixed and random estimates for all hypothesis tests. As burnout was conceptualized and measured with three subscales, we first examined burnout as a latent construct with three indicators as an outcome of workplace telepressure (χ24 = 7.38, p = .11; CFI = .99; RMSEA = .035). Consistent with past research, between-person levels of workplace telepressure significantly predicted more burnout (b = 0.28, SE = .12, p = .02). The effect was not significant at the within-person level (b = 0.06, SE = .06, p = .25). Decomposing the burnout into specific subscales provided evidence that the effects suggested that the between-person effect was driven primarily by a strong and significant relationship between workplace telepressure and physical exhaustion (b = 0.35, SE = .11, p = .022). When within-person effects were controlled, neither cognitive weariness (b = 0.20, SE = .10, p = .05) nor emotional exhaustion (b = 0.14, SE = .10, p = .19) showed significant between-person relationships with telepressure. None of the relationships between workplace telepressure and burnout was significant at the within-person level of analysis (ps > .16). It is worth noting, however, that the relationship between workplace telepressure and cognitive weariness reached statistical significance when the non-significant within-person effect of telepressure was excluded from the model specification (while still estimating within-person variance; b = 0.28, SE = .12, p = .02). Together, the results support Hypothesis 1a such that workplace telepressure predicts more physical exhaustion and cognitive weariness at the between-person level of analysis.
Table 3

Fixed and random estimates for hypothesis tests using multilevel structural equation modeling

 

Null model

Bivariate relationships

Indirect effect model

Random

Fixed

Random

Fixed

Random

TP

TP

PD

Outcome

Between variance

Within variance

Between effect

Within effect

Between variance

Within variance

Between effect

Within effect

Between effect

Within effect

Between variance

Within variance

Psychological detachment

0.44*

0.53*

−0.19*

0.06

0.42*

0.52*

Burnout (latent)

1.33*

0.31*

0.28*

0.08

1.30*

0.31*

0.20

0.07

−0.43*

−0.09

1.23*

0.30*

Physical fatigue

1.65*

0.57*

0.35*

0.08

1.57*

0.57*

0.25*

0.08

0.50*

−0.08

1.47*

0.57*

Cognitive weariness

1.19*

0.57*

0.20

0.08

1.17*

0.57*

0.14

0.09

−0.33*

−0.13

1.12*

0.56*

Emotional exhaustion

1.24*

0.86*

0.14

−0.03

1.22*

0.86*

0.08

−0.04

−0.30*

0.02

1.18*

0.86*

Work Engagement (latent)

0.72*

0.24*

0.04

0.06

0.72*

0.24*

0.06

0.06

0.07

0.15*

0.71*

0.23*

Vigor

0.85*

0.39*

0.03

0.05

0.85*

0.39*

0.06

0.04

0.15

0.17*

0.84*

0.37*

Dedication

0.83*

0.41*

0.03

0.03

0.83*

0.41*

0.05

0.03

0.07

0.13*

0.83*

0.41*

Absorption

0.61*

0.38*

0.16*

0.11*

0.59*

0.37*

0.15

0.10*

−0.06

0.15*

0.59*

0.36*

Poor sleep quality

2.02*

0.77*

0.42*

0.002

1.99*

0.77*

0.27*

0.01

−0.78*

−0.16

1.65*

0.76*

Sleep inconsistency

0.49*

0.32*

0.11

0.04

0.48*

0.32*

0.09

0.04

−0.12

0.01

0.48*

0.32*

All estimates are unstandardized. Between effects involve only between-person components of each variable in the model. Within effects include only within-person components of each variable in the model

*p < .05

Similarly, results supported Hypothesis 1b at the between-person, but not within-person, level of analysis. Workplace telepressure at the between-person level predicted poor sleep quality (b = 0.42, SE = .12, p < .001) but not sleep inconsistency (b = 0.11, SE = .07, p = .10). Neither sleep outcome was significantly related to workplace telepressure at the within-person level (ps > .56).

The between-person levels of workplace telepressure did not predict the latent composite for the three subscales of work engagement (b = 0.04, SE = .08, p = .60; overall model χ2(4) = 15.49, p = .004; CFI = .99; RMSEA = .065). The effect also was not significant at the within-person level (b = 0.06, SE = .05, p = .21). However, decomposing work engagement into specific subscales provided evidence of a positive between-person effect (b = 0.16, SE = .08, p = .04) and within-person effect (b = 0.11, SE = .05, p = .03) uniquely for the absorption subscale of work engagement. These relationships suggest that workplace telepressure showed a positive relationship with absorption at both the between-person and within-person level of analysis. Workplace telepressure was not related to the other subscales at the between-person or within-person levels (ps > .36). The results lend partial support to Hypothesis 2, specifically suggesting that more workplace telepressure is associated with more absorption (but not vigor or dedication) measures of work engagement at both the between-person and within-person level of analysis. Thus, Hypothesis 2 received partial support.

Workplace telepressure means significantly predicted lower levels of psychological detachment at the between-person level (b = −0.19, SE = .06, p = .002). Supporting Hypothesis 3, higher mean levels of workplace telepressure predicted less psychological detachment from work. At the within-person level of analysis, workplace telepressure did not predict psychological detachment in this study (b = 0.06, SE = .08, p = .43).

Hypothesis Testing: Indirect Effects

Hypotheses 4a was tested with indirect effects of workplace telepressure on exhaustion and poor sleep quality through psychological detachment. The predictor (workplace telepressure) and proposed intervening variable (psychological detachment) were significantly related only at the between-person level for these outcomes. Thus, the indirect effects were estimated with the relationship between the predictor and intervening variable (a path) at the between-person level of analysis. See Figs. 1 and 2 for model specification and Table 3 for all model estimates.
Fig. 1

Multilevel direct and indirect effects of workplace telepressure on burnout through psychological detachment

Fig. 2

Multilevel direct and indirect effects of workplace telepressure on sleep through psychological detachment

As expected, workplace telepressure exhibited a significant and positive indirect effect on burnout as a latent variable composite through psychological detachment at the between-person level (ab = .09, SE = .04, p = .03; overall model χ2(8) = 11.22, p = .19; CFI = .99; RMSEA = .02). None of the required relationships among variables was significant at the within-person level; thus, indirect effects at the within-person level were not computed. At the between-person level, the indirect effects were positive and significant for the physical exhaustion subscale (ab = 0.10, SE = .05, p = .04). The effect on cognitive weariness (ab = 0.06, SE = .03, p = .05) and emotional exhaustion (ab = 0.06, SE = .03, p = .09) were not significant (See Fig. 1).

The indirect effect of workplace telepressure through psychological detachment on poor sleep quality also was statistically significant (ab = 0.15, SE = .06, p = .01). The indirect effect was not significant for sleep inconsistency (ab = 0.02, SE = .02, p = .23; see Fig. 2). The results for burnout and sleep quality outcomes suggest that the impact of workplace telepressure on negative well-being outcomes observed in bivariate relationships above may be in part driven by poor work recovery as measured by psychological detachment.

To test Hypothesis 4b, indirect effects were also estimated for the effects of workplace telepressure on work engagement. In previous analyses, workplace telepressure showed relationships with absorption at both the between-person and within-person levels of analysis. However, psychological detachment was only related to telepressure at the between-person level, and detachment only related to work engagement (absorption) at the within-person level (see Fig. 3). To test the indirect effect where the first path (a) is at the between-person level and the second path (b) is at the within-person level, the MSEM analysis was adjusted to conduct a traditional multilevel model. Specifically, the b path was constrained to be equal across the between-person and within-person levels of analysis (see Fig. 4).
Fig. 3

Multilevel direct and indirect effects of workplace telepressure on work engagement through psychological detachment (unconstrained MSEM)

Fig. 4

Multilevel direct and indirect effects of workplace telepressure on work engagement through psychological detachment (constrained to traditional multilevel model)

When examined as a latent composite variable, work engagement was indirectly predicted by workplace telepressure through psychological detachment at the within-person level (ab = −0.03, SE = .021 p = .03; overall model χ2(9) = 23.76, p = .005; CFI = .98; RMSEA = .049). We decomposed the latent composite into separate subscales for work engagement.

Note that the bivariate relationship between workplace telepressure and work engagement (at the between-person and within-person levels) was positive and significant only for the absorption subscale of work engagement.1 The indirect effects of between-person telepressure through psychological detachment was negative and significant for vigor (ab = −0.03, SE = .01, p = .02) and dedication (ab = −0.02, SE = .01, p = .048). The indirect effect was not significant for absorption (ab = −0.02, SE = .01, p = .08). However, when the non-significant within-person component of telepressure was removed from the model, all three indirect effects were statistically significant (ps < .01). Also note that the indirect effects are opposite in sign to the bivariate relationship between workplace telepressure and absorption. This suggests that only the part of workplace telepressure contributing to psychological detachment might have a negative impact on work engagement. Alternatively, only the part of workplace telepressure that does not interfere with psychological detachment yielded a positive relationship with absorption.

Alternative Explanations

Frequency for ICT use at work and ICT connection demands were examined as potential alternative explanations in the hypothesis testing models. Frequency of ICT use was positively related to workplace telepressure (b = 0.14, SE = .05, p = .003), psychological detachment (b = −0.14, SE = .04, p = .001), physical exhaustion (b = −0.16, SE = .08, p = .046), vigor (b = 0.15, SE = .06, p = .01), dedication (b = 0.13, SE = .06, p = .03), and absorption (b = 0.11, SE = .05, p = .02). However, no interactions between workplace telepressure and frequency of ICT use emerged. Importantly, the significant relations between telepressure and detachment (b = −0.16, SE = .06, p = .01), physical exhaustion (b = 0.40, SE = .11, p < .001), cognitive weariness (b = 0.22, SE = .11, p = .04), poor sleep quality (b = 0.43, SE = .12, p < .001), and absorption (within-person level only; b = 0.11, SE = .05, p = .03) still held after controlling for ICT use. Also, the previously noted indirect effects through psychological detachment on physical exhaustion (ab = .10, SE = .04, p = .03), cognitive weariness (without within-person relationships in model; ab = .07, SE = .03, p = .04), and poor sleep quality (ab = .14, SE = .06, p = .02) remained significant. However, the indirect effect on telepressure on absorption was no longer statistically significant when controlling for frequency of ICT (ab = −.02, SE = .01, p = .07).

ICT connection demands also were related to workplace telepressure (b = 0.31, SE = .07, p < .001), psychological detachment (b = −0.25, SE = .06, p < .001) and absorption (b = 0.18, SE = .08, p = .02). Similar to workplace telepressure, ICT connection demands was associated with less psychological detachment from work and more work engagement as absorption. When ICT connection demands was included with telepressure in predicting outcomes, the previous relationship between telepressure and psychological detachment (b = −0.14, SE = .06, p = .03) held as significant; however, the relationship between workplace telepressure and absorption at the between-person level of analysis dropped to a non-significant level (b = 0.12, SE = .09, p = .15). As the ICT connection demands variable was measured only at the between-person level, the within-person association of workplace telepressure on absorption was not affected when ICT connection demands was included in the model. The findings suggest that ICT connection demands could serve as an alternative explanation for the relationship between workplace telepressure and absorption at the between-person level.

Although there were no linear bivariate relationships between ICT connection demands and the exhaustion outcomes, analyses revealed significant interactions between workplace telepressure (between-person) and ICT connection demands for all three exhaustion variables: physical (b = −0.28, SE = .14, p = .04), cognitive (b = −0.28, SE = .14, p = .04), and emotional (b = −0.44, SE = .14, p = .001).2 At high levels of ICT connection demands (one standard deviation above the mean), workplace telepressure did not predict physical (b = 0.14, SE = .17, p = .40), cognitive (b = −0.01, SE = .16, p = .93), or emotional exhaustion (b = −0.23, SE = .15, p = .15). However, at low levels of ICT connection demands (one standard deviation below the mean), workplace telepressure predicted more physical (b = 0.58, SE = .14, p < .001), cognitive (b = 0.43, SE = .16, p = .01), and emotional exhaustion (b = 0.45, SE = .15, p = .002). In all three outcomes, the interaction pattern suggests that the impact of workplace telepressure that typically occurs on exhaustion (i. e., at average levels of ICT connection demands) decreases as ICT connection demands increase, and increases as ICT connection demands decrease.

Discussion

Summary of Findings

The present study aimed to directly test the presumed intervening mechanism of impaired recovery via psychological detachment as an explanation for why workplace telepressure is associated with negative worker well-being.

As an initial step, we examined the extent to which workplace telepressure was stable across three measurements over a two-month time period. We demonstrated that workplace telepressure might change to some extent over time. We observed a large ICC, indicting significant consistency in workplace telepressure reports across the three-month period. Yet, approximately half of the variance remained as within-person variation. A preliminary cross-lagged panel analysis revealed that the stability of constructs across the three waves of data precluded evidence for lagged relationships among the telepressure, detachment, and well-being outcome variables. Thus, data were analyzed using a multilevel structural equation modeling approach to account for possible between-person and within-person effects of telepressure on outcomes. However, this decision limited our ability to draw strong conclusions about the relationship directionality.

Similar to past research (Barber and Santuzzi 2015), higher levels of workplace telepressure were associated with more physical and cognitive exhaustion and sleep quality problems at the between-person level of analysis. The within-person variation did not predict exhaustion or sleep outcomes. Also consistent with past findings on challenge demands (Crawford et al. 2010; Van den Broeck et al. 2010), workplace telepressure showed a small positive association with work engagement, namely the absorption subscale. The relationship was evident at both the between-person and within-person levels of analysis. These results support the assertion that workplace telepressure may be a reaction to a challenge demand (as compared to hindrance) that is associated with increased levels of both work engagement and work exhaustion.

Furthermore, results supported psychological detachment as an intervening variable that may partially explain why workplace telepressure predicts negative worker outcomes. At the between-person level, workplace telepressure led to lower psychological detachment, which in turn predicted more physical exhaustion and sleep quality problems. These findings are consistent with past research showing the negative relationship between poor detachment from workplace stressors and well-being, in accordance with the stressor-detachment model (Sonnentag and Fritz 2007, 2015).

In contrast to the exhaustion and sleep quality results, indirect effect of workplace telepressure on work engagement through psychological detachment was significant at the within-person (but not between-person) level. Although only the absorption subscale for work engagement showed a bivariate relationship with workplace telepressure (at the between-person and within-person levels), the indirect effects were negative and significant for the within-person factors of work engagement (absorption, vigor and dedication) after removing the non-significant within-person effect of telepressure on those factors. Note that the significant bivariate association between telepressure and absorption was positive, whereas the indirect effects through psychological detachment on work engagement were negative. This suggests that the between-person variance in workplace telepressure that was unrelated to psychological detachment demonstrated a positive association with the absorption subscale of work engagement, whereas the part of workplace telepressure that is related to psychological detachment demonstrated a negative effect on work engagement. Although seemingly contradictory, a similar pattern was reported in past research on job involvement and work engagement (Kühnel et al. 2009). In that work, job involvement yielded a positive relationship with engagement, but showed a negative indirect effect through lower detachment. More work involvement may be associated with more engagement as long as it does not interfere with recovery, in which case it may backfire and impair work engagement. We observed a similar phenomenon with workplace telepressure. To the extent that telepressure related to lower detachment, the association between telepressure and work engagement was negative.

The fact that the indirect effect to work engagement variables only emerged at the within-person level of analysis suggests the negative effect of workplace telepressure through psychological detachment might be specific to the time period at which work engagement was measured. Whereas general levels of workplace telepressure related to general levels of exhaustion and sleep quality averaged over time, negative effects on work engagement seem to be reflected in time-specific experiences of work engagement. This has important implications for intervention design when addressing either well-being or engagement outcomes of workplace telepressure. Interventions aimed to protect work engagement consequences of telepressure may require a focus on employee inconsistency in engagement rather than general levels of work engagement.

Finally, we also examined whether frequency of ICT use or ICT connection demands at work could serve as alternative explanations for the results hypothesized to be driven by workplace telepressure. Both ICT use and demands were directly related to more workplace telepressure and less psychological detachment. ICT use and ICT connection demands were positively related to some of the engagement subscales, but were not significantly related to exhaustion variables. These findings are similar to those in related research showing no direct effect of ICT demands on well-being (Day et al. 2012).

ICT connection demands—but not ICT use—moderated the effects of workplace telepressure on worker exhaustion, such that telepressure had a greater positive relationship with all three exhaustion subscales (physical, cognitive, and emotional) when ICT demands were low. Although seemingly counterintuitive, these results suggest workplace telepressure is more problematic for workers who have low, rather than high, overall ICT demands. One explanation may be that employees with low ICT demands may be especially sensitive to the occasional request for immediate response due to expectancy violations (Burgoon and Hale 1988). By comparison, employees who have jobs where they regularly face high expectations for response time and availability may also learn to adopt effective coping strategies to reduce exhaustion (i.e., similar to stress inoculation training; Freedy and Hobfoll 1994). Our sample varied greatly in the reported ICT use and ICT demands at work, with scores near the lowest and highest end of the response scales. However, information on the extent to which such use or demands change over time was not collected. The role of expectancy violations as a route to exhaustion as a function of changes in ICT-related expectations and workplace telepressure should be tested in future research studies. Such research might identify whether new ICT users and other workers who are not accustomed to ICT demands are particularly vulnerable to negative well-being outcomes from workplace telepressure.

Implications

Together, the hypothesis tests and exploratory results have several important implications for continued research and organizational practices. Broadly speaking, the findings from this research suggest that employees need effective recovery strategies from demands to respond quickly to work-related ICT messages. The failure to recover from work might yield unintended negative consequences for employee exhaustion, sleep, and work engagement. Our study demonstrated that workplace telepressure is associated with negative well-being outcomes through its negative effects on psychological detachment from work demands. To the extent that workplace telepressure interferes with psychological detachment from work, the expected employee outcomes include more physical exhaustion, poor sleep quality, and less work engagement. These results suggest that a critical component to intervention is ensuring that employees experience psychological detachment from work activities to reduce the negative association between telepressure and well-being.

For example, one training program that has shown to improve sleep quality and reduce stress explicitly teaches employees how to psychologically detach after work, among other after-work recovery strategies (Hahn et al. 2011). Another intervention helps workers detach from high workloads by creating plans for how unfinished tasks will be done the following day (Smit and Barber 2016). Further research on workplace telepressure may want to adopt these types of “detachment” interventions using experimental designs.

Another promising strategy to support worker well-being may be to encourage psychological detachment as a work design strategy that reduces overall ICT response expectation and availability demands. For example, predictable time off (PTO) is one such work reorganization strategy that has been shown to allow people time to disconnect fully from the work environment, even in highly responsive service industries (Perlow and Porter 2009). In PTO environments, employees work as a team to discuss how to create and enforce times when particular employees will be unavailable during a given week or month. This approach is likely to increase workers’ regular experience of psychological detachment by making them feel they are not “on call” at all times.

The predicted paradox of work engagement and telepressure was also evident in the present study. Workplace telepressure was indeed related to higher levels of work engagement (absorption), consistent with a motivational process. It is reasonable to predict that workers who experience more preoccupation and urge to respond to work messages are also likely to report feeling more engaged in work due to repeated communications with work colleagues and perhaps feelings of accomplishment. However, if the cost of workplace telepressure is lower recovery, work engagement might actually suffer. This was demonstrated in our findings regarding how workplace telepressure, psychological detachment, and work engagement variables were related. Although workplace telepressure might be associated with more absorption, the extent to which telepressure reduces psychological detachment from work can yield the opposite outcome in the form of lower work engagement. Practitioners should carefully implement ICT tools and practices that benefit worker engagement without risking a cost to psychological detachment and recovery from work. If ICT practices decrease psychological detachment from work, workers may experience a net cost to their well-being. Given that the negative effects seem to emerge at the within-person level, the extent to which telepressure negatively affects work engagement may vary depending on the time point of measurement. The competitive organization must be able to maximize the positive aspects of ICT tools (i.e., flexibility) while also reducing negative well-being effects on workers due to workplace telepressure (i.e., lack of detachment).

Limitations and Future Directions

Despite the contributions that the three-wave survey provided, several limitations should be noted. As the first multiple-measurement study on workplace telepressure, the appropriate time lags between measurement periods were somewhat uninformed. Past work on work demands tends to find more statistically significant relationships among shorter time lags (1–12 months; de Lange et al. 2003; Sonnentag and Frese 2003; Zapf et al. 1996). Thus, we erred on the side of testing 1 month lags as a starting point for this research given the stability of workplace telepressure effects was unknown, yet recent evidence showed a modest test-retest correlation for general reports of telepressure across a one-month lag (Barber and Santuzzi 2015).

The stability of workplace telepressure at three measurement periods across 2 months yielded mixed conclusions about the appropriateness of the one-month lag. The repeated measures ANOVA suggested that some differences occurred in reported workplace telepressure within participants over the time periods. Despite those results, workplace telepressure showed strong stability effects over the time periods in the CLPA results. Additionally, the tests for hypotheses showed no within-person effects of workplace telepressure on any of the outcome variables. This suggests that, if there is within-person variation on workplace telepressure, it did not predict any meaningful outcomes in the present study. Thus, one possibility is that workplace telepressure is a stable individual difference (see Grawitch et al. 2017), but its measure was susceptible to systematic measurement error in this study.

Another possibility is that the two-month study period was not long enough to detect significant changes in workplace telepressure and its relationship to well-being outcomes. Past research using longitudinal designs to examine workplace stressors and well-being outcomes reports one-month intervals as relatively short, compared to studies examining effects over a year or longer time interval (de Lange et al. 2003; Zapf et al. 1996). Both cross-sectional and lagged relationships may strengthen over a period of several years with chronic exposure to a workplace stressor (Ford et al. 2014). Extending the study period to span several years may allow for more reliable variation in workplace telepressure and the potential to observe accumulation effects on well-being outcomes.

Still another reason is that measurement periods within the study period were not well aligned to systematic changes in workplace telepressure. Measurement periods in future research should coincide with discrete changes in specific employees’ work demands (changes in supervisors, workgroups, organization, or type of work) to capture maximum fluctuations in workplace telepressure. This would suggest that workplace telepressure might change in short periods of time (e.g., day-to-day), but the changes are temporary and would not be evident when measured several days or weeks after the critical event in the work environment. Indeed, recent research found negative psychological consequences of ICT events on a day-to-day basis (Braukmann et al. in press). Continued research on the stability of telepressure particularly in response to changing workplace environments and demands is needed to understand the onset and duration of workplace telepressure. Perhaps such research questions can be addressed with longitudinal data designs that rely on event-contingent approaches (i.e., multiple measurements taken after discrete events, such as job changes) rather than the interval-contingent approach studied here (Fisher and To 2012).

On a related note, the within-person variation in psychological detachment and well-being outcomes was limited to differences across three measurement periods. Although within-person indirect effects were evident for work engagement variables in this study, the bivariate relationships between those variables and workplace telepressure were not reliable for two of three subscales. With additional measurements, the precise nature of within-person variation in psychological detachment and well-being outcomes could be examined, such as in testing duration or accumulation effects. We recommend that future research expand the design of this study to measure employees at more than three periods to allow for more reliable within-person variation in workplace telepressure, psychological detachment, and complex associations with well-being outcomes. Unlike workplace telepressure, past research has shown psychological detachment to vary day-to-day (Sonnentag and Fritz 2015), thus providing additional precedent to continue the investigation of within-person variation in detachment and the consequences for worker well-being.

Finally, past research showed that workplace telepressure levels were related to perceived organizational norms (Barber and Santuzzi 2015). Thus, the levels of workplace telepressure in the present study might have been consistent because the workplace environment—and thus the norms–-were constant throughout the study. It is also possible that, regardless of the actual norms, the workers’ perceptions of the norms remained consistent over measurement periods. The extent to which workplace telepressure was consistent due to consistent organizational norms or individual differences in perceptions of norms could not be explored in this study. However, exploratory analyses suggested that work environments with low ICT connection demands might be more susceptible to negative workplace telepressure effects than those with high ICT connection demands. Future research should examine employees clustered within organizations to tease apart the extent to which telepressure is consistent due to the workplace environment or individual differences. Such a design also might detect whether certain job types are more susceptible to the negative well-being outcomes of workplace telepressure and whether some worker individual differences predict susceptibility to telepressure in specific job types or work environments. Moreover, such research could identify organizational strategies that might buffer the negative well-being outcomes for workers who experience workplace telepressure.

Conclusion

Overall, our findings provide evidence that workplace telepressure can be understood in terms of health impairment processes related to job demands due to lack of recovery. Specifically, workplace telepressure has a negative relationship with worker well-being through its interference with psychological detachment from work. Workplace telepressure seems to be associated with lower levels of psychological detachment, which in turn predicts higher levels of physical exhaustion, poor sleep quality, and lower levels of work engagement. We encourage organizations to consider how to sustain positive motivational benefits of workplace telepressure over time with appropriate recovery opportunities for employees to avoid negative health effects.

Footnotes

  1. 1.

    We explored whether detachment might serve as a moderator rather than mediator for the engagement outcomes based on alternative propositions of the stressor-detachment model (Sonnentag and Fritz 2015). Results from traditional multilevel model analysis showed a significant interaction only for work engagement as dedication; b = 0.10, SE = .04, p = .03). The simple slope for telepressure on engagement at the average level of psychological detachment was near zero (b = 0.06, SE = .08, p = .48). At low levels of detachment (1.75 SD below the mean), telepressure did not predict dedication, b = 0.06, SE = .11, p = .33. However, at very high levels of detachment (1.75 SD above the mean), more telepressure did predict more dedication, b = 0.22, SE = .11, p = .04. Note that telepressure did not predict absorption at lower levels of detachment (e.g., 1 SD and 1.5 SD above the mean). The results suggest that dedication increases with higher levels of telepressure only when high levels of psychological detachment are also present.

  2. 2.

    None of the interaction effects on burnout variables were mediated by psychological detachment (ps > .09).

Notes

Acknowledgements

The project was supported by a Research and Artistry grant awarded to the authors by Northern Illinois University. The authors are grateful to Sarah Bailey for her assistance with data collection and data management for this project.

Compliance with Ethical Standards

Conflict of Interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

References

  1. Bakker, A. B., & Bal, M. P. (2010). Weekly work engagement and performance: a study among starting teachers. Journal of Occupational and Organizational Psychology, 83(1), 189–206.  https://doi.org/10.1348/096317909X402596.CrossRefGoogle Scholar
  2. Bakker, A. B., & Demerouti, E. (2007). The job demands-resources model: state of the art. Journal of Managerial Psychology, 22(3), 309–328.  https://doi.org/10.1108/02683940710733115.CrossRefGoogle Scholar
  3. Barber, L. K., & Jenkins, J. S. (2014). Creating technological boundaries to protect bedtime: examining work–home boundary management, psychological detachment and sleep. Stress and Health, 30(3), 259–264.CrossRefPubMedGoogle Scholar
  4. Barber, L. K., & Santuzzi, A. M. (2015). Please respond ASAP: workplace telepressure and employee recovery. Journal of Occupational Health Psychology, 20(2), 172–189.  https://doi.org/10.1037/a0038278.CrossRefPubMedGoogle Scholar
  5. Barber, L. K., & Santuzzi, A. M. (2017). Telepressure and college student employment: the costs of staying connected across social contexts. Stress and Health, 33(1), 14–23.CrossRefPubMedGoogle Scholar
  6. Barley, S. R., Meyerson, D. E., & Grodal, S. (2011). E-mail as a source and symbol of stress. Organization Science, 22(4), 887–906.  https://doi.org/10.1287/orsc.1100.0573.CrossRefGoogle Scholar
  7. Bauer, D. J., & Curran, P. J. (2005). Probing interactions in fixed and multilevel regression: inferential and graphical techniques. Multivariate Behavioral Research, 40(3), 373–400.CrossRefPubMedGoogle Scholar
  8. Beal, D. J. (2015). ESM 2.0: State of the art and future potential of experience sampling methods in organizational research. Annual Review of Organizational Psychology and Organizational Behavior, 2(1), 383–407.CrossRefGoogle Scholar
  9. Braukmann, J., Schmitt, A., Ďuranová, L., & Ohly, S. (in press). Identifying ICT-related affective events across life domains and examining their unique relationships with employee recovery. Journal of Business and Psychology.  https://doi.org/10.1007/s10869-017-9508-7.CrossRefGoogle Scholar
  10. Burgoon, J. K., & Hale, J. L. (1988). Nonverbal expectancy violations: model elaboration and application to immediacy behaviors. Communication Monographs, 55(1), 58–79.  https://doi.org/10.1080/03637758809376158.CrossRefGoogle Scholar
  11. Cavanaugh, M. A., Boswell, W. R., Roehling, M. V., & Boudreau, J. W. (2000). An empirical examination of self-reported work stress among US managers. Journal of Applied Psychology, 85(1), 65–74.CrossRefPubMedGoogle Scholar
  12. Christian, M. S., Garza, A. S., & Slaughter, J. E. (2011). Work engagement: a quantitative review and test of its relations with task and contextual performance. Personnel Psychology, 64(1), 89–136.  https://doi.org/10.1111/j.1744-6570.2010.01203.x.CrossRefGoogle Scholar
  13. Crawford, E. R., LePine, J. A., & Rich, B. L. (2010). Linking job demands and resources to employee engagement and burnout: a theoretical extension and meta-analytic test. Journal of Applied Psychology, 95(5), 834–848.CrossRefPubMedGoogle Scholar
  14. Day, A., Scott, N., & Kelloway, E. K. (2010). Information and communication technology: implications for job stress and employee well-being. Research in Occupational Stress and Well-being, 8, 317–350.  https://doi.org/10.1108/S1479-3555(2010)0000008011.CrossRefGoogle Scholar
  15. Day, A., Paquet, S., Scott, N., & Hambley, L. (2012). Perceived information and communication technology (ICT) demands on employee outcomes: the moderating effect of organizational ICT support. Journal of Occupational Health Psychology, 17(4), 473.  https://doi.org/10.1037/a0029837.CrossRefPubMedGoogle Scholar
  16. de Lange, A. H., Taris, T. W., Kompier, M. A., Houtman, I. L., & Bongers, P. M. (2003). “The very best of the millennium”: longitudinal research and the demand-control-(support) model. Journal of Occupational Health Psychology, 8(4), 282–305.CrossRefPubMedGoogle Scholar
  17. Derks, D., & Bakker, A. B. (2014). Smartphone use, work–home interference, and burnout: a diary study on the role of recovery. Applied Psychology, 63, 411–440.  https://doi.org/10.1111/j.1464-0597.2012.00530.x.CrossRefGoogle Scholar
  18. Derks, D., ten Brummelhuis, L. L., Zecic, D., & Bakker, A. B. (2014). Switching on and off … : does smartphone use obstruct the possibility to engage in recovery activities? European Journal of Work and Organizational Psychology, 23(1), 80–90.  https://doi.org/10.1080/1359432X.2012.711013.CrossRefGoogle Scholar
  19. Ferguson, M., Carlson, D., Boswell, W., Whitten, D., Butts, M. M., & Kacmar, K. M. M. (2016). Tethered to work: a family systems approach linking mobile device use to turnover intentions. Journal of Applied Psychology, 101(4), 520–534.  https://doi.org/10.1037/apl0000075.CrossRefPubMedGoogle Scholar
  20. Finkel, S. E. (1995). Causal analysis with panel data. Thousand Oaks: Sage.CrossRefGoogle Scholar
  21. Fisher, C. D., & To, M. L. (2012). Using experience sampling methodology in organizational behavior. Journal of Organizational Behavior, 33(7), 865–877.  https://doi.org/10.1002/job.1803.CrossRefGoogle Scholar
  22. Ford, M. T., Matthews, R. A., Wooldridge, J. D., Mishra, V., Kakar, U. M., & Strahan, S. R. (2014). How do occupational stressor-strain effects vary with time? A review and meta-analysis of the relevance of time lags in longitudinal studies. Work and Stress, 28(1), 9–30.CrossRefGoogle Scholar
  23. Freedy, J. R., & Hobfoll, S. E. (1994). Stress inoculation for reduction of burnout: a conservation of resources approach. Anxiety, Stress, and Coping, 6, 311–325.  https://doi.org/10.1080/10615809408248805.CrossRefGoogle Scholar
  24. Fritz, C., & Sonnentag, S. (2006). Recovery, well-being, and performance-related outcomes: the role of workload and vacation experiences. Journal of Applied Psychology, 91(4), 936–945.  https://doi.org/10.1037/0021-9010.91.4.936.CrossRefPubMedGoogle Scholar
  25. Fritz, C., Yankelevich, M., Zarubin, A., & Barger, P. (2010). Happy, healthy, and productive: the role of detachment from work during nonwork time. Journal of Applied Psychology, 95(5), 977–983.  https://doi.org/10.1037/a0019462.CrossRefPubMedGoogle Scholar
  26. Gilson, L. L., Maynard, M. T., Jones Young, N. C., Vartiainen, M., & Hakonen, M. (2015). Virtual teams research: 10 years, 10 themes, and 10 opportunities. Journal of Management, 41(5), 1313–1337.  https://doi.org/10.1177/0149206314559946.CrossRefGoogle Scholar
  27. Goodman, J. K., Cryder, C. E., & Cheema, A. (2013). Data collection in a flat world: the strengths and weaknesses of mechanical Turk samples. Journal of Behavioral Decision Making, 26, 213–224.  https://doi.org/10.1002/bdm.1753.CrossRefGoogle Scholar
  28. Grawitch, M. J., Werth, P. M., Palmer, S. N., Erb, K. R., & Lavigne, K. N. (2017). Self-imposed pressure or organizational norms? Further examination of the construct of workplace telepressure. Stress and Health, 1–14.  https://doi.org/10.1002/smi.2792.CrossRefPubMedGoogle Scholar
  29. Hahn, V. C., Binnewies, C., Sonnentag, S., & Mojza, E. J. (2011). Learning how to recover from job stress: effects of a recovery training program on recovery, recovery-related self-efficacy, and well-being. Journal of Occupational Health Psychology, 16(2), 202–216.CrossRefPubMedGoogle Scholar
  30. Harpaz, I. (2002). Advantages and disadvantages of telecommuting for the individual, organization and society. Work Study, 51(2), 74–80.  https://doi.org/10.1108/00438020210418791.CrossRefGoogle Scholar
  31. Huang, J. L., Bowling, N. A., Liu, M., & Li, Y. (2015). Detecting insufficient effort responding with an infrequency scale: evaluating validity and participant reactions. Journal of Business and Psychology, 30(2), 299–311.CrossRefGoogle Scholar
  32. Jenkins, C. D., Stanton, B. A., Niemcryk, S. J., & Rose, R. M. (1988). A scale for the estimation of sleep problems in clinical research. Journal of Clinical Epidemiology, 41(4), 313–321.  https://doi.org/10.1016/0895-4356(88)90138-2.CrossRefPubMedGoogle Scholar
  33. Kenny, D. (1975). Cross-lagged panel correlation: a test for spuriousness. Psychological Bulletin, 82, 887–903.  https://doi.org/10.1037/0033-2909.82.6.887.CrossRefGoogle Scholar
  34. Kühnel, J., Sonnentag, S., & Westman, M. (2009). Does work engagement increase after a short respite? The role of job involvement as a double-edged sword. Journal of Occupational and Organizational Psychology, 82(3), 575–594.  https://doi.org/10.1348/096317908X349362.CrossRefGoogle Scholar
  35. Landers, R. N., & Behrend, T. S. (2015). An inconvenient truth: arbitrary distinctions between organizational, mechanical Turk, and other convenience samples. Industrial and Organizational Psychology, 8(2), 142–164.  https://doi.org/10.1017/iop.2015.13.CrossRefGoogle Scholar
  36. Mastin, D. F., Bryson, J., & Corwyn, R. (2006). Assessment of sleep hygiene using the sleep hygiene index. Journal of Behavioral Medicine, 29(3), 223–227.  https://doi.org/10.1007/s10865-9047-6.CrossRefPubMedGoogle Scholar
  37. Mazmanian, M., Orlikowski, W. J., & Yates, J. (2013). The autonomy paradox: the implications of mobile email devices for knowledge professionals. Organization Science, 24(5), 1337–1357.  https://doi.org/10.1287/orsc.1120.0806.CrossRefGoogle Scholar
  38. Meade, A. W., & Craig, S. B. (2012). Identifying careless responses in survey data. Psychological Methods, 17(3), 437–455.  https://doi.org/10.1037/a0028085.CrossRefPubMedGoogle Scholar
  39. Meijman, T. F., & Mulder, G. (1998). Psychological aspects of workload. In P. J. D. Drenth & H. Thierry (Eds.), Handbook of work and organizational psychology, Work psychology (Vol. 2, pp. 5–33). Hove, England: Psychology Press.Google Scholar
  40. Olson-Buchanan, J. B., & Boswell, W. R. (2006). Blurring boundaries: correlates of integration segmentation between work and family. Journal of Vocational Behavior, 68(3), 432–445.  https://doi.org/10.1016/j.jvb.2005.10.006.CrossRefGoogle Scholar
  41. Park, Y., Fritz, C., & Jex, S. M. (2011). Relationships between work-home segmentation and psychological detachment from work: the role of communication technology use at home. Journal of Occupational Health Psychology, 16(4), 457–467.  https://doi.org/10.1037/a0023594.CrossRefPubMedGoogle Scholar
  42. Perlow, L. A., & Porter, J. L. (2009). Making time off predictable-and required. Harvard Business Review, 87(10), 102–109.PubMedGoogle Scholar
  43. Peugh, J. L., & Enders, C. K. (2004). Missing data in educational research: a review of reporting practices and suggestions for improvement. Review of Educational Research, 74(4), 525–556.  https://doi.org/10.3102/00346543074004525.CrossRefGoogle Scholar
  44. Preacher, K. J., Zyphur, M. J., & Zhang, Z. (2010). A general multilevel SEM framework for assessing multilevel mediation. Psychological Methods, 15, 209–233.  https://doi.org/10.1037/a0020141.CrossRefPubMedGoogle Scholar
  45. Preacher, K. J., Zhang, Z., & Zyphur, M. J. (2015). Multilevel structural equation models for assessing moderation within and across levels of analysis. Psychological Methods, 20(4), 1–17.  https://doi.org/10.1037/met0000052.CrossRefGoogle Scholar
  46. Querstret, D., & Cropley, M. (2012). Exploring the relationship between work-related rumination, sleep quality, and work-related fatigue. Journal of Occupational Health Psychology, 17(3), 341–353.  https://doi.org/10.1037/a0028552.CrossRefPubMedGoogle Scholar
  47. Schaufeli, W. B., Bakker, A. B., & Salanova, M. (2006). The measurement of work engagement with a short questionnaire: a cross-national study. Educational and Psychological Measurement, 66(4), 701–716.  https://doi.org/10.1177/0013164405282471.CrossRefGoogle Scholar
  48. Shirom, A., & Melamed, S. (2006). A comparison of the construct validity of two burnout measures in two groups of professionals. International Journal of Stress Management, 13(2), 176–200.  https://doi.org/10.1037/1072-5245.13.2.176.CrossRefGoogle Scholar
  49. Smit, B. W., & Barber, L. K. (2016). Psychologically detaching despite high workloads: the role of attentional processes. Journal of Occupational Psychology, 21(4), 432–442.  https://doi.org/10.1037/ocp0000019.CrossRefGoogle Scholar
  50. Sonnentag, S. (2003). Recovery, work engagement, and proactive behavior: a new look at the interface between nonwork and work. Journal of Applied Psychology, 88(3), 518–528.  https://doi.org/10.1037/0021-9010.88.3.518.CrossRefPubMedGoogle Scholar
  51. Sonnentag, S., & Frese, M. (2003). Stress in organizations. Handbook of Psychology (Vol. 3, pp. 453–491). John Wiley & Sons, Inc.Google Scholar
  52. Sonnentag, S., & Fritz, C. (2007). The recovery experience questionnaire: development and validation of a measure for assessing recuperation and unwinding from work. Journal of Occupational Health Psychology, 12(3), 204–221.  https://doi.org/10.1037/1076-8998.12.3.204.CrossRefPubMedGoogle Scholar
  53. Sonnentag, S., & Fritz, C. (2015). Recovery from job stress: the stressor-detachment model as an integrative framework. Journal of Organizational Behavior, 36(1), 72–103.  https://doi.org/10.1002/job.1924.CrossRefGoogle Scholar
  54. Sonnentag, S., Binnewies, C., & Mojza, E. J. (2010a). Staying well and engaged when demands are high: the role of psychological detachment. Journal of Applied Psychology, 95(5), 965–976.  https://doi.org/10.1037/a0020032.CrossRefPubMedGoogle Scholar
  55. Sonnentag, S., Kuttler, I., & Fritz, C. (2010b). Job stressors, emotional exhaustion, and need for recovery: a multi-source study on the benefits of psychological detachment. Journal of Vocational Behavior, 76(3), 355–365.  https://doi.org/10.1016/j.jvb.2009.06.005.CrossRefGoogle Scholar
  56. Van den Broeck, A., De Cuyper, N., De Witte, H., & Vansteenkiste, M. (2010). Not all job demands are equal: differentiating job hindrances and job challenges in the job demands-resources model. European Journal of Work and Organizational Psychology, 19, 735–759.CrossRefGoogle Scholar
  57. Zapf, D., Dormann, C., & Frese, M. (1996). Longitudinal studies in organizational stress research: a review of the literature with reference to methodological issues. Journal of Occupational Health Psychology, 1(2), 145–169.CrossRefPubMedGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2018

Authors and Affiliations

  1. 1.Department of PsychologyNorthern Illinois UniversityDeKalbUSA
  2. 2.Department of PsychologySan Diego StateSan DiegoUSA

Personalised recommendations