You are on page 1of 11

Journal of Occupational Health Psychology © 2014 American Psychological Association

2015, Vol. 20, No. 1, 82–92 1076-8998/15/$12.00 http://dx.doi.org/10.1037/a0037890

Setting a Good Example: Supervisors as Work-Life-Friendly Role Models


Within the Context of Boundary Management
Anna R. Koch and Carmen Binnewies
University of Muenster

This multisource, multilevel study examined the importance of supervisors as work-life-friendly role
models for employees’ boundary management. Particularly, we tested whether supervisors’ work-home
segmentation behavior represents work-life-friendly role modeling for their employees. Furthermore, we
tested whether work-life-friendly role modeling is positively related to employees’ work-home segmen-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

tation behavior. Also, we examined whether work-life-friendly role modeling is positively related to
This document is copyrighted by the American Psychological Association or one of its allied publishers.

employees’ well-being in terms of feeling less exhausted and disengaged. In total, 237 employees and
their 75 supervisors participated in our study. Results from hierarchical linear models revealed that
supervisors who showed more segmentation behavior to separate work and home were more likely
perceived as work-life-friendly role models. Employees with work-life-friendly role models were more
likely to segment between work and home, and they felt less exhausted and disengaged. One may
conclude that supervisors as work-life-friendly role models are highly important for employees’ work-
home segmentation behavior and gatekeepers to implement a work-life-friendly organizational culture.

Keywords: work-life-friendly role modeling, boundary management, work-home segmentation behavior,


exhaustion, disengagement

Currently, work can be done anywhere and anytime. Boundaries typically cannot rely on formal systems of job descriptions to
between work and home domains are becoming increasingly cultivate boundary management, there should be a strong influence
blurred (Ashforth, Kreiner, & Fugate, 2000). For example, the rise of role models (Yaffe & Kark, 2011).
of communication technology allows wireless Internet access and Until now, knowledge about supervisors as role models for their
facilitates work being done at home (Kurland & Bailey, 1999). employees’ work-home segmentation is scarce, although there
Employees must step up to the new challenge of successfully have been calls for studies on the effect of supervisors as role
organizing the boundaries between their work and home domains. models within the context of boundary management (Hammer,
As most human learning occurs through observational learning Kossek, Yragui, Bodner, & Hanson, 2009). In addition, research
(Bandura, 1977) employees often rely on their supervisors as role aimed to identify starting points for implementing a work-life-
models for learning how to successfully organize their work-home friendly organizational culture (Thompson, Beauvais, & Lyness,
boundaries (Kreiner, Hollensbe, & Sheep, 2009; Park, Fritz, & Jex, 1999). Supervisors constitute promising starting points (O’Neill et
2011). Supervisors are part of employees’ organizational environ- al., 2009). Considering organizational work-life interventions, it
ment, which boundary theory emphasizes as an influential ante- would be highly important to know whether supervisors with high
cedent for employees’ boundary management (Clark, 2000; Ko- work-home segmentation behavior represent work-life-friendly
ssek & Lautsch, 2012). Supervisors embody organizational values role models within their work group. We address these calls and
(Powell & Mainiero, 1999; Scandura & Lankau, 1997) and allow focus on supervisors as role models within the context of boundary
employees to view them as positive examples of organizational management. Relying on social learning (Bandura, 1977) and
boundary management norms. Particularly, because organizations recovery theories (Meijman & Mulder, 1998), we argue that su-
pervisors with high work-home segmentation behavior who define
clear respite from work when being at home represent work-life-
friendly role models. Although the work-home segmentation of
This article was published Online First September 8, 2014. supervisors is not behavior that is directly aimed at employees, it
Anna R. Koch and Carmen Binnewies, Institute of Psychology, Univer- is visible and thus observable by employees (Nippert-Eng, 1996).
sity of Muenster. Employees can observe supervisors giving an example of how to
This study is part of Anna R. Koch’s dissertation. Preliminary results of manage both life domains successfully in terms of defining work-
this study were presented at the 16th Conference of the European Associ- free respites when at home. We propose that work-life-friendly
ation of Work and Organizational Psychology in Muenster, Germany. role modeling is positively related to employees’ individual work-
We thank Wiebke Boess, Anna Sophie Herrmann, Ina Krueger and
home segmentation behavior and well-being (low amounts of
Astrid Wirth for their involvement in data collection. We thank Verena C.
Hahn and Eva Brosch for helpful comments on earlier versions of this
exhaustion and disengagement). Work-life-friendly role models
paper. should facilitate the work-home segmentation behavior of employ-
Correspondence concerning this article should be addressed to Anna R. ees in terms of taking respites, during which recovery can occur.
Koch, Institute of Psychology, University of Muenster, Fliednerstr. 21, Moreover, work-life-friendly role models may give employees the
48149 Muenster, Germany. E-mail: anna.r.koch@uni-muenster.de feeling of being fairly treated because of an adequate amount of

82
SETTING A GOOD EXAMPLE 83

segmentation from work when they are at home, and increase Eng, 1996, p. 7). Thus, supervisors’ work-home segmentation
positive emotions and identification with the organization. behavior can be recognized by employees through direct and
By including the concept of supervisors’ work-life-friendly role indirect indicators (Friedman & Lobel, 2003). Such indicators
modeling into boundary management research, our study contrib- could be supervisors’ direct communication, observed work and
utes to the literature in two ways. First, our study captures the nonwork hours, work results and received and written work-related
relationship between work-life-friendly role modeling and em- correspondence during nonwork time.
ployees’ work-home segmentation behavior. Studying work-life-
friendly role modeling as one element and potential predictor from
Recovery Perspective on Work-Home
the organizational environment will enrich theory building around
Segmentation Behavior
the construct of boundary management. Hence, our study contrib-
utes to boundary literature and helps to better understand the In general, there is not just one right way to manage boundaries.
phenomenon of boundary management itself. Second, our study Research suggests that boundary management highly depends on
examines whether supervisors who show a high degree of segmen- employees’ preferences for segmentation (Park et al., 2011). How-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

tation behavior between the work domain and home domain are ever, we argue that from a recovery perspective at least a slight
This document is copyrighted by the American Psychological Association or one of its allied publishers.

experienced as work-life-friendly role models within their work degree of segmentation between work and home domain is essen-
groups. Focusing on the relationship between supervisors’ own tial to release load reactions and secure well-being (Meijman &
work-home segmentation behavior and work-life-friendly role Mulder, 1998). According to the Effort-Recovery model (Meijman
modeling allows us to draw a more comprehensive picture of the & Mulder, 1998), recovery occurs with the absence of work
potential consequences of boundary management. Our study stressors. If an employee stops working and depleting resources,
thereby aims to identify new starting points for implementing a the functional system returns to the baseline level and recovery
work-life-friendly organizational culture. automatically occurs (Binnewies & Sonnentag, 2008; Meijman &
Mulder, 1998). Therefore, according to the Effort-Recovery model
(Meijman & Mulder, 1998) recovery is essential to segment life
Boundary Management
domains and find at least some respite from work when at home.
Boundary theory indicates that employees create and maintain Recovery research supports this assumption and shows that recov-
boundaries between their life domains, such as work and home ery during nonwork time is beneficial for well-being and health
(Bulger, Matthews, & Hoffman, 2007; Clark, 2000; Nippert-Eng, (Sonnentag & Bayer, 2005; Sonnentag & Fritz, 2007). However, if
1996). Boundaries can be constructed along a continuum from time for recovery is not sufficient, an employee’s functional sys-
weak ones (high integration between domains) to strong ones (high tem does not return to the baseline level and the performance of
segmentation between domains). Employees differ in their behav- upcoming work tasks is impeded (Binnewies & Sonnentag, 2008;
iors and strategies for integrating and segmenting life domains Meijman & Mulder, 1998). For example, job-related activities
(Hecht & Allen, 2009; Nippert-Eng, 1996). These behavioral strat- during nonwork time further draw on resources needed at work
egies aimed at integrating and segmenting life domains have been and inhibit recovery (Sonnentag, 2001). Consequently, an em-
defined as boundary management (Ashforth et al., 2000; Bulger et ployee must invest compensatory effort in order to maintain the
al., 2007; Clark, 2000; Nippert-Eng, 1996). Within the context of fulfillment of tasks and goals (Meijman & Mulder, 1998). Insuf-
work-life-balance, boundary management (e.g., bringing elements ficient recovery and increased fatigue can accumulate, when re-
of one domain into the other) comprises two typical behaviors, covery is lacking over longer periods of time (Meijman & Mulder,
referring to the integration of home into work and work into home 1998). Over time, negative consequences, such as severe health
(Bulger et al., 2007; Hecht & Allen, 2009). Home-work segmen- complaints that cannot be easily reversed during the usual rest
tation behavior refers to the degree to which employees prevent periods can result (Burke & Fiksenbaum, 2008). In conclusion,
various aspects of the home domain from entering the work from a recovery perspective work-home segmentation in terms of
domain. Second, work-home segmentation behavior refers to the finding respite from work is needed to recover from work stress
degree to which employees prevent various aspects of work from (Park et al., 2011). High integration of work into the home domain
entering the home domain. In our study, we focus on the second without respite harms recovery (Park et al., 2011).
dimension: work-home segmentation behavior. Work-home-
segmentation behavior refers to the extent that employees work out
Supervisors as Work-Life-Friendly Role Models
of office, write or receive work-related correspondence at home,
and define clear respites (Hecht & Allen, 2009). Employees with Social learning theory emphasizes the importance of role mod-
high work-home segmentation behavior draw a line between work els for employees’ behaviors and attitudes within the organiza-
and home domains (Kreiner et al., 2009). For example, employees tional context (Bandura, 1977). Supervisors are emphasized as
take respites during which they are not available for work issues particularly important role models, for example by transforma-
while at home. In contrast, employees with low work-home seg- tional leadership theories (Bass & Avolio, 1994; Judge & Bono,
mentation behavior are likely to integrate work into the home 2000). Supervisors’ behavior shapes organizational norms and
domain, for example by doing work-related correspondence at embodies organizational values (Hammer et al., 2009; Powell &
home (Hecht & Allen, 2009). Boundary management is visible and Mainiero, 1999; Scandura & Lankau, 1997). For example, super-
thus observable by others (Nippert-Eng, 1996). Nippert-Eng visors embody organizational valued boundary management.
(1996) states that boundary management “is first and foremost a Therefore, we argue that supervisors who segment more between
mental activity, but it must be enacted and enhanced through a work and home represent stronger work-life-friendly role models
largely visible collection of essential, practical activities” (Nippert- within their work groups.
84 KOCH AND BINNEWIES

We define work-life-friendly role models as those who provide Demerouti, Bakker, Nachreiner, & Schaufeli, 2001). Exhaustion is
examples of strategies and behaviors that allow employees to a consequence of intense strain, and a long-term consequence of
recover from work stress. Supervisors who are perceived as work- prolonged exposure to certain job demands (Demerouti et al.,
life-friendly role models set a good example in terms of work-life 2001). Disengagement refers to distancing oneself from one’s
balance, for example by taking respites during nonwork time and work in general, work tasks and work content (Demerouti et al.,
refraining from work-related activities. 2001). Low disengagement is associated with high identification
In our study, we conceptualize work-life-friendly role modeling with work and willingness to stay in the same occupation (Demer-
as a shared group-level variable that reflects work group members’ outi, Mostert, & Bakker, 2010). In general, employees rely on role
perceptions of the extent to which their supervisor constitutes a models by forming expectations for themselves as well as guide-
work-life-friendly role model. This definition implies that employ- lines for future behavior (Bandura, 1977), which in turn affects
ees working in the same work group are likely to experience their well-being and behavior. A role model’s behavior is copied
similar supervisor boundary management. This definition is con- particularly at times when rewarding consequences are expected
sistent with Chen and Bliese (2002) who conceptualized leadership (Bandura, 1977). Supervisors represent organizational values
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

behavior as shared group-level perception and implies that super- (Powell & Mainiero, 1999; Scandura & Lankau, 1997), have high
This document is copyrighted by the American Psychological Association or one of its allied publishers.

visors’ work-home segmentation behavior is visible within their decisional power, and are often responsible for hiring, evaluating
work group. Work-life role modeling comprises supervisors’ mod- and promoting employees. As a consequence, they constitute role
eling behaviors on the job (Hammer et al., 2009). We propose that models whose behavior is likely to be imitated (Bandura, 1977;
supervisors’ work-home segmentation behavior is a particularly Weiss, 1978). Previous research has shown that work-home seg-
important modeling behavior. Supervisors who highly segment mentation behavior depends not only on individual preferences for
between work and home domains create work-free spaces in their integration and segmentation but also on organizational norms
private lives (e.g., engaging in leisure and family activities) during (Park et al., 2011), as represented by the supervisor.
which they are not available for work-related issues. These super- We argue that employees should be more likely to intentionally
visor behaviors provide employees with the impression that the define respites and believe that they do not have to be always
organization values private life and that taking time off does not available when working with a work-life-friendly role model. In
hinder organizational success (Friedman & Lobel, 2003). Super-
order to be successful in the organization, employees are likely to
visors who highly integrate work into the home domain with an
adapt their behavior according to the work-life-friendly role model
‘always-on’ mentality do not intentionally create work-free spaces
(Park et al., 2011) by segmenting stronger between work and home
in their private lives. These supervisors are likely to engage in
domains. Moreover, employees should feel fairly treated and ex-
work issues at any time (Hecht & Allen, 2009). Supervisors with
perience support for their private lives, as organizations allow time
high work-home integration behavior may therefore imply that the
for respites. Positive emotions and identification with the organi-
organization expects employees to be always available even during
zation should thereby also be higher. Also, during work-free
nonwork time, probably to the detriment of family obligations and
respites load reactions (e.g., fatigue) can be released and recovery
leisure activities (Friedman & Lobel, 2003). Therefore, these su-
can occur, which should benefit employees’ well-being (Cohen-
pervisors may embody the perception that there is low organiza-
Charash & Spector, 2001; Meijman & Mulder, 1998). In sum, we
tional support for employees’ private lives and recovery from work
stress. argue that work-life-friendly role modeling should be positively
Empirical evidence supports our proposition: Kirby and Krone related to employees’ work-home segmentation behavior and well-
(2002) showed that other than what supervisors communicate, being (low amounts of exhaustion and disengagement).
employees look at their supervisors’ behavior (e.g., hours worked A work-life-unfriendly role model makes it more likely that
overtime) regarding their use of alternative work arrangements employees conclude that their organization expects an “always-
(e.g., flextime). Also Hammer and colleagues (2009) found that on” mentality. Therefore, it will be more likely that employees
supervisors are experienced as role models within the context of integrate work into their home domain, for example, by taking
work-life-balance. fewer work-free respites. These employees are more likely to feel
unfairly treated by disproportionate expectations of work-home
Hypothesis 1: Supervisors with high work-home segmentation integration. Employees are more likely to feel distressed and less
behavior represent work-life-friendly role models within their committed to their organization (Cohen-Charash & Spector, 2001).
work group. Moreover, if employees do not take work-free respites, they cannot
recover from load reactions, which may result in impaired well-
being (Meijman & Mulder, 1998). Therefore, we argue that work-
Work-Life-Friendly Role Modeling Is Related to
life-unfriendly role models should be negatively related to employ-
Employees’ Work-Home Segmentation
ees’ work-home integration and reduce employees’ well-being
Behavior and Well-Being (high amounts of exhaustion and disengagement).
We argue that work-life-friendly role models constitute an im- There is initial empirical evidence that supports our assump-
portant resource for employees in the organizational context tions. Park and colleagues (2011) found that the organizational
(Hammer et al., 2009) and should be positively related to employ- segmentation norm affects whether employees can refrain from
ees’ own work-home segmentation behavior. Moreover, work-life work-related thoughts during nonwork time. Burke (2006) found
role modeling should be positively associated with employees’ that employees who experienced their organization (and supervi-
well-being. Regarding employees’ well-being, we focused on two sor) as valuing work-life-balance felt less forced to work extended
core dimensions: emotional exhaustion and disengagement (cf. hours.
SETTING A GOOD EXAMPLE 85

Hypothesis 2: Work-life-friendly role modeling will have a Measures


direct positive relationship with employees’ work-home seg-
mentation behavior. Supervisors completed one questionnaire, which comprised
their work-home segmentation behavior and demographic data.
Hypothesis 3: Work-life-friendly role modeling will have a Employee data was measured with two questionnaires: a predictor
direct negative relationship with employees’ exhaustion and and an outcome variable questionnaire. In the predictor variable
disengagement. questionnaire, employees’ perception of their supervisors as work-
life-friendly role models was measured. Furthermore, employees’
job involvement as a control variable and demographic data were
Method collected. In the outcome variable questionnaire that was com-
We used a multisource design with employees and their super- pleted 2 days after the predictor variable questionnaire, employees’
visors to test our hypotheses. Participants were white-collar em- work-home segmentation behavior as well as their exhaustion and
ployees and supervisors from various German organizations. To disengagement were measured. We used two separate question-
naires for employees to reduce common method bias. As recom-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

recruit participants we contacted and informed managers from


mended by Podsakoff, MacKenzie, Lee, and Podsakoff (2003), we
This document is copyrighted by the American Psychological Association or one of its allied publishers.

various companies (e.g., mostly local and federal administrations,


health insurance companies and care providers) about our study on separated the measurement of the predictor (role modeling percep-
“work-life balance and recovery.” After managers expressed in- tion) from the outcomes (employees’ work-home segmentation
terest in participation, we provided them with information leaflets behavior and well-being). Our research design is displayed in
that they used to inform employees. Employees could then register Figure 1.
online for study participation. Upon registration we sent employ- Supervisors’ questionnaire. Supervisors’ work-home seg-
ees packages including instruction leaflets, paper-and-pencil ques- mentation behavior was assessed with Hecht and Allen’s (2009)
tionnaires and postpaid return envelopes. Employees completed eight-item boundary strength at home scale. A sample item is “I
the predictor variable questionnaire immediately, whereas the out- often do work at home” (reverse coded). Answering categories
come variable questionnaire was completed 2 days later. Employ- were “1 ⫽ strongly disagree” to “7 ⫽ strongly agree.” Cronbach’s
ees also received questionnaires for their supervisors and were alpha was .88.
asked to invite their supervisors to participate in the study by Employees’ predictor variable questionnaire. Work-life-
giving them the questionnaires. Like employees, supervisors could friendly role modeling was assessed with a subscale of Hammer
then decide whether or not they wanted to participate in the study. and colleagues’ (2009) family supportive supervisor behavior
All questionnaires were sent back via postpaid envelopes. Employ- (FSSB) scale. The subscale contains three items. A sample item is
ees’ and supervisors’ questionnaires were matched with a code. “My supervisor is a good role model for work and nonwork
In total, 403 employees and 81 supervisors participated in our balance.” Answering categories were “1 ⫽ strongly disagree” to
study. A total of 149 employees1 had to be dropped from the “5 ⫽ strongly agree.” Cronbach’s alpha was .88.
analyses, as they could not be matched with supervisor data. For Employees’ job involvement was measured as a control variable.
all study variables as well as for all demographic variables, we Boundary theory and recent research findings (Ashforth et al.,
conducted t-tests in order to compare employees whose supervi- 2000; Hecht & Allen, 2009) suggest that strong involvement
sors had participated with employees whose supervisors had not within the work domain makes integration of work into the home
participated. T-tests showed that employees whose supervisors had domain more likely. Employees who strongly identify themselves
participated did not significantly differ from those employees with their job seem to be particularly willing to put their private
whose supervisors had not participated with respect to job involve- life on hold in favor of work issues. Job involvement was assessed
ment, work-life role modeling, work-home segmentation behavior with a five-item short scale of Kanungo’s job involvement ques-
and exhaustion. The only significant difference between the groups tionnaire (1982). This scale measures the degree of psychological
was that disengagement was higher for employees whose super- importance of one’s job using a five-point Likert scale from “1 ⫽
visors had not participated. One possible explanation for this strongly disagree” to “5 ⫽ strongly agree.” A sample item is “I
finding is that less highly disengaged employees did invite their consider my job to be very central to my existence.” Cronbach’s
supervisors to participate. alpha was .87.
Another 17 employees and six supervisors had to be excluded Supervisors’ emotional work-life support was measured as a
because of missing data on study variables. Thus, our final sample further control variable. Work-life research (e.g., Hammer et al.,
consisted of 237 employees and 75 supervisors. The majority of 2009; Thompson et al., 1999) has shown that supervisors’ work-
employees was female (67.4%). Employees’ mean age was 41.93 life support benefits their employees in terms of work-home issues.
years (SD ⫽ 13.15). On average, 31.6% of employees lived To rule out proposed relationships simply being because of super-
together with a partner and 37.6% lived together with a partner and visors’ expressed support we decided to control for supervisors’
children. The average number of children was 0.82 (SD ⫽ 0.98).2
On average, employees had worked 10 years (SD ⫽ 9.78) in their 1
In our study, employees who participated received the supervisor’s
current company and 1.26% held a leadership position (SD ⫽ questionnaire and were asked to forward it to their supervisor. Thus, the
0.99). With regard to supervisors, 45.5% were female and on dropout of the 149 employees without matchable supervisor data could
average 49.25 years old (SD ⫽ 11.50). The average employment have been caused by both employees and supervisors. Because of this
procedure and reasons of anonymity, the rate of nonresponding supervisors
duration of supervisors within their current company was 15.81 cannot be determined.
years (SD ⫽ 9.86). On average, 12.57 employees per supervisor 2
Neither employees’ living situation nor their number of children was
participated (SD ⫽ 12.53). significantly related to our study’s predictor and outcome variables.
86 KOCH AND BINNEWIES

Supervisors’ Employees’ Employees’


quesonnaire predictor variable quesonnaire outcome variable questionnaire

- Work-life friendly role- - Employees’ work-home


modeling segmentaon behavior

Supervisors’ work-home - Employees’ job involvement - Employees’ exhauson


segmentaon behavior
- Emoonal work-life support - Employees’ disengagement

Figure 1. Conceptual model and research design.

emotional work-life support within our analyses. Supervisors’ To test our hypotheses, all predictor variables were centered
emotional work-life support was measured with a subscale of around the grand mean.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Hammer and colleagues’ (2009) family supportive supervisor be-


This document is copyrighted by the American Psychological Association or one of its allied publishers.

havior (FSSB) scale. The subscale measures supervisors’ care and Results
verbalized support toward employees’ work-life balance with four
items. A sample item is “My supervisor makes me feel comfort- Means, standard deviations, Cronbach’s alphas, intraclass cor-
able talking to him or her about my conflicts between work and relations and zero-order correlations are displayed in Table 1. To
nonwork.” Answering categories were “1 ⫽ strongly disagree” to test for the relationship between supervisors’ work-home segmen-
“5 ⫽ strongly agree.” Cronbach’s alpha was .92. tation behavior and work-life-friendly role modeling (Hypothesis
Employees’ outcome variable questionnaire. Employees’ 1) we conducted HLM analyses. In the Null model, the intercept
work-home segmentation behavior was assessed with Hecht and was the only predictor. In Model 1, we entered employees’ job
Allen’s (2009) eight-item boundary strength at home scale. A involvement and supervisors’ emotional work-life support as con-
sample item is “I often do work at home” (reverse coded). An- trol variables. In Model 2, we entered supervisors’ work-home
swering categories were “1 ⫽ strongly disagree” to “7 ⫽ strongly segmentation behavior as Level 2 predictor variable. Supervisors’
agree.” Cronbach’s alpha was .82. emotional work-life support in Model 1 (b ⫽ 0.57, p ⬍ .01) and
Exhaustion and disengagement were assessed with the Olden- supervisors’ work-home segmentation in Model 2 (b ⫽ 0.29, p ⬍
burg Burnout Inventory (Demerouti, Bakker, Vardakou, & Kantas .01) emerged as significant positive predictors of work-life-
Aristotelis, 2003). Emotional exhaustion was measured with eight friendly role modeling (see Table 2). Thus, supervisors who pro-
items. A sample item is “After my work, I usually feel worn out vided support and segmented work from the home domain were
and weary.” Disengagement was also measured with eight items. experienced as stronger work-life-friendly role models within their
An example item is “I usually talk about my work in a derogatory work groups, which supports Hypothesis 1.
way.” Answering categories for both dimensions were “1 – totally To test whether work-life-friendly role modeling is related to
disagree” to “4 – totally agree.” Cronbach’s alphas were .82 for employees’ work-home segmentation behavior and well-being
exhaustion and .78 for disengagement. (Hypotheses 2–3) we also used HLM analyses. In the Null model,
the intercept was the only predictor. In Model 1, we entered
employees’ job involvement and supervisors’ emotional work-life
Data Analysis
support as control variables. In Model 2, we entered supervisors’
We had data at two levels: the employee level (Level 1) and the work-home segmentation behavior and work-life-friendly role
supervisor level (Level 2). Employee-level data was nested within modeling as predictor variables.
supervisors. In order not to overestimate statistical significance Table 3 displays the results for employees’ work-home segmen-
(Urbach & Austin, 2005) we applied hierarchical linear models to tation behavior as an outcome variable. In Model 1, job involve-
test our hypotheses (Snijders & Bosker, 1999). We used the HLM ment was a significant negative (b ⫽ ⫺.22, p ⬍ .001) predictor of
software (HLM Version 6; Raudenbush, Bryk, Cheong, & Cong- employees’ work-home segmentation behavior. In Model 2, work-
don, 2004) to analyze the data with hierarchical linear modeling. life-friendly role modeling was a significant positive predictor

Table 1
Means, Standard Deviations (SDs), Cronbach’s Alphas, Intra Class Correlations (ICCs) and Correlations Between Study Variables

Variables Mean SD ␣ ICC 1 2 3 4 5 6

1. Employees’ job involvement 4.12 1.16 0.87 0.10


2. Emotional work-life support 3.56 0.97 0.92 0.08 .18ⴱⴱ
3. Supervisors’ work-home segmentation behavior 3.15 0.82 0.88 .06 ⫺.05
4. Work-life-friendly role modeling 3.25 0.98 0.88 0.14 .20ⴱⴱ .56ⴱⴱ .24ⴱⴱ
5. Employees’ work-home segmentation behavior 3.85 0.85 0.82 0.08 ⫺.29ⴱⴱ .08 .04 .14ⴱ
6. Employees’ exhaustion 2.14 0.55 0.82 0.06 ⫺.02 ⫺.15ⴱ ⫺.02 ⫺.23ⴱⴱ ⫺.15ⴱ
7. Employees’ disengagement 1.89 0.53 0.78 0.13 ⫺.27ⴱⴱ ⫺.27ⴱⴱ ⫺.04 ⫺.27ⴱⴱ .17ⴱ ⫺.36ⴱⴱ
Note. N at the employee level ⫽ 237. N at the supervisor level ⫽ 75.

p ⬍ .05. ⴱⴱ p ⬍ .01.
SETTING A GOOD EXAMPLE 87

Table 2
Multilevel Estimates for Models Predicting Work-Life-Friendly Role Modeling

Nullmodel Model 1 Model 2


Estim. SE t Estim. SE t Estim. SE t

Intercept 3.24 0.08 38.90 3.26 0.08 42.18 3.27 0.07 48.52
Employees’ job involvement 0.05 0.05 1.05 0.06 0.05 1.14
Emotional work-life support 0.57ⴱⴱⴱ 0.08 7.02 0.58ⴱⴱⴱ 0.08 7.49
Supervisors’ work-home segmentation behavior 0.29ⴱⴱ 0.10 2.84
⫺2ⴱlog (lh) 650.25 551.85 541.96
⌬⫺2ⴱlog 98.40ⴱⴱⴱ 9.89ⴱⴱ
⌬ DF 2 1
Level 1 Intercept Variance (SE) 0.82 (0.09) 0.50 (0.05) 0.51 (0.05)
Level 2 Intercept Variance (SE) 0.13 (0.08) 0.18 (0.07) 0.11 (0.05)
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Note. N at employee level ⫽ 237. N at supervisor level ⫽ 75. Unstandardized estimates are reported.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

ⴱⴱ
p ⬍ .01. ⴱⴱⴱ p ⬍ .001.

(b ⫽ 0.17, p ⬍ .05). Supervisors’ work-home segmentation be- havior. Referring to exhaustion and disengagement as outcome
havior was not significantly related to employees’ work-home variables, Sobel tests did not reveal significant indirect effects.
segmentation behavior. Consequently, Hypothesis 2 was support- In the context of two-level data, Multilevel SEM constitutes a
ed: Work-life-friendly role modeling was positively related to more sophisticated way of testing multilevel mediation (Preacher,
employees’ work-home segmentation behavior. Zyphur, & Zhang, 2010; Preacher, Zhang, & Zyphur, 2011). To
Table 4 displays the results for employees’ exhaustion as an test for the indirect relationships between supervisors’ work-home
outcome variable. In Model 1, supervisors’ emotional work-life segmentation behavior and employees’ outcomes via work-life
support (b ⫽ ⫺.08, p ⬍ .05) and in Model 2, work-life-friendly role modeling we also conducted Multilevel SEM, using Mplus 7.0
role modeling emerged as significant negative predictors (Muthén & Muthén, 1998 –2012). Specifically, we entered work-
(b ⫽ ⫺.14, p ⬍ .01). Supervisors’ work-home segmentation life role modeling as a latent variable. Concerning all three out-
behavior was not a significant predictor. come variables, Multilevel SEM showed similar relationships as
Table 5 displays the results for employees’ disengagement as an the Sobel tests. However, concerning employees’ work-home seg-
outcome variable. In Model 1, job involvement (b ⫽ ⫺.10, p ⬍ mentation behavior as an outcome variable, the indirect effect of
.001) and supervisors’ emotional work-life support (b ⫽ ⫺.11, p ⬍ supervisors’ work-home segmentation behavior via work-life role
.01) emerged as negative predictors of employees’ disengagement. modeling was only marginally significant (b ⫽ .04, SE ⫽ .02, p ⫽
In Model 2, work-life-friendly role modeling (b ⫽ ⫺.09, p ⬍ .05) .075). The slightly smaller effect may be because of the small
but not supervisors’ work-home segmentation behavior was a sample size at Level 2 (75 supervisors), which limits statistical
significant negative predictor. In sum, work-life-friendly role mod- power. Nevertheless, Multilevel SEM supports the proposition that
eling was positively related to employees’ well-being, such that work-life role modeling functions as a mediator between supervi-
employees felt less exhausted and disengaged. Thus, Hypothesis 3 sors’ and employees’ work-home segmentation behavior.
was confirmed. We also tested the effects of supervisors’ and employees’ gen-
der on our proposed relationships. We tested for main effects,
Testing for Indirect Effects of Supervisors’ cross-level interactions (supervisors’ gender and work-life role
modeling) and effects of specific gender pairings but could not
Work-Home Segmentation Behavior and Employees’
identify any significant gender effects.
Work-Home Segmentation Behavior and Well-Being
Because we found positive relationships between supervisors’ Discussion
work-home segmentation behavior and work-life role modeling, as
well as between work-life role modeling and employees’ outcome This multisource, multilevel study examined the importance of
variables (work-home segmentation behavior, exhaustion, disen- supervisors as role models within the context of boundary man-
gagement), the question arises whether work-life-friendly role agement. Recent research highly emphasizes the importance of
modeling is the linking mechanism and functions as a mediator.3 supervisors as role models for their employees’ boundary between
Therefore, we tested for indirect effects of supervisors’ work-home work and home domains (Hammer et al., 2009; Kirby & Krone,
segmentation behavior on employees’ work-home segmentation 2002; Regan, 1994). However, to the best of our knowledge our
behavior and well-being, conducting Sobel tests (Sobel, 1982) and study is the first to examine the relationship between supervisors’
Multilevel SEM. Concerning employees’ work-home segmenta- actual work-home segmentation behavior and employees’ percep-
tion behavior as an outcome variable, Sobel tests revealed an tion of work-life role modeling. By our study, we can enrich theory
indirect effect of supervisors’ work-home segmentation behavior
(Sobel’s z ⫽ 2.04, p ⬍ .05). This finding indicates that work-life 3
Additional analyses showed that work-life role modeling was not a
role modeling mediates the relationship of supervisors’ work- significant moderator between supervisors’ work-home segmentation and
home segmentation and employees’ work-home segmentation be- our three outcome variables.
88 KOCH AND BINNEWIES

Table 3
Multilevel Estimates for Models Predicting Employees’ Work-Home Segmentation Behavior

Nullmodel Model 1 Model 2


Estim. SE t Estim. SE t Estim. SE t

Intercept 3.85 0.07 56.64 3.85 0.06 63.63 3.86 0.06 59.86
Employees’ job involvement ⫺0.22ⴱⴱⴱ 0.05 ⫺4.76 ⫺0.23 0.05 ⫺4.97
Emotional work-life support 0.14 0.07 1.95 0.06 0.06 0.91
Supervisors’ work-home segmentation behavior 0.12 0.08 1.48
Work-life-friendly role modeling 0.17ⴱ 0.08 2.08
⫺2ⴱlog (lh) 590.62 566.42 556.01
⌬ ⫺2ⴱlog 24.20ⴱⴱⴱ 10.41ⴱⴱ
⌬ DF 2 2
Level 1 Intercept Variance (SE) 0.67 (0.07) 0.62 (0.06) 0.57 (0.06)
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Level 2 Intercept Variance (SE) 0.06 (0.05) 0.03 (0.03) 0.06 (0.04)
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Note. N at employee level ⫽ 237. N at supervisor level ⫽ 75. Unstandardized estimates are reported.

p ⬍ .05. ⴱⴱ p ⬍ .01. ⴱⴱⴱ p ⬍ .001.

building around the construct of boundary management and iden- a broader sense. Our results are consistent with findings from
tify starting points for implementing a work-life-friendly organi- Hammer and colleagues (2009) who found that family-supportive
zational culture within organizations. As boundary theory suggests supervisor behavior, from which role modeling is one subcompo-
(Kossek & Lautsch, 2012), we found that work-life-friendly role nent, decreases employees’ work-home conflict. Employees with a
modeling is positively related to employees’ individual work- family-supportive role model reported less conflict from work-
home segmentation behavior and well-being. The perception of related issues entering the home domain. However, Hammer and
working with a work-life-friendly role model seems to give em- colleagues (2009) acknowledged that predictors of supervisors’
ployees the feeling that it is socially accepted within the organi- role modeling as well as underlying mechanisms between super-
zation to take time away from work and that organizations require visors’ support and employees’ boundary management should be
an appropriate amount of work-home integration. When working identified by further research.
with a work-life-friendly role model, employees were more likely Our results show that supervisors’ work-home segmentation
to segment between life domains in terms of taking respites. Also, behavior is related to work-life role modeling. Supervisors with
these employees reported better well-being, in terms of feeling less high work-home segmentation behavior draw a sharper line be-
exhausted and disengaged. Lockwood (2006) suggested that fe- tween life domains. Employees have the feeling that these super-
male employees may derive particular benefit from the example of visors are able to meet the needs from both life domains and
a female supervisor with family duties as a role model. However, balance both domains successfully. This finding is highly impor-
our findings do not support these assumptions, as we could not find tant, as supervisors’ work-home segmentation is not a behavior
any significant effects of gender on our study variables and pro- that is aimed at employees, compared with the leadership behav-
posed relationships. iors or emotional work-life support that we also identified as being
It may be concluded from our results that it is functionally significantly related to work-life role modeling. Many supervisors
strategic for organizations to stimulate supervisors’ work-life- are probably not even aware that their work-home segmentation
friendly role modeling to benefit their employees’ work-home behavior is recognized and has an effect on employees. Our
segmentation behavior, well-being, and their work-life-balance in findings are in line with research suggesting that supervisors’ own

Table 4
Multilevel Estimates for Models Predicting Employees’ Exhaustion

Nullmodel Model 1 Model 2


Estim. SE t Estim. SE t Estim. SE t

Intercept 2.17 0.04 51.17 2.17 0.04 50.38 2.17 0.04 51.21
Employees’ job involvement ⫺0.01 0.03 ⫺0.26 0.00 0.03 0.04
Emotional work-life support ⫺0.08ⴱ 0.04 ⫺2.28 ⫺0.01 0.04 ⫺0.14
Supervisors’ work-home segmentation behavior 0.01 0.05 0.11
Work-life-friendly role modeling ⫺0.14ⴱⴱ 0.05 ⫺3.05
⫺2ⴱlog (lh) 387.95 382.76 373.10
⌬ ⫺2ⴱlog 5.19 9.66ⴱⴱ
⌬ DF 1 2 2
Level 1 Intercept Variance (SE) 0.29 (0.03) 0.28 (0.03) 0.27 (0.03)
Level 2 Intercept Variance (SE) 0.02 (0.02) 0.02 (0.02) 0.02 (0.02)
Note. N at employee level ⫽ 237. N at supervisor level ⫽ 75. Unstandardized estimates are reported.

p ⬍ .05. ⴱⴱ p ⬍ .01.
SETTING A GOOD EXAMPLE 89

Table 5
Multilevel Estimates for Models Predicting Employees’ Disengagement

Nullmodel Model 1 Model 2


Estim. SE t Estim. SE t Estim. SE t

Intercept 1.89 0.04 42.20 1.90 0.04 49.68 1.90 0.04 49.14
Employees’ job involvement ⫺0.10ⴱⴱⴱ 0.02 ⫺4.97 ⫺0.10ⴱⴱⴱ 0.02 ⫺4.34
Emotional work-life support ⫺0.11ⴱⴱ 0.04 ⫺2.87 ⫺0.06 0.05 ⫺1.32
Supervisors’ work-home segmentation behavior 0.01 0.04 0.31
Work-life-friendly role modeling ⫺0.09ⴱ 0.04 ⫺2.09
⫺2ⴱlog (lh) 360.71 333.82 329.50
⌬ ⫺2ⴱlog 26.89ⴱⴱⴱ 4.31
⌬ DF 1 2 2
Level 1 Intercept Variance (SE) 0.24 (0.03) 0.23 (0.02) 0.22 (0.02)
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Level 2 Intercept Variance (SE) 0.04 (0.02) 0.02 (0.01) 0.02 (0.01)
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Note. N at employee level ⫽ 237. N at supervisor level ⫽ 75. Unstandardized estimates are reported.

p ⬍ .05. ⴱⴱ p ⬍ .01. ⴱⴱⴱ p ⬍ .001.

behavior has an effect on employees, over and above what super- is not enough (Dikkers et al., 2007; Kirby & Krone, 2002; Lewis,
visors directly communicate to employees (Kirby & Krone, 2002). 1997). Regan (1994) suggested that cultural change would only
Friedman and Lobel (2003) suggest that “no matter what you say, come about when supervisors also embrace work-life values, both
people look at what you do. And if you’re working 15 hours a day in what they say and in what they do. Our study brings up initial
and not taking any vacation, it will be believed that emulating that evidence for this assumption, and shows that supervisors with high
in the organization is a way to success” (p.88). Our study’s work-home segmentation behavior constitute work-life-friendly
findings support these suggestions: although supervisors’ emo- role models. Our study also contributes to boundary theory and
tional work-life support also relates to work-life role modeling, shows that elements from the organizational environment, in terms
supervisors’ work-home segmentation behavior was related to of work-life role models, are related to employees’ work-home
work-life role modeling above and beyond the expressed support. segmentation behavior.
Supervisors’ work-home segmentation behavior was measured as In line with previous research, our results show that individual
a more distal construct than emotional work-life support, which predictors, such as employees’ own job involvement (Hecht &
yields common method variance with work-life role modeling as Allen, 2009), are also related to work-home segmentation behav-
both constructs were measured in the same employee question- ior. Additionally, correlations show a significant positive correla-
naire. tion between employees’ disengagement and work-home segmen-
We also found a significant indirect relationship between super- tation behavior (r ⫽ .17). Disengaged employees experience
visors’ work-home segmentation behavior and employees’ work- negative attitudes toward their work (Demerouti et al., 2001) and
home segmentation behavior via work-life role modeling, and thus could therefore be more likely to implement a high work-home
identified work-life role modeling as mediator according to the boundary to minimize negative work-home spillover. However,
definition of Hayes (2009); Judd and Kenney (2010) and this explanation should be further tested in future research. Fur-
MacKinnon, Krull, and Lockwood (2000). We can conclude from thermore, future studies should focus on examining individual and
this result that supervisors’ work-home segmentation behavior organizational predictors together and how they interact with each
represents work-life-friendly role modeling within their work
other. For example, Kossek and Lautsch (2012) suggested that the
group, which in turn is positively related to their employees’
boundary management of employees might be a function of indi-
work-home segmentation behavior. Regarding the well-being of
vidual preferences in relation to the organizational context in
employees, no indirect relationships could be identified. In terms
which these styles are enacted. Referring to our study, employees’
of our findings, it could be suggested that there is no direct
individual preference for segmentation may be a moderator in the
relationship, as supervisors’ work-home segmentation behavior
relationship between work-life-friendly role modeling and em-
and employees’ well-being constitute distal constructs. We can
ployees’ outcomes. Employees who prefer segmentation between
conclude from this result that work-life-friendly role modeling is
life domains may particularly benefit from a work-life-friendly
not the connecting mechanism between supervisors’ work-home
role model and experience a work-life-unfriendly role model as a
segmentation behavior and employees’ well-being.
source of stress (Kossek & Lautsch, 2012). Furthermore, based on
our results regarding supervisors as role models in the context of
Theoretical Implications boundary management, future research should further investigate
Our results suggest that social learning (Bandura, 1977) and the role of other domain members, such as the role of spouses from
recovery (Meijman & Mulder, 1998) theory can be useful frame- the home domain or the role of colleagues from the work domain
works to describe organizational influence on employees’ work- (Park et al., 2011).
home segmentation behavior. Many organizations wish to imple- Research also suggests that organizational climate that is sup-
ment a work-life-friendly organizational culture. Research portive of employees’ work-life-balance affects employees’ work-
suggests that only implementing organizational-work-life benefits home segmentation behavior. For example, Thompson and col-
90 KOCH AND BINNEWIES

leagues (1999) showed that a work-life-friendly climate fosters longitudinal design in future research to confirm the current find-
employees’ use of work-life benefits (e.g., telework) and decreases ing in more detail. For example, the effect could be more precisely
work-life conflict. In terms of work-life role modeling, the impor- studied in a sample of job beginners who are having their first
tance of climate supporting employees to deal with work-life contact with organizational boundary management values.
issues is promising. It could be suggested that employees’ partic- Second, we focused exclusively on the supervisor-employee rela-
ularly imitate a role model’s behavior when this is in line with the tionship in our study, although there may be other social influences on
work-life climate. Whether supervisors feel affected by climate in employees within the organizational context. For example, Park and
terms of their own work-home segmentation behavior and which colleagues (2011) showed that a social norm within the work group
behaviors of organizational actors shape the climate would be that is also shaped by coworkers affects employees’ detachment at
important additional research questions within this context. In line home. Still, in focusing on the supervisor-employee relationship it is
with previous research, it can be suggested that work-life climate neither suggested that the supervisor is the only potential role model
affects both supervisors’ and employees’ work-home segmentation nor that the influence process occurs only in a downward direction.
behavior (Thompson et al., 1999).
However, if role modeling does occur in organizations, the
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

supervisor-employee relationship is an interaction in which this is


This document is copyrighted by the American Psychological Association or one of its allied publishers.

Practical Implications likely to be found and is therefore a good starting point (Weiss, 1978).
Third, although we could confirm a positive relationship be-
Our findings are highly important with respect to practical
tween supervisors’ work-home segmentation behavior and em-
implications. Organizations should keep in mind that because of
their status supervisors can be powerful change agents in making ployees’ shared work-life role modeling perception, we cannot
workplaces more work-life-friendly. Supervisors can act as gate- conclusively say which specific behaviors of supervisors’ work-
keepers for the availability and implementation of organizational home segmentation (e.g., working after hours, received correspon-
work-life benefits and as change agents for informal supportive dence) shape employees’ perception. As we controlled for super-
organizational cultures. Therefore, organizations may wish to en- visors’ emotional work-life support in our analyses, we ruled out
able their supervisors to set a positive example in the context of one alternative influence on work-life role modeling. Our study is
boundary management. Organizations should make sure that su- one first step to enhance the understanding of work-life role
pervisors are not expected to be always available, and have op- modeling in the context of work-home segmentation behavior and
portunities to switch off when at home. Supervisors’ high work- should stimulate further research in this field. Whereas our study
home segmentation behavior represents work-life-friendly role focused on whether supervisors’ work-home segmentation behav-
modeling, which in turn is positively related to employees’ work- ior is related to employees’ work-life role modeling perception at
home segmentation behavior. Also, work-life-friendly role mod- all, future research should continue our effort by directly asking
eling was positively related to employees’ well-being. Therefore, employees about crucial signals of supervisors’ work-home seg-
efforts to enhance supervisors’ work-home segmentation behavior mentation behavior. Also a comparison with supervisors’ own
would probably not only benefit supervisors themselves but also perspective on visible indicators would be interesting to find out
the system as a whole. For example, Burke (2001) emphasizes that how much awareness supervisors have concerning their role mod-
it is important that organizations stop taking “pride in developing eling effect. Furthermore, it has to be taken into account that
cultures where long work hours and sacrifices in nonwork lives are supervisors’ work-home segmentation is one important but prob-
seen as requirements for success and advancement” (p. 639). ably not the only modeling behavior that shapes employees’ per-
Keeping a boundary between work and home allows switching off ception of having a work-life-friendly role model. Therefore, fur-
from work demands, recovery and continued motivation in the ther modeling behaviors on the job as well as further confounding
long run (Hecht & Allen, 2009; Park et al., 2011). variables above emotional work-life support should be examined.
However, responsibility for employees’ well-being and work- Thereby, the understanding of the relationship between supervi-
life balance in a broader sense should not be exclusively attributed
sors’ work-home segmentation behavior and work-life role mod-
to supervisors. Organizations should also take additional steps to
eling can be expanded.
foster their employees’ well-being and work-life balance, for ex-
In addition, our sample consisted of employees and supervisors
ample by avoiding negative career consequences in terms of pri-
from the health and service sector. Most of the employees who
vate obligations and also by limiting organizational time demands
participated were female. Thus, there may be concerns about the
(Thompson et al., 1999).
generalizability of our study. However, additional analyses did not
reveal relations between employees’ and supervisors’ gender and
Limitations and Future Research our study’s predictor and outcome variables. Furthermore, we had
Any study such as ours has some limitations, and this must be to drop 149 employees as they could not be matched with super-
taken into account when considering the results. One limitation of visor data that introduces the risk of a systematic sample bias.
our study is the use of a cross-sectional design in which causal However, t-tests showed that employees whose supervisors had
interpretations are limited. For example, we cannot ensure from participated did not significantly differ from those employees
our study design whether there really is a downward process from whose supervisors had not participated (with the exception of
supervisors on employees or whether the effect is vice versa. disengagement). Also, our dropout is still in the range of what
Although this study provides multilevel findings of supervisors’ Nulty (2008) describes as being typically found in paper-based
behavior on work groups’ role modeling perception that are the- settings. Nevertheless, future research should replicate our find-
oretically plausible (Weiss, 1978), we recommend the use of a ings in a more representative sample to validate our findings.
SETTING A GOOD EXAMPLE 91

References Hayes, A. F. (2009). Beyond Baron and Kenny: Statistical mediation


analysis in the new millennium. Communication Monographs, 76, 408 –
Ashforth, B. E., Kreiner, G. E., & Fugate, M. (2000). All in a day’s work: 420. doi:10.1080/03637750903310360
Boundaries and micro role transitions. The Academy of Management Hecht, T. D., & Allen, N. J. (2009). A longitudinal examination of the
Review, 25, 472– 491. doi:10.2307/259305 work-nonwork boundary strength construct. Journal of Organizational
Bandura, A. (1977). Social learning theory. Englewood Cliffs: Prentice Behavior, 30, 839 – 862. doi:10.1002/job.579
Hall. Judd, C. M., & Kenney, D. A. (2010). Data analysis in social psychology:
Bass, B. M., & Avolio, B. J. (1994). Transformational leadership and Recent and recurring issues. In S. T. Fiske, D. T. Gilbert, & G. Lindzey
organizational culture. International Journal of Public Administration, (Eds.), The handbook of social psychology (5th ed., Vol. 1, pp. 115–
17, 541–554. doi:10.1080/01900699408524907 139). Hoboken, NJ: Wiley.
Binnewies, C., & Sonnentag, S. (2008). Recovery after work. Unwinding Judge, T. A., & Bono, J. E. (2000). Five-factor model of personality and
from daily job stress. In R. J. Burke & C. L. Cooper (Eds.), The long transformational leadership. Journal of Applied Psychology, 85, 751–
work hours culture - Causes, consequences and choices (pp. 275–293). 765. doi:10.1037/0021-9010.85.5.751
Bingley, UK: Emerald. Kanungo, R. N. (1982). Measurement of job and work involvement.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Bulger, C. A., Matthews, R. A., & Hoffman, M. E. (2007). Work and Journal of Applied Psychology, 67, 341–349. doi:10.1037/0021-9010.67
This document is copyrighted by the American Psychological Association or one of its allied publishers.

personal life boundary management: Boundary strength, work/personal .3.341


life balance and the segmentation-integration continuum. Journal of Kirby, E. L., & Krone, K. (2002). “The policy exists but you can’t really
Occupational Health Psychology, 12, 365–375. doi:10.1037/1076-8998 use it”: Communication and the structuration of work-family policies.
.12.4.365 Journal of Applied Communication Research, 30, 50 –77. doi:10.1080/
Burke, R. J. (2001). Workaholism in organizations: The role of organiza- 00909880216577
tional values. Personnel Review, 30, 637– 645. doi:10.1108/ Kossek, E. E., & Lautsch, B. A. (2012). Work-family boundary manage-
EUM0000000005977 ment styles in organizations: A cross-level model. Organizational Psy-
Burke, R. J. (2006). Organizational culture: A key to the success of chology Review, 2, 152–171. doi:10.1177/2041386611436264
work-life integration. In F. Jones, R. J. Burke, & M. Westьman (Eds.), Kreiner, G. E., Hollensbe, E. C., & Sheep, M. L. (2009). Balancing borders
Work-life balance. A psychological perspective (pp. 235–260). Hove & and bridges: Negotiating the work-home interface via boundary work
New York: Psychology Press. tactics. Academy of Management Journal, 52, 704 –730. doi:10.5465/
Burke, R. J., & Fiksenbaum, L. (2008). Work hours, work intensity, and AMJ.2009.43669916
work addiction: Costs and benefits. In R. J. Burke & C. L. Cooper (Eds.), Kurland, N. B., & Bailey, D. E. (1999). The advantages and challenges of
The long work hours culture. Causes, consequences and choices (1st ed., working here, there anywhere, and anytime. Organizational Dynamics,
pp. 3–36). Bingley, UK: Emerald Group Pub. 28, 53– 68. doi:10.1016/S0090-2616(00)80016-9
Chen, G., & Bliese, P. D. (2002). The role of different levels of leadership Lewis, S. (1997). ‘Family friendly’ employment policies: A route to
in predicting self- and collective efficacy: Evidence for discontinuity. changing organizational culture or playing about at the margins? Gen-
Journal of Applied Psychology, 87, 549 –556. doi:10.1037/0021-9010 der, Work and Organization, 4, 13–23. doi:10.1111/1468-0432.00020
.87.3.549 Lockwood, P. (2006). “Someone like me can be successful”: Do college
Clark, S. C. (2000). Work/family border theory: A new theory of work/ students need same-gender role models? Psychology of Women Quar-
family balance. Human Relations, 53, 747–770. doi:10.1177/ terly, 30, 36 – 46. doi:10.1111/j.1471-6402.2006.00260.x
0018726700536001 MacKinnon, D. P., Krull, J. L., & Lockwood, C. M. (2000). Equivalence
Cohen-Charash, Y., & Spector, P. E. (2001). The role of justice in orga- of the mediation, confounding, and suppression effect. Prevention Sci-
nizations: A meta-analysis. Organizational Behavior and Human Deci- ence, 1, 173–181. doi:10.1023/A:1026595011371
sion Processes, 86, 278 –321. doi:10.1006/obhd.2001.2958 Meijman, T., & Mulder, G. (1998). Psychological aspects of workload. In
Demerouti, E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001). P. J. D. Drenth, H. Thierry, & C. J. de Wolff (Eds.), Handbook of work
The job demands-resources model of burnout. Journal of Applied Psy- and organizational psychology: Vol. 2. Work psychology (2nd ed., pp.
chology, 86, 499 –512. doi:10.1037/0021-9010.86.3.499 5–33). Hove: Psychology Press.
Demerouti, E., Bakker, A. B., Vardakou, I., & Kantas Aristotelis. (2003). Muthén, B. O., & Muthén, L. K. (1998 –2010). Mplus user’s guide (6th
The convergent validity of two burnout instruments: A multitrait- ed.). Los Angeles, CA: Muthén & Muthén.
multimethod analysis. European Journal of Psychological Assessment, Nippert-Eng, C. E. (1996). Home and work: Negotiating boundaries
19, 12–23. doi:10.1027//1015-5759.19.1.12 through everyday life. Chicago, IL: University of Chicago Press. doi:
Demerouti, E., Mostert, K., & Bakker, A. B. (2010). Burnout and work 10.7208/chicago/9780226581477.001.0001
engagement: A thorough investigation of the independency of both Nulty, D. D. (2008). The adequacy of response rates to online and paper
constructs. Journal of Occupational Health Psychology, 15, 209 –222. surveys: What can be done? Assessment & Evaluation in Higher Edu-
doi:10.1037/a0019408 cation, 33, 301–314. doi:10.1080/02602930701293231
Dikkers, J. S. E., Geurts, S. A. E., Dulk, L. D., Peper, B., Taris, T. W., O’Neill, J. W., Harrison, M. M., Cleveland, J., Almeida, D., Stawski, R.,
& Kompier, M. A. J. (2007). Dimensions of work– home culture and & Crouter, A. C. (2009). Work–family climate, organizational commit-
their relations with the use of work– home arrangements and work– ment, and turnover: Multilevel contagion effects of leaders. Journal of
home interaction. Work & Stress, 21, 155–172. doi:10.1080/ Vocational Behavior, 74, 18 –29. doi:10.1016/j.jvb.2008.10.004
02678370701442190 Park, Y., Fritz, C., & Jex, S. M. (2011). Relationships between work-home
Friedman, S. D., & Lobel, S. (2003). The happy workaholic: A role model segmentation and psychological detachment from work: The role of
for employees. Academy of Management Executive, 17, 87–98. doi: communication technology use at home. Journal of Occupational
10.5465/AME.2003.10954764 Health Psychology, 16, 457– 467. doi:10.1037/a0023594
Hammer, L. B., Kossek, E. E., Yragui, N. L., Bodner, T. E., & Hanson, Podsakoff, P. M., MacKenzie, S. B., Jeong-Yeon, L., & Podsakoff, N.
G. C. (2009). Development and validation of a multidimensional mea- (2003). Common method biases in behavioral research: A critical review
sure of family supportive supervisor behaviors (FSSB). Journal of of the literature and recommended remedies. Journal of Applied Psy-
Management, 35, 837– 856. doi:10.1177/0149206308328510 chology, 88, 879 –903. doi:10.1037/0021-9010.88.5.879
92 KOCH AND BINNEWIES

Powell, G. N., & Mainiero, L. A. (1999). Managerial decision making Sonnentag, S., & Bayer, U. (2005). Switching off mentally: Predictors and
regarding alternative work arrangements. Journal of Occupational and consequences of psychological detachment from work during off-job
Organizational Psychology, 72, 41–56. doi:10.1348/096317999166482 time. Journal of Occupational Health Psychology, 10, 393– 414. doi:
Preacher, K. J., Zhang, Z., & Zyphur, M. J. (2011). Alternative methods for 10.1037/1076-8998.10.4.393
assessing mediation in multilevel data: The advantages of multilevel Sonnentag, S., & Fritz, C. (2007). The recovery experience questionnaire:
SEM. Structural Equation Modeling: A Multidisciplinary Journal, 18, Development and validation of a measure for assessing recuperation and
161–182. doi:10.1080/10705511.2011.557329 unwinding from work. Journal of Occupational Health Psychology, 12,
Preacher, K. J., Zyphur, M. J., & Zhang, Z. (2010). A general multilevel 204 –221. doi:10.1037/1076-8998.12.3.204
SEM framework for assessing multilevel mediation. Psychological Thompson, C. A., Beauvais, L. L., & Lyness, K. S. (1999). When work–
Methods, 15, 209 –233. doi:10.1037/a0020141 family benefits are not enough: The influence of work–family culture on
Raudenbush, S. W., Bryk, A., Cheong, Y. F., & Congdon, R. (2004). benefit utilization, organizational attachment, and work–family conflict.
HLM6: Hierarchical linear and nonlinear modeling. Chicago, IL: Sci- Journal of Vocational Behavior, 54, 392– 415. doi:10.1006/jvbe.1998
entific Software International. .1681
Regan, M. (1994). Beware the work/family culture shock. Personnel Urbach, D. R., & Austin, P. C. (2005). Conventional models overestimate
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Journal, 73, 35–36. the statistical significance of volume-outcome associations, compared


Scandura, T. A., & Lankau, M. J. (1997). Relationships of gender, family with multilevel models. Journal of Clinical Epidemiology, 58, 391– 400.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

responsibility and flexible work hours to organizational commitment and doi:10.1016/j.jclinepi.2004.12.001


job satisfaction. Journal of Organizational Behavior, 18, 377–391. doi: Weiss, H. M. (1978). Social learning of work values in organizations.
10.1002/(SICI)1099-1379(199707)18:4⬍377::AID-JOB807⬎3.0.CO; Journal of Applied Psychology, 63, 711–718. doi:10.1037/0021-9010.63
2-1 .6.711
Snijders, T. A. B., & Bosker, R. J. (1999). Multilevel analysis: An intro- Yaffe, T., & Kark, R. (2011). Leading by example: The case of leader
duction to basic and advanced multilevel modeling. London: Sage. OCB. Journal of Applied Psychology, 96, 806 – 826. doi:10.1037/
Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in a0022464
structural equation models. Sociological Methodology, 13, 290 –312.
doi:10.2307/270723
Sonnentag, S. (2001). Work, recovery activities, and individual well-being: Received November 29, 2013
A diary study. Journal of Occupational Health Psychology, 6, 196 –202. Revision received August 4, 2014
doi:10.1037/1076-8998.6.3.196 Accepted August 6, 2014 䡲

Showcase your work in APA’s newest database.

Make your tests available to other researchers and students; get wider recognition for your work.
“PsycTESTS is going to be an outstanding resource for psychology,” said Ronald F. Levant,
PhD. “I was among the first to provide some of my tests and was happy to do so. They will be
available for others to use—and will relieve me of the administrative tasks of providing them to
individuals.”
Visit http://www.apa.org/pubs/databases/psyctests/call-for-tests.aspx
to learn more about PsycTESTS and how you can participate.
Questions? Call 1-800-374-2722 or write to tests@apa.org.
Not since PsycARTICLES has a database been so eagerly anticipated!

You might also like