Work-Life Balance : The Good and the Bad of Boundary Management

Work-life balance is an important issue in today’s world and the different strategies used by people to manage their work and their personal life can have a great impact. Two studies were conducted (study 1: n = 117; study 2: n = 293) to examine how boundary segmentation preferences (studies 1 & 2) and boundary integration strategies (study 2) affect work-family conflict and enrichment. Results from structural equation modeling partly confirmed the hypothetical model in both studies. Study 1 showed that work-home segmentation preference negatively predicted work-family enrichment, while home-work segmentation preference negatively predicted family-work enrichment. Study 2 provided similar results, as it showed that work-home segmentation preference negatively predicted work-family enrichment. It also showed that work-home segmentation preference positively predicted work-family conflict and home-work segmentation preference positively predicted work-family enrichment, while work-life integration strategy positively predicted work-family conflict, family-work conflict, work-family enrichment and family-work enrichment. No significant relationship was found between life-work integration strategy and any of the dependent variables. Findings from these studies highlight the importance of using appropriate boundary management strategies in order to promote a better work-life balance. They also enhance current knowledge related to boundary management and work-life balance by examining relationships with work-family enrichment.


Introduction
Roles at work and at home have changed significantly in the last decades, which have led to a growing interest in how people manage to take on multiples roles (e.g., Barnett & Hyde, 2001;Edwards & Rothbard, 2000;Greenhaus & Parasuraman, 1999).Boundaries are becoming more and more permeable, in large part because of the ever-increasing number of teleworkers (Allen, Cho, & Meier, 2014).Even for those who work on-site in an office, technology gives them the flexibility-formally or informally-to answer personal e-mails, calls and text messages during working hours; on the other hand, they can answer professional e-mails, calls and text messages during the week-end and when they're on vacation (Kossek & Lautsch, 2008).It's up to the individual to find the right balance between work and personal life (Major & Germano, 2006), especially since it's now normal to expect some blending of the different roles (Allen et al., 2014;Ollier-Malaterre, Rothbard, & Berg, 2013;Olson-Buchanan & Boswell, 2006).This aspect of modern-day life can lead to conflict between those roles and with the demands that come with them (e.g., work-family conflict; Greenhaus & Beutell, 1985;Carlson, Kacmar, & Williams, 2000).However, boundary permeability can also enrich the different spheres of life (Greenhaus & Powell, 2006).For example, our work can enrich our personal life, and vice-versa.When people find the right balance between work and their personal lives, they benefit from all the roles they take on (e.g., work-family enrichment; Greenhaus & Powell, 2006).Considering the significance of work-life balance, understanding how boundary management impacts all aspects of life is becoming more and more important.Several studies have examined how boundary management can have an impact on work-family conflict (Clark, 2002;Bulger, Matthews, & Hoffman, 2007;Kinman & Jones, 2008;Kossek, Lautsch, & Eaton, 2006;Kossek, Ruderman, Braddy, & Hannum, 2012;Li, Miao, Zhao, & Lehto, 2013;Matthews & Barnes-Farrell, 2010;Olson-Buchanan & Boswell, 2006;Powell & Greenhaus, 2010), but none of them, to our knowledge, has yet to explore the impact of boundary management on work-family enrichment.management strategies related to the different spheres of life can impact work-family conflict as well as work-family enrichment.More specifically, the goals are to examine: 1) the impact of boundary segmentation preferences, 2) the impact of various boundary management strategies on work-family enrichment and work-family conflict.Results will be discussed in relation to the theories of boundary management (Ashforth, Kreiner, & Fugate, 2000), work-family conflict (Greenhaus & Beutell, 1985) and work-family enrichment (Greenhaus & Powell, 2006).

Work-Family Balance
As noted above, work-family balance has become an important issue nowadays.In fact, the shift we're seeing in today's world has also led to a change in the way we define the concept of family (Barnett & Hyde, 2011;Bourdeau & Houlfort, 2015;Grzywacz & Marks, 2000;Rothausen, 1999;Teachman, Tedrow, & Crowder, 2000).However, most research on work-life balance is focused on married workers with children (Casper, Eby, Bordeaux, Lockwood, & Burnett, 2007) and tends to leave out single adults who don't have any children.In this article, family is defined as a unit which may include couples with or without children as well as single parents, since these family structures, although they differ from the concept of traditional family, are all affected by issues related to work-life balance (Bourdeau & Houlfort, 2015).In order to fit our contemporary definition of family, to help enhance and broaden current knowledge on work-life balance and to reduce the limitations of the term "work-family", we use the terms "work-family", "work-home" and "work-life" interchangeably and view "work-family" issues as part of an important work-life relationship.
Work-family balance can be looked at in terms of work-family conflict or from a more positive point of view, namely work-family enrichment (Greenhaus & Powell, 2006).It should be noted that both of these are two-way processes: the impact can be seen both from work to family and from family to work.

Work-Family Conflict
Work-Family Conflict (WFC) and Family-Work Conflict (FWC) result from the demands of the multiple roles one has to take on (both at work and at home), which can lead to conflict (Greenhaus & Beutell, 1985).The negative impact of these conflicts is significant for the individual as well as the organization (or the family).Studies have shown that conflict was positively associated to negative outcomes, like absenteeism, burnout and decline in job satisfaction and life satisfaction (Allen, Herst, Bruck, & Sutton, 2000;Byron, 2005;Kossek & Ozeki, 1998;Kossek, Pichler, Bodner, & Hammer, 2011).According to Carlson et al. (2000), conflict includes the following components: time (e.g., "My work keeps me from my family activities more than I would like"), behavior (e.g., "The problem-solving behaviors that work for me at home does not seem to be as useful at work") and stress (e.g., "Due to all the pressures at work, sometimes when I come home I am too stressed to do things I enjoy").

Work-Family Enrichment
Despite potential conflict between the different spheres of life, it is also possible to enrich one role by using resources developed in another to meet requirements more appropriately (Greenhaus & Powell, 2006).Since work-family enrichment is a two-way process, experiences in one life domain can lead to a better quality of life in another domain, and vice-versa (Carlson, Kacmar, Wayne, & Grzywacz, 2006;Greenhaus & Powell, 2006).For example, studies have shown that enrichment is positively associated to positive impacts (e.g., affective commitment, emotional commitment towards the organization, satisfaction at work, at home and in life in general) and negatively associated to negative outcomes (e.g., negative affect and physical health problems; Bourdeau & Houlfort, 2015;McNall, Nicklin, & Masuda, 2010).According to Carlson et al. (2006), each direction in the process (WFE and FWE) has three components.WFE would include development (both on a personal and intellectual level, for example, "My involvement in my work helps me to understand different viewpoints and this helps me be a better family member"), affect (positive emotions experienced within the role, for example, "My involvement in my work puts me in a good mood and this helps me be a better family member") and capital (psychological capital acquired through the role, for example, "My involvement in my work helps me feel personally fulfilled and this helps me be a better family member"), whereas FWE would include development (both on a personal and intellectual level, for example, "My involvement in my family helps me to gain knowledge and this helps me be a better worker"), affect (positive emotions experienced within the role, for example, "My involvement in my family puts me in a good mood and this helps me be a better worker") and efficiency (developed with family management, for example, "My involvement in my family requires me to avoid wasting time at work and this helps me be a better worker").

Managing Work-Family Boundaries
Nippert-Eng (1996) suggests that in an effort to organize the different roles related to family and work, people tend to build mental and sometimes even physical barriers as a way to seperate their family and work environments, which they do using two opposite approaches: integration and segmentation.Boundaries can be spatial, temporal, cognitive, behavioral, relational or emotional, among other types (Ashforth et al., 2000;Clark, 2000;Hall & Richter, 1988;Nippert-Eng, 1996).With boundary management, people can organize and separate the demands and expectations that come with each role (Kossek, Noe, & DeMarr, 1999).For example, an individual can keep two calendars, one for work and one for his personal life (separation), or put in an office at home (integration).We should note that boundary management is also a two-way process, as are work-family conflict and enrichment (Ashforth et al., 2000;Hecht & Allen, 2009;Olson-Buchanan & Boswell, 2006), although some people are able to separate or integrate in one direction (e.g., from work to home) but not in the other (Hecht & Allen, 2009;Kossek et al., 2012).There are three different boundary management strategies: 1) separating work from family; 2) integrating work and family; 3) a hybrid approach where the individual alternates between the first two approaches (Kossek & Lautsch, 2008).Since there isn't any scale that measures the hybrid approach and since this article isn't intended to create one, this research focuses on the first two strategies, namely separation and integration of work and personal life.Boundary segmentation between these two spheres of life will be defined in terms of preference (i.e., work-home segmentation preference and home-work segmentation preference), whereas boundary integration will be defined in terms of the strategies that are used (i.e., work-life integration and life-work integration).

Impacts of Work-Family Boundary Management
According to Ashforth et al. (2000), integrating work and family in a single space-time (spatial and temporal dimensions) makes the boundaries between the two permeable, which means that work can be disrupted by family demands and tasks (and vice-versa).This can increase conflict between these two spheres.A study by Olson-Buchanan and Boswell (2006) found that boundary permeability from work to family was associated with more WFC, whereas boundary permeability from family to work was associated with less WFC.Another study showed that work boundary permeability was associated with more WFC, which wasn't the case with family boundary permeability (Clark, 2002).In addition, higher work permeability seems to be associated with more FWC, whereas higher family permeability seems to be associated with more WFC (Bulger et al., 2007;Matthews & Barnes-Farrell, 2010).Conversely, segmentation of the two spheres of life could help reduce conflict.In fact, studies have shown that real segmentation between work and family responsibilities was associated with less WFC (Kinman & Jones, 2008;Kossek et al., 2012;Powell & Greenhaus, 2010) and FWC (Kossek et al., 2006;Kossek et al., 2012) and to a better balance between work and family (Li et al., 2013).However, boundary segmentation could thwart enrichment because of the barriers put in place by the individual.To our knowledge, this hasn't been studied directly as of yet.Moreover, boundary integration could increase enrichment because of boundary permeability, which allows more exchanges between the two spheres of life.Even though Ashforth et al. (2000) state that integrating work and family doesn't necessarily mean less work-family conflict, one might wonder if this type of integration could increase work-family enrichment.As far as we know, no study has yet to examine this.

Our Research
The goal of this research was to fill this gap by examining how boundary segmentation preferences and integration strategies used by workers can contribute to conflict (both WFC and FWC) as well as enrichment (both WFE and FWE) in the work-family interface.This will build on existing knowledge on work-family balance by looking at whether mechanisms applied by workers can reduce conflict while at the same time increase enrichment in their professional and personal lives.Two studies were conducted to better define the impact of segmentation preferences (studies 1 and 2) and integration strategies (study 2) on conflict and enrichment experienced by workers.

Research Questions and Hypothetical Model
Study 1 examined how boundary segmentation preferences could help predict work-life balance in terms of conflict and enrichment.Based on results from previous studies on bidirectional work-family conflict and boundary permeability (Bulger et al., 2007;Clark, 2002;Kinman & Jones, 2008;Kossek et al., 2006;Kossek et al., 2012;Li et al., 2013;Matthews & Barnes-Farrell, 2010;Olson-Buchanan & Boswel, 2006;Powell & Greenhaus, 2010), a model (see Figure 1) is proposed where the more workers have a significant preference for segmentation between different life spheres, the less work-family conflict (WFC and FWC) and work-family enrichment (WFE and FWE) they would experience.Participants 1 (n = 173) were French-speaking workers from various fields.Most of them (n = 112) were elementary and high school teachers from several schools across the province of Quebec willing to take part in studies on their experiences related to work.This group of participants is especially relevant to our study, since teachers often have to do some work at home (preparing classes, grading student work and exams, etc.).The rest of the sample (n = 61) included workers from different fields who had been recruited via Facebook.They were recruited in order to have a more diverse group with different experiences in terms of work-life balance, which would allow better generalization of the results.
Participants from the two subsamples were invited to take part in a short study on their experiences related to work.The online software Survey Monkey was used to collect the data.As a thank you for their help, all participants were entered in a draft to win one of three family passes to a water park or a zoo or a prepaid VISA card (value of about US$100).
Those who accessed the questionnaire without answering any questions were removed from the sample (n = 22).In accordance with the premises established by Tabachnick and Fidell (2013), participants with more than 25% of missing data (n = 31) were also removed.In addition, two participants were removed because they had more than 5% of missing data following a non-random answering pattern.Missing data from the rest of the sample were replaced by the mean value, since they represented less than 5% of the data and since Little's MCAR (missing completely at random) test provided a result that was not significant, χ 2 = 242.47(df = 215, p = .10).Finally, one participant with extreme and multivariate values was identified (n = 1) and removed from the sample.The final sample (n = 117) included 87 elementary and high school teachers as well as 30 workers from different fields who were recruited via Facebook.Ninety-seven of them were women (83%), nineteen were men (16%) and one didn't specify his or her gender (0.85%).Participants were between 21 and 65 years old (mean age 40.27 years; SD = 9.73) and seventy-two of them had at least one child.

Measurements (1) Boundary segmentation preference
A modified French version of the scale developed by Kreiner (2006) was used to measure boundary segmentation preference.The initial scale from Kreiner (2006) has four items that measure only work-home segmentation, whereas the French version also includes a subscale measuring home-work segmentation preference.This subscale is based on the four initial items, which were modified to reflect the direction of home-work segmentation.An example of work-home segmentation preference would be "I prefer to keep work life at work", whereas "I prefer to keep my private life at home" would be an example of home-work segmentation preference.Participants had to point out to what extent they agreed with each statement using a Likert-type scale from 1 to 7, 1 being "Strongly disagree" and 7 being "Strongly agree".Higher scores meant a higher preference for boundary segmentation.
(2) Work-family conflict A French version of the work-family conflict scale developed by Carlson et al. (2000) was used in this study.
Two subscales with nine items each were used to measure work-family conflict (e.g., "My work keeps me from my family activities more than I would like") and family-work conflict (e.g., "Due to stress at home, I am often preoccupied with family matters at work").Participants had to point out to what extent they agreed with each statement using a Likert-type scale from 1 to 5, 1 being "Strongly disagree" and 5 being "Strongly agree".The higher the score, the more participants experienced conflict (whether work-family or family-work).
(3) Work-family enrichment The work-family enrichment scale developed by Carlson et al. (2006) and validated in French by Bourdeau and Houlfort (2015) was used during this study.Two subscales with nine items each were used to measure work-family enrichment (e.g., "My involvement in my work helps me to understand different viewpoints and this helps me be a better family member") and family-work enrichment (e.g., "My involvement in my family puts me in a good mood and this helps me be a better worker").Participants had to point out to what extent they agreed with each statement using a Likert-type scale from 1 to 5, 1 being "Strongly disagree" and 5 being "Strongly agree".The higher the score, the more participants experienced enrichment (whether work-family or family-work).
(4) Control variables Some variables, which may impact the ones under study and create a bias in the results, need to be taken into account.The type of participants (teachers from the pool of volunteers versus workers from different fields recruited via Facebook), gender and family status (having children or not) were therefore controlled.Because of the particular nature of this study, which examines work-family balance and boundary segmentation preference, gender and family status may in fact have an impact on these variables (e.g., Byron, 2005;Kossek & Ozeki, 1998;MacEwen & Barling, 1994;Martins, Eddleston, & Veiga, 2002).

Preliminary Results
Preliminary analyses were performed with SPSS v.21 (IBM Corp., 2012).The normality of each variable was assessed by screening the scores of skewness and kurtosis.All variables were normally distributed, except for family-work enrichment, for which the kurtosis score was higher than 1.The variable was transformed in order to correct the negative skew noticed on the histogram.Table 1 shows means, standard deviations, Cronbach's alphas and correlations for all variables.An independent sample t-test showed no difference between the subsamples with regards to the type of participant (teachers versus workers from other fields), but did show differences with regards to gender and having children or not.

Main Results
In order to check if gender and having children or not affected the present results, two sets of analyses were conducted: one path analysis controlling for gender and for having children or not and another path analysis where no controlled variables were included.Comparison of the models that were tested showed no difference, so the results reported here are the ones from the initial hypothetical model.To test the hypothetical model, a path analysis using MPLUS 6.12 (Muthén & Muthén, 1998-2011) was conducted.The hypothetical model included two exogenous variables (work-home segmentation preference and home-work segmentation preference) and four endogenous variables (work-family conflict, family-work conflict, work-family enrichment and family-work enrichment).It also included direct paths from work-home segmentation preference and home-work segmentation preference to work-family conflict, family-work conflict, work-family enrichment and family-work enrichment.Covariances were allowed between both directions of conflict and between both directions of enrichment.The model was tested using maximum likelihood estimation with robust standard errors (MLR estimation).The goodness-of-fit was assessed using four indices: the Comparative Fit Index (CFI), the Tucker-Lewis Index (TLI), the Root Mean Square Error of Approximation (RMSEA) and the Standardized Root Mean Square Residuals (SRMR).The model provided an excellent fit to the data (χ 2 (14) = 148.52,p < .001;CFI = 1.00;TLI = 1.00;RMSEA = .00;SRMR = .00).
Results are summarized in Figure 2.They show that work-home segmentation preference and home-work segmentation preference negatively predicted work-family enrichment and family-work enrichment, respectively, while no relationships were found with conflict variables.These non-significant findings are however in line with other studies where Kreiner's measurement was used (e.g., Kreiner, 2006;Powel & Greenhaus, 2010;Shockley & Allen, 2012).In sum, these results provide partial support for the hypothetical model.Separating work from personal life prevents positive spillover from work to home.The reverse holds true for spillover from home to work.In addition, these results did not support the hypothetical positive effect of segmentation, namely protecting workers from conflict.

Research Questions
Results from Study 1 provided partial support for the proposed model.In Study 2, we set out to build on these results in two ways.First, by replicating results from Study 1 with a larger sample and second, by adding boundary integration strategies to the hypothesized model.One of the reasons why no significant relationship was found in Study1 between boundary segmentation preferences and conflict variables might be that the individual's preference can't be applied in reality, hence the importance of adding variables for boundary integration strategies.The approach implemented by workers refers to their actual behaviors, and not only to preference.Since these behaviors don't refer to a simple preference, they could in all likelihood have a bigger impact on work-family conflict.The same model as in Study 1 was hypothesized, but the assumption that people who use a work-life boundary integration strategy would not only experience more conflict, but more enrichment as well was added (see Figure 3).Participants who accessed the survey without providing any answers for the scales used for the analysis were removed from the sample (n = 26).In accordance with the premises established by Tabachnick and Fidell (2013), participants with more than 25% of missing data (n = 12) were also removed.One univariate outlier and two multivariate outliers were also identified and removed from the sample (n = 3).The final sample (n = 293) included 132 women (45.1%) and 161 men (54.9%).The mean age was 30.75 years (SD = 9.10) and seventy participants had at least one child.

Measurements
(1) Segmentation preference, work-family conflict and work-family enrichment.
The scales used in Study 1 were also used in Study 2 to assess segmentation preference, work-family conflict and work-family enrichment among workers.
Two scales of the Work-Life Indicator (Kossek et al., 2012) were used in this study: the Nonwork Interrupting Work Behaviors scale (5 items; e.g., "I respond to personal communications (e.g., emails, texts, and phone calls) during work") and the Work Interrupting Nonwork Behaviors scale (5 items; e.g., "I regularly bring work home").The first scale was conceptualized in this study in terms of life-work integration strategy while the second was conceptualized in terms of work-life integration strategy.Participants rated each item on a five-point scale ranging from 1 (strongly disagree) to 5 (strongly agree).Higher scores meant a higher use of integration strategies.

Preliminary Results
Preliminary analyses were performed with SPSS v.21 (IBM Corp., 2012).The normality of each variable was assessed by screening the scores of skewness and kurtosis.All variables were normally distributed, except for work-home segmentation preference, for which the kurtosis score was higher than -1 (-1.01).The variable was transformed in order to correct the negative skew noticed on the histogram.Table 2 shows means, standard deviations, Cronbach's alphas and correlations for all variables.As in Study 1, an independent sample t-test did show differences with regards to gender and having children or not.

Main Results
In order to check if this affected the results in Study 2, two sets of analyses were conducted: one path analysis controlling for gender and for having children or not and another path analysis where no controlled variables were included.As in Study 1, comparison of the models that were tested showed no difference, so the results reported here are the ones from the initial hypothetical model of Study 2.
To test the hypothetical model, a path analysis using MPLUS 6.12 (Muthén & Muthén, 1998-2011) was conducted.The hypothetical model included four exogenous variables (work-home segmentation preference, home-work segmentation preference, work-life integration strategy and life-work integration strategy) and four endogenous variables (work-family conflict, family-work conflict, work-family enrichment and family-work enrichment).It also included direct paths from work-home segmentation preference and home-work segmentation preference to work-family conflict, family-work conflict, work-family enrichment and family-work enrichment.Covariances were allowed between both directions of conflict and between both directions of enrichment.The model was tested using maximum likelihood estimation with robust standard errors (MLR estimation).The goodness-of-fit was assessed using four indices: the Comparative Fit Index (CFI), the Tucker-Lewis Index (TLI), the Root Mean Square Error of Approximation (RMSEA) and the Standardized Root Mean Square Residuals (SRMR).The model provided an excellent fit to the data (χ 2 (22) = 457.87,p < .001;CFI = 1.00;TLI = 1.00;RMSEA = .00;SRMR = .00).In sum, results from Study 2 provide partial support for the hypothetical model.Separating work from personal life thwarts work-family enrichment, thus taking away any benefits of a potential positive spillover from work to family, and does not protect from conflict (WFC and FWC).However, preventing personal matters from intruding on work (family-work segmentation preference) was found to increase work-family enrichment, although it could not prevent workers from experiencing conflict.Concrete behaviors such as work-life integration strategy (work interrupting nonwork behaviors) positively predicted conflict (WFC and FWC) and enrichment (WFE and FWE).Although this type of strategy allows for positive spillover from work to home (and vice-versa), it also increases conflict between work and personal life.

Discussion
These studies were intended to fill gaps in the literature with regards to segmentation preferences and boundary integration strategies, in relation to work-family balance (bidirectional work-family conflict and enrichment).
Even though several studies had examined the impact of boundary segmentation preferences on bidirectional work-family conflict, there hadn't been any studies to our knowledge on boundary segmentation preference in relation to bidirectional work-family enrichment.Since this type of enrichment can have a positive impact on workers and organizations, it was important to determine if preferences or strategies that increase segmentation between work and personal life could impact enrichment negatively while helping to avoid conflict.

Boundary Segmentation Preferences and Work-Family Enrichment
Results from both studies show that boundary segmentation preferences can thwart work-family enrichment.Indeed, Study 1 shows that work-home segmentation preference negatively predicted work-family enrichment while home-work segmentation preference negatively predicted family-work enrichment.In Study 2, results partly replicated these findings as work-home segmentation preference negatively predicted work-family enrichment, but unlike Study 1, no relationship was found between home-work segmentation preference and family-work enrichment.Instead, home-work segmentation preference positively predicted work-family enrichment.To our knowledge, these are the first studies to examine the relationship between segmentation preferences and enrichment.According to the work-family enrichment theory (Carlson et al., 2006;Greenhaus & Powell, 2006), family life can be enhanced, or enriched, by integrating abilities developed at work and using them at home, and vice-versa.Results from both studies show that people who prefer to keep work and personal life separate (e.g., not taking any business calls while at home) would deprive themselves from this enrichment, which could in fact promote integration of the different roles and spheres of life.These results add a bit of perspective, since current thinking is that a clear separation between work and personal life is the best way to find balance between the two.However, results from Study 2 suggest that segmentation isn't always negative.
Preventing personal matters from intruding on work would allow workers to benefit from positive spillover of WFE.Perhaps this type of boundary management could increase concentration at work, and thus facilitate experiences of positive emotions and the development of abilities that are easier to transfer to personal life.More research is needed on this subject to fully understand this interesting finding.

Boundary Segmentation Preferences and Work-Family Conflict
Although findings from Study 1 did not show any relationship between segmentation preferences and conflicts, results from Study 2 showed that preference for work-family segmentation positively predicted WFC, whereas a marginal positive relationship was found between preference for family-work segmentation and FWC.To our knowledge, this is the first time that significant relationships have been found between segmentation preferences and conflicts, thus confirming the need to pursue this line of research.These findings suggest that separating work and personal life by making sure the two don't overlap does not necessarily prevent conflict.In other words, preventing conflict might not be as simple as building impermeable boundaries between the different spheres of life.It is however important to note that the concept of segmentation preferences does not assess a person's actual behaviors.Our work and family environments do not always allow us to use our preferred mode of segmentation.We can't always keep family matters from intruding on our work.For instance, if we get a call informing us that our child is sick and needs to be picked up at school and brought home, it will certainly have an impact on our work.Despite our preference for segmentation, our environment and circumstances force us to integrate both our work and our personal life.

Boundary Integration Strategies and Work-Family Enrichment
Previous research has shown that work-life integration strategy (work interrupting nonwork behaviors) and life-work integration strategy (nonwork interrupting work behaviors) were associated to more positive family-to-work spillover (Kossek et al., 2012).Results from Study 2 support those findings and show that work-life integration strategy positively predicted WFE and FWE, whereas a marginal positive relationship was found between life-work integration strategy and WFE.This suggests that these strategies allow positive spillover from work to home (and vice-versa) and that integrating work and personal life can truly be beneficial.These findings deepen the work-family enrichment theory (Carlson et al., 2006;Greenhaus & Powell, 2006) by showing that real integration strategies, unlike segmentation preferences, allow people to efficiently use abilities and skills developed at work and at home to enhance their professional and personal lives.

Boundary Integration Strategies and Work-Family Conflict
Results from Study 2 show that work-life integration strategy (work interrupting nonwork behaviors) positively predicted WFC and FWC but no relationship was found between life-work integration strategy (nonwork interrupting work behaviors) and any of the conflict variables.These findings are somewhat similar to those of Kossek et al. (2012).Work-life integration strategy was positively associated with more conflicts (WFC and FWC), whereas life-work integration strategy was positively associated with more FWC.This discrepancy could mean that the nature of the sphere that spills over to the other is important: people would experience more conflict when their personal life is disrupted by work than the other way around.Qualitative research that would investigate "how" conflict is experienced when work disrupts nonwork activities (and vice-versa) would certainly help us better understand the phenomenological experience related to these disruptions, and why (or when) some disruptions seem to generate more conflict than others.

Theoretical Implications
Adding the scale developed by Carlson et al. (2006) for bidirectional work-family enrichment to a study on boundary management allows to combine the two theories into a single model, which hadn't been done yet.To our knowledge, this is also the first time that significant relationships have been found between segmentation preferences and conflicts.Interestingly, segmentation strategies seem to exacerbate workers' experience of conflict between their work and their personal lives.These findings indicate that this line of research is worth pursuing in order to fully understand under which circumstances these strategies might be useful and when they become harmful.Overall, findings in this paper highlight the need to analyze work-life balance as a whole, by examining both conflict and enrichment as well as segmentation preferences and integration strategies in order to have a more complete picture of the dynamics behind work-life balance.Future research should also investigate the perception of boundary control in order to examine what prevents people from using their actual preferences as a boundary management strategy and the impact of being able to do so or not.

Limitations and Future Studies
Despite the positives from our studies, some limitations should be pointed out.First, the size of the sample in Study 1 could reduce the statistical power of some analyses and conceal some impacts which could be significant in a bigger sample, although the size of the sample from Study 2 and partial replication of relationships identified in Study 1 could somewhat offset this limitation.Another limitation is the correlational nature of the two studies, which prevents any determination of causality and directionality in the relationships established between the work-family interface variables and the ones related to boundary management.Finally, there's the fact that all the data are derived from self-reported information.This approach can alter perception with regards to conflict (or the absence thereof).It could help to collect information from third parties as well (e.g., spouse) in order to triangulate the data.
Future studies should use prospective designs in order to make a determination with regards to long-term effects of segmentation preferences and integration strategies on work-family interface variables.It would also be a good idea to examine the hybrid approach for boundary management (alternating between integration and segmentation; Kossek & Lautsch, 2008, 2012).We would expect people using a hybrid approach to experience less conflict and more enrichment, and that they would alternate between segmentation and integration in a relatively strategic fashion by adapting to the ever-changing demands and resources of each sphere of life in order to reduce conflict and increase enrichment.In this context, it would be interesting to study a specific group of people particularly impacted by boundary integration and segmentation, like teleworkers.We would expect these relationships to be stronger for teleworkers than for other workers, since teleworking involves integrating spatial and temporal boundaries, which makes them much more permeable.

Conclusion
The main objective of this research was to examine how segmentation preferences and boundary integration strategies between different spheres of life could contribute to a better work-family balance.It provided a better overall picture of the dynamics behind work-family balance.Main findings suggest that an increased segmentation or an increased integration of the different spheres of life does not mean a better balance.Although both of these approaches can promote enrichment from one sphere to the other, they also increase the risk of conflict.Work-family balance thus remains an important issue in modern society and affects most workers, who have to juggle with a professional life and a personal life that are both very demanding.

Figure
Figure 1.Theoretical model

Figure
Figure 3. Theoretical model

Figure 4 .
As was the case in Study 1, they show that work-home segmentation preference negatively predicted work-family enrichment.But, unlike Study 1, no relationship was found between home-work segmentation and family-work enrichment.However, home-work segmentation preference positively predicted work-family enrichment and positively predicted family-work conflict, although only marginally.Also, work-home segmentation preference positively predicted work-family conflict.As for work-life integration strategy, it positively predicted work-family conflict, family-work conflict, work-family enrichment and family-work enrichment.On the other hand, life-work integration strategy positively predicted work-family enrichment, but only marginally.

Table 1 .
Study 1: Means, standard deviations and correlations between variables

Table 2 .
Study 2: Means, standard deviations and correlations between variables