Please, Do Not Interrupt Me: Work–Family Balance and Segmentation Behavior as Mediators of Boundary Violations and Teleworkers’ Burnout and Flourishing

: The lockdown, in the COVID-19 pandemic, is considered an external crisis that evokes innumerous changes in individuals lives. One of the changes is the work and family dynamics. Based on boundary theory we examine the mediated role of work and family balance and boundary segmentation behavior in the relationship between boundary violations and teleworkers’ stress and well-being. However, because women and men live their work and family differently, gender may condition the way teleworkers lead with boundary violations and boundary segmentation. Hypotheses were tested through moderated mediation modeling using data collected of 456 teleworkers during lockdown. In line with our expectations, teleworkers who have suffered most boundary violations were those with least boundary segmentation behaviors and with least work-family balance which, in turn was related to higher burnout and lower ﬂourishing. Furthermore, gender was found to moderate the relationship between boundary violations from work-to-family and segmentation behavior in the same direction and this relationship was stronger for females than for males. We discuss implications for future research and for managing teleworkers, creating sustainability, both during a crise and stable days.


Introduction
The outbreak of the COVID-19 pandemic and consequent social distancing measures forced a shift toward mass telework in most countries all around the world. Telework is considered a flexible arrangement to promote a healthy work and family relationship, allowing an employee to perform work away from the office for a part of the work week while keeping in contact via information and communication technologies (ICT) [1]. Despite this practice being a flexible arrangement, its effects on the work and family relationship remain inconclusive [1][2][3].
However, the conditions of telework during the lockdown period may not apply to a post-outbreak teleworking condition. In fact, not only were employees forced to work from home overnight, including situations where dual earner couples began teleworking at the same time, but also children were in the same space as a result of the closure of schools and the implementation of "distance learning" [4]. Hence, most employees were suddenly not only confronted with the challenge of working under different conditions but were simultaneously obliged to give support to their children. Furthermore, some employees are likely to have faced problems related to a lack of space, i.e., not having a private room in which to work. As such, instead of incorporating its pre-outbreak principles of being a family friendly measure, teleworking during lockdown may, in fact, have had a negative impact on families. It is, therefore, not surprising that both the lockdown and telework obligations may have had a negative impact on employees' well-being [5,6]. goals (SDG) advocated by the United Nations [29], which, as already highlighted, are more important than ever for the recovery of the impact of COVID-19. Firstly, as already explained, this study will contribute to the specific socio-occupational health aim-"health and well-being" (SDG 3). Secondly, by exploring the moderating role of gender, it will contribute to the goal of gender equality (SDG 5) by seeking to bring evidence, which, in turn, will give rise to intervention in the gender role domain. Thirdly, the United Nations [29] also highlights the "employment, decent work for all and social protection" aim (SGD 8). Telework and the boundary violations that may arise, thereof, may provide further understanding of the limits to be established in these working conditions, thus contributing to the promotion of decent work. Overall, it is the aim of this study to provide evidence to support the promotion of inclusive and sustainable working conditions that ensure the well-being of workers.

Boundary Violations, Burnout and Flourishing: The Mediating Role of Boundary Segmentation Behavior
Telework, which was formerly regarded as a voluntary practice where people choose to work remotely at their own discretion [30], became an imposition with the outbreak of COVID-19. Thus, when implemented during a lockdown period, telework must be understood in light of this very specific context [30]. An important factor to consider in such a setting is the exposure of employees to more interruption, be it from work or from the family, referred to in the literature as boundary violations. In fact, interruptions from familyto-work may increase as a result of family members sharing the same space during the workday [10]. In fact, telework during lockdown fosters more work-to-family interruptions since technologies enable swift communication with just a click and often with no time limit. For instance, some employees under normal circumstances would find the interruption of a family dinner or their personal space by work intolerable; however, during lockdown, these situations became normalized as employees felt the obligation to compensate for their physical absence from the workspace [5,6]. Likewise, during lockdown, family members naturally tend to interrupt each other during working hours, since all the members are in the same space and their coexistence promotes more communication. Additionally, family household duties also increased over this period [6], possibly exposing employees to more unplanned interruptions during the work day. The boundary theory [12] posits that individuals establish boundaries as a means to simplify and order their environment, and this phenomenon is described by Zerubavel [31] as the creation of "mental fences". Thus, boundary violations emerge as a challenge to these "mental fences" that maintain the boundaries and may therefore undermine the well-being of employees.
More specifically, well-being can be understood through ill-being, e.g., burnout, but also through well-being, e.g., flourishing. Burnout is defined as "a state of exhaustion in which one is cynical about the value of one's occupation and doubtful of one's capacity to perform" ( [32], p. 20). In a situation of boundary violation, individuals need to apply transitions from one role to the other role more frequently. This phenomenon is referred to by Louis and Sutton ( [33], p. 55) as "switching cognitive gears", and these exchanges can generate more cognitive and emotional strain (i.e., burnout) by requiring more mental effort to reconnect with the interrupted tasks.
From a positive perspective, flourishing reflects psychological well-being, which "is thought to represent optimal human functioning" (p. 251), including the meaning and purpose, supportive and rewarding relationships, engagement and interest, contribution to others' wellbeing, competency, self-acceptance, optimism and feeling respected dimensions [34]. However, exposure to boundary violations can decrease the flourishing state, as an interruption may initially be a sign of disrespect for the family and/or professional role, affecting individuals' role identity, i.e., socially constructed definitions of self-in-role consisting of core or central features [25]. Furthermore, in order to understand the negative effect that boundary violations may have on flourishing, it is important to understand what Ashforth et al. [25], in line with the explanation of role identities, refer to as "contrast", i.e., the greater the contrast between professional role identities and family, the greater the magnitude of transition from one to another. For example, decision making, self-reliance and stability are required of a telecommuting supervisor; however, this professional may have to interrupt work in order to fulfil a father/mother role where he/she is expected to be more affectionate and warm. This duality may lead individuals to feel that they are not corresponding to both roles and, therefore, not offering the support expected of them, thus feeling that they are neither working optimally at a professional level nor in the family.
Furthermore, the boundary theory [11,12] and work-family border theory [13] highlight the importance of boundaries in helping to understand the impact of boundary violations on well-being. More specifically, a boundary is a line of differentiation between domains that can be physical, temporal and psychological [15]. Boundaries are influenced by flexibility, i.e., the extent to which spatial and temporal are pliable, and permeability, i.e., the extent to which a person physically located in one domain may be psychologically or behaviorally involved in another domain [15]. Hence, when boundaries have greater flexibility and permeability, blending occurs. Thus, segmentors have less flexible and less permeable boundaries, while integrators have more flexible and more permeable boundaries. In this vein, boundary management enactment represents the degree to which individuals actually keep work and family domains separate as part of an active effort to manage work and nonwork roles [15]. For instance, to adopt a segmented strategy is to attribute distinct schedules to work and family activities, and to adopt an integration strategy is to maintain work and family within the same schedule [12,15]. In more concrete terms, segmentation is related to less work-family conflict and has positive effects on well-being and health [15]. As advanced by Laustsch, Kossek and Eaton [35], the segmentation of telework is the best strategy to prevent interruptions. In this regard, Ashford et al. [25] argue that high segmentation renders the transition more difficult since this obliges a processual response from employees. Thus, boundary violations are expected to have a negative impact on segmentation. Taken together, segmentation behavior in a telework and lockdown situation is likely to induce greater well-being (less burnout and more flourishing); however, in turn, this work context fosters more boundary violations that are detrimental to segmentation. Hence, the following hypothesis was established: Hypothesis 1 (H1). Boundary segmentation behavior mediates the relationship between boundary violations and well-being (i.e., burnout and flourishing)

1.
Boundary segmentation behavior from work to family mediates the relationship between boundary violations from work to family and burnout.

2.
Boundary segmentation behavior from family to work mediates the relationship between boundary violations from family to work and burnout.

3.
Boundary segmentation behavior from work to family mediates the relationship between boundary violations from work to family and flourishing.

4.
Boundary segmentation behavior from family to work mediates the relationship between boundary violations from family to work and flourishing.

Boundary Violations, Burnout and Flourishing: The Mediating Role of Work-Family Balance
Work-family balance is considered an important and ongoing issue for organizations, particularly in times of COVID-19, during which telework has been imposed with negative consequences for the work-family relationship. For instance, the study of Schieman, Badaway, Milkie and Bierman [36] with a sample of Canadian workers showed that a pandemic situation increased work-family conflict, a negative antecedent of work-family balance. In the same vein, Vazari, Casper, Wayne and Matthews [37], having collected data preand post-COVID-19, found that, in the context of the pandemic, employees' work-family interference increased.
Work-family balance is defined as high effectiveness and satisfaction across multiple roles, including "effectiveness (good functioning, being successful and productive) and/or positive affect (satisfying, happy, healthy) in both work and family roles" ( [20], p. 173).
Individuals may be career-and family focused or career (or family)-focused and find a balance when they feel effective and satisfied in a role (or roles) that are of highest priority [6]. Consequently, the aforementioned boundary violations may be understood as an invasion of the priority individuals are placing on a specific role, thus resulting in less balance.
On the other hand, by integrating an overall inter-role valuation of congruity between the work and family domains [38] work-family balance contributes to reducing burnout and increasing flourishing. Regarding burnout, individuals who establish a balance between their work and family domains are deemed to be in harmony [39] and, therefore, do not experience ill-being. On the contrary, this harmony is part of their state of optimal functioning and the feeling that they are living a meaningful life since they manage to balance the roles that give them identity; in other words, they are in a state of flourishing.
In this regard, the literature review of Sirgy and Lee [40] on the antecedents and consequences of work-family balance shows that the latter has low levels of burnout and positive effects on well-being variables, such as, for example, life satisfaction, high family performance, and high marital satisfaction.
Hence, the following hypothesis was established: Hypothesis 2 (H2). Work-family balance mediates the relationship between boundary violations and well-being (i.e., burnout and flourishing)

5.
Work-family balance mediates the relationship between boundary violations from work to family and burnout. 6.
Work-family balance mediates the relationship between boundary violations from family to work and burnout. 7.
Work-family balance mediates the relationship between boundary violations from work to family and flourishing. 8.
Work-family balance mediates the relationship between boundary violations from family to work and flourishing.

Boundary Violations, Segmentation Behavior and Work-Family Balance: The Moderating Role of Gender
The relationship between work and family may differ for males and females [24] therefore, gender may condition these assumed mediated relationships. In fact, role theory suggests that individuals play a variety of roles on a daily basis. These roles may be inherently incompatible, as they comprise various time and behavioral expectations [27]. Generally, males tend to consider their predominant professional role in relation to their identity, thus underestimating the roles associated with their family role [18]. On the other hand, family responsibilities continue to be predominantly attributed to females [41]. Hence, females tend to place more value on organizational strategies and resources, in particular work-family balance practices, as they facilitate a balance between these two areas [23]. Conversely, males have a greater tendency to separate their professional and family roles and to attach less importance and value to the possibility of balancing these two domains [24].
In fact, the idea emerges in many debates concerning work-family issues that managing the work-family interface is more challenging for females than males. Furthermore, recently, Alon, Doepke, Olmstead-Rumsey and Tertilt [22] demonstrated highly alarming results regarding gender inequality during the COVID-19 pandemic. Some factors suggest that this pandemic will have a disproportionate negative effect on females and their employment opportunities: firstly, the fall in employment related to social distancing measures has had a considerable impact on sectors with high female employment shares; secondly, the closure of schools and daycare centers has hugely increased child care needs, which has a particularly significant impact on working mothers; and finally, single mothers, who are often in a disadvantaged economic position to begin with, are likely to take the biggest hit.
Therefore, it is the aim of this study to explore the role of gender in the pattern of associations between boundary violations and segmentation and WFB in the two directions: family-to-work and work-to-family. Hence, it is expected that: Hypothesis 3 (H3). The relationship between boundary violations, segmentation behavior in both directions-family-to-work and work-to-family-and WFB will be moderated by gender:

9.
Boundary violations from work to family will affect segmentation behavior from work-to family, mostly for females than for males; 10. Boundary violations from work to family will affect WFB, mostly for females than for males; 11. Boundary violations from family to work will affect segmentation behavior from family to work, mostly for males than for females; 12. Boundary violations from family to work will affect WFB, mostly for males than for females; Our theoretical model is represented in Figure 1a,b. secondly, the closure of schools and daycare centers has hugely increased child care needs, which has a particularly significant impact on working mothers; and finally, single mothers, who are often in a disadvantaged economic position to begin with, are likely to take the biggest hit. Therefore, it is the aim of this study to explore the role of gender in the pattern of associations between boundary violations and segmentation and WFB in the two directions: family-to-work and work-to-family. Hence, it is expected that: Hypothesis 3 (H3). The relationship between boundary violations, segmentation behavior in both directions-family-to-work and work-to-family-and WFB will be moderated by gender: 9. Boundary violations from work to family will affect segmentation behavior from work-to family, mostly for females than for males; 10. Boundary violations from work to family will affect WFB, mostly for females than for males; 11. Boundary violations from family to work will affect segmentation behavior from family to work, mostly for males than for females; 12. Boundary violations from family to work will affect WFB, mostly for males than for females; Our theoretical model is represented in Figure 1a

Procedure and Sample
The data collection for this study was carried out as part of a research project conducted within the scope of a project entitled "Work-Family boundary dynamics in nontraditional jobs" funded by Portuguese Science Foundation (PTDC/PSI-GER/32367/2017). To such end, associated companies were contacted by project researchers to participate in the study. The employees of the companies that are in teleworking who agreed to participate were notified by HR of the objectives of the study and were invited to take part in the study. Through the SurveyMonkey platform, a link was generated, which directed participants to an online survey. Finally, the employees were informed that their participation was voluntary, confidential and anonymous.
This study was released by the HR department of several companies in the service area, together with the employees who, during the 1st lockdown, had shifted to a full-time teleworking arrangement. The sample represents around 59% of the contacted employees. The majority of the sample work in consultancy services (53.3%) and in health services (30%). Regarding gender, the majority of the sample (73.5%) were females, which is faithful to the reality of the distribution of men and women according to the Portuguese national statistical institute by sector, with women, compared to men, having increased their expression in tertiary sector jobs in recent years (INE, 2020) [42]. Further, 53.3% were married or in a stable relationship and 50% without children.

Measures Boundary Violations
Boundary violations were measured with 3 items for each direction (i.e., familyto-work and work-to-family) taken from Hunter, Cark and Carlson [43]. In particular, respondents were asked to indicate the extent to which they agreed with each of four items representing boundary violations from family to work (e.g., " . . . my personal/family life has violated the work-family boundary more than I would like.") and boundary segmentation behavior from work to family (e.g., " . . . my work has violated the familywork boundary more than I would like."). The response options for boundary violations were presented on a Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree), with higher scores reflecting a higher degree of violations. The reliability estimate for boundary violations from work to family was 0.91 and for boundary violations from family-to-work, 0.84.
Boundary segmentation behavior: Boundary segmentation behavior was measured with 4 items for each direction (i.e., family-to-work and work-to-family) taken from Powell and Greenhaus [18]. In particular, respondents were asked to indicate the extent to which they agreed with each of four items representing enacted boundary segmentation behavior from work to family (e.g., "I only think about work when I'm working") and boundary segmentation behavior from family-to-work (e.g., "I leave my personal life for moments of free time."). The response options for enacted boundary segmentation behavior were presented on a Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree), with higher scores reflecting a higher degree of segmentation behaviors. The reliability estimate for boundary segmentation behavior from work to family was 0.78 and for boundary segmentation behavior from family-to-work, 0.67.
Work-family balance: WFB was measured with 4 items based on measures previously used by Allen and Kiburz [44] and Greenhaus, Ziegert and Allen [45]. An example of an item is: "I balance my work and family responsibilities so that one does not upset the other." The response options for the work-family balance scale were presented on a Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree), with higher scores reflecting a higher degree of self-rated work-family balance. The reliability estimate was 0.86.
Burnout: Burnout was measured using the Shirom-Melamed Burnout Measure (SMBM) [46] that contains 4 items to measure physical exhaustion (e.g., "I feel tired"), 4 items for emotional exhaustion (e.g., "I feel unable to have a good relationship with others (e.g., co-workers, boss, customers, etc.)") and 4 items to measure mental exhaustion (e.g., "I find it difficult to concentrate"). This scale was previously validated for the Portuguese population [47]. The response options for the SMBQ 7-point Likert scale range from 1 (hardly ever) to 7 (almost always), with higher scores reflecting a higher degree of self-rated burnout. The reliability estimate was 0.96.  [34], to which participants responded using a five-point Likert scale (1 = Strongly disagree to 5 = Strongly agree). This instrument consists of eight items (e.g., "I lead a life with purpose and meaning"), with good internal consistency (α = 0.81).
Control Variables: Previous studies show that marital status and having/not having children may influence both the work-family relationship and well-being [49][50][51]. Thus, these variables were controlled.

Statistical Analyses
First, due to the fact that all the measures were assessed as self-reports, a confirmatory factor analysis was conducted to examine whether the measures indeed represented different constructs. Confirmatory factor analysis (CFA) [52] with structural equation modeling methods were implemented with Mplus 7.2 [53]. The maximum likelihood estimation provides the well-known global fit statistics for structural equation modeling methods: comparative fit index (CFI; satisfactory values of 0.90 and above), Tucker-Lewis index (TLI; satisfactory values of 0.90 and above) and root mean squared error of approximation (RMSEA; satisfactory values below 0.08) [54]. In general, our theoretical model (boundary violations from work to family; boundary violations from family to work; boundary segmentation behavior from work to family; boundary segmentation behavior from family to work; work-family balance; burnout and flourishing) was compared with a one-factor model. Second, to analyze the moderated mediation relationships, model 7 of the computational tool process, developed by Hayes [55], was used. Considering all the relationships established in our hypotheses, it was possible to establish eight distinct models to analyze the hypotheses under study. Variables were mean-centered.

Confirmatory Factor Analysis
Our seven-factor theoretical model revealed an appropriate fit to the observed levels (χ 2 (670) = 1398.15, p < 0.01, CFI = 0.93; TLI = 0.93; RMSEA = 0.05). This measurement model was then compared to the one-factor model, which presupposes the saturation of all the items in a single latent variable that revealed a poorer fit (χ 2 (703) = 5767.67, p < 0.01, CFI = 0.53; TLI = 0.50; RMSEA = 0.13). Thus, the theoretical model provided the best fit for the data, and the difference from the alternative model was significant ∆χ 2 (33) = 4369.52, p < 0.01.
Mean values, standard deviation and correlations between the variables under study are described in Table 1.

Hypothesis Testing
Regarding H1 (a), the relationship between boundary violations from work to family with boundary segmentation behavior from work to family was found to be negative and significant (β = −0.31, p < 0.001). However, the relationship between boundary segmentation behavior from work to family with burnout was nonsignificant. Furthermore, the relationship between boundary violations from work to family and burnout was positive and significant (β = 0.62, p < 0.001). Taking all the results together, H1 (a) was refuted.
Concerning H1 (b), the relationship between boundary violations from family to work and boundary segmentation behavior from family to work, as expected, they were found to be negative and significant (β = −0.32, p < 0.001), and the relationship between segmentation behavior from family to work and burnout was also negative and significant (β = −0.37, p < 0.001). Moreover, the indirect effects were significant (for both males and females). Thus, H1 (b) was supported, but only partially, as the relationship between boundary violations from family to work, and burnout was positive and significant (β = 0.57, p < 0.001).
Regarding H1 (c), the relationship between boundary violations from work to family and boundary segmentation behavior from work to family were found to be negative and significant (β = −0.32, p < 0.001), but the relationship between boundary segmentation behavior from work to family and flourishing was nonsignificant. Thus, H1 (c) was refuted. However, it should be noted that boundary violations from work to family presented a negative and significant relationship with flourishing (β = −0.11, p < 0.001).
As for H1 (d), the relationship between boundary violations from family to work and boundary segmentation behavior from family to work were found to be negative and significant (β = −0.32, p < 0.001), and the relationship between boundary segmentation behavior from family to work and flourishing was positive and significant (β = 0.17, p < 0.001). Moreover, the indirect effects were significant (for both males and females). Thus, H1 (d) was supported, but only partially, as boundary violations from family to work, and flourishing also presented a negative and significant relationship (β = −0.05, p < 0.05).
Regarding H2 (a), the relationship between boundary violations from work to family and work-family balance was found to be negative and significant (β = −0.64, p < 0.001), and the relationship between work-family balance and burnout was also negative and significant (β = −0.47, p < 0.001) Furthermore, the relationship between boundary violations from work to family and burnout was positive and significant (β = 0.32, p < 0.001), and the indirect effects (for both males and females) were significant. Taking the results together, H2 (a) was partially supported.
Concerning H2 (b), the relationship between boundary violations from family to work and work-family balance was found, as expected, to be negative and significant (β = −0.68, p < 0.001), and the relationship between work-family balance and burnout was negative and significant (β = −0.48, p < 0.001). Moreover, the indirect effects were significant (from both males and females). Thus, H2 (b) was supported, but only partially, as the relationship between boundary violations from family to work, and burnout was positive and significant (β = 0.36, p < 0.001).
As for H2 (c), the relationship between boundary violations from work to family and work-family balance was found to be negative and significant (β = −0.65, p < 0.001), and the relationship between work-family balance and flourishing was positive and significant (β = 0.18, p < 0.001). Additionally, the indirect effects were significant (for both males and females). Thus, H (c) was totally supported.
As far as H2 (d) is concerned, the relationship between boundary violations from family to work and work-family balance was found to be negative and significant (β = −0.70, p < 0.001), and the relationship between work-family balance and flourishing was positive and significant (β = 0.18, p < 0.001). Additionally, the indirect effects were significant (for both males and females). Thus, H2 (d) was totally supported.
Finally, regarding our moderation hypothesis (H3), the interaction between boundary violations and boundary segmentation behavior was observed to be moderated by gender only in the work-to-family direction. Thus, our H3 (c) and H3 (d) were refuted, as they focus on the family-to-work direction.
Furthermore, gender moderation was not observed between boundary violations (in both directions) and work-family balance; hence, our H3 (b) and H3 (d) were refuted.
As for H3 (a), the hypothesis was supported, as may be seen in the Figure 2 in a situation of low violations from work to family, females were observed to adopt more segmentation behaviors compared to males. However, in a situation of higher violations from work-to-family, females were more negatively affected as their segmentation behavior from work-to-family decreased compared to men. positive and significant (β = 0.18, p < 0.001). Additionally, the indirect effects were significant (for both males and females). Thus, H2 (d) was totally supported.
Finally, regarding our moderation hypothesis (H3), the interaction between boundary violations and boundary segmentation behavior was observed to be moderated by gender only in the work-to-family direction. Thus, our H3 (c) and H3 (d) were refuted, as they focus on the family-to-work direction.
Furthermore, gender moderation was not observed between boundary violations (in both directions) and work-family balance; hence, our H3 (b) and H3 (d) were refuted.
As for H3 (a), the hypothesis was supported, as may be seen in the Figure 2 in a situation of low violations from work to family, females were observed to adopt more segmentation behaviors compared to males. However, in a situation of higher violations from work-to-family, females were more negatively affected as their segmentation behavior from work-to-family decreased compared to men. Regarding the control variables, it should be noted that having children was positively and significantly related to burnout and negatively and significantly related to segmentation behavior in the two directions under study in the different models.

Discussion
In the present study, the aim was to test work-family segmentation behavior and work-family balance as underlying mechanisms to explain the relationship between boundary violations and well-being (i.e., burnout and flourishing) in a sample of teleworkers during lockdown. Furthermore, it sought to test the moderating role of gender in the relationship between boundary violations and segmentation behavior/work-family balance. First, support was observed for mediation leading boundary violations from family to work to well-being (i.e., burnout and flourishing) mediated by segmentation behavior from family to work. Second, support was found for mediation leading boundary violations from both directions to well-being (i.e., burnout and flourishing) mediated by work-family balance. Third, gender was found to moderate the relationship between boundary violations from work to family and segmentation behavior from work to family, and this relationship was stronger for females than for males. Regarding the control variables, it should be noted that having children was positively and significantly related to burnout and negatively and significantly related to segmentation behavior in the two directions under study in the different models.

Discussion
In the present study, the aim was to test work-family segmentation behavior and workfamily balance as underlying mechanisms to explain the relationship between boundary violations and well-being (i.e., burnout and flourishing) in a sample of teleworkers during lockdown. Furthermore, it sought to test the moderating role of gender in the relationship between boundary violations and segmentation behavior/work-family balance. First, support was observed for mediation leading boundary violations from family to work to well-being (i.e., burnout and flourishing) mediated by segmentation behavior from family to work. Second, support was found for mediation leading boundary violations from both directions to well-being (i.e., burnout and flourishing) mediated by work-family balance. Third, gender was found to moderate the relationship between boundary violations from work to family and segmentation behavior from work to family, and this relationship was stronger for females than for males.

Boundary Violations (for Both Directions) Are Detrimental to Segmentation Behavior, Work-Family Balance and to Well-Being
As hypothesized, our study indicated that boundary violations (for both directions) are detrimental to the segmentation and work-family balance of the teleworkers in lockdown. This result is in line with the boundary theory [11,12] and the work-family border theory [13], which stress the importance of delineating borders, such as "mental fences" ( [31], p. 2) for individuals, to such an extent that, when they are subject to violations, segmentation and balance are impaired. Furthermore, boundary violations revealed a direct and positive relationship with burnout and a negative relationship with flourishing, pointing to their detrimental effect for well-being. In fact, this result shows that violations bring a burden of effort that leads to burnout and, in addition, affect role identity [11]. It is also in line with the idea that the "contrast" requirement [11] in role switching does not allow teleworkers to function optimally.

Segmentation
Behavior from Work to Family Is not Significantly Related to Well-Being, While Segmentation Behavior from Family to Work Fosters Well-Being As far as the relationship between boundary management by segmentation behavior and well-being is concerned, the results were antagonistic. On the one hand, when teleworkers successfully establish the limits of their personal affairs when they are at work, this helps them to prevent burnout and maintain their optimal functioning, in line with what has been advanced in previous studies [11,35]. However, on the other hand, leaving work matters aside when they are enjoying their free time (i.e., segmentation behavior from work to family) has no impact on well-being. This result appears to be an indicator of the specific time of collection of this sample, namely during lockdown. With the possibility of other family members working simultaneously in the same space, the importance of establishing clear boundaries to prevent invasion of the family in the work domain is emphasized more than the other way around. In addition, in the situation of a sudden shift to telework, many employees may feel the need to prove that they are available at all times and that they are able to adapt to the challenge of teleworking and its respective demands. Additionally, as stated by Hodder [56] and Kiffin et al. [57], demands are intensified in teleworking situations, and as COVID-19 brought job insecurity and financial strain at the same time, workers felt an obligation to respond to this intensification and, therefore, make themselves available even in times of crisis, leisure and family periods.

Work-Family Balance Mediates the Relationship between Boundary Violations and Well-Being
As predicted in our hypotheses, work-family balance acts as a mechanism to explain the relationship between boundary violations and well-being. As highlighted by the workfamily relationship literature [40], this mediating role of work-family balance reinforces its importance as an indicator of well-being, thus preventing burnout and promoting flourishing, while additionally reinforcing the aforementioned relationship between boundary violations and work-family balance. The results of the present study show that this mediation was total in the positive pole of the well-being domain (i.e., flourishing) but only partial in the negative pole (i.e., burnout). This contrast may be due to the fact that boundary violations are a negative variable and, therefore, are more easily significantly related to the negative pole of well-being.

Gender Moderates the Relationship between Boundary Violations and Segmentation Behavior from Work to Family, But Not from Family to Work
In line with our hypothesis based on role theory [18,23,24,41], segmentation is affected by violations from work to family more strongly in the case of females when compared to males. Thus, as observed in this study, when females' family life (the predominant domain due to their traditional role) is affected by their work activity, it is more difficult for them to establish boundaries between the domains, leading to more interdomain management difficulties.
On the other hand, with regard to family violations in the work sphere, the results do not show significant differences between males and females. In this case, family violations in the work domain were expected to affect males more than females. This result may be due to a trend toward significant changes in the traditional division of gender roles and in the personal way they are experienced. Beyond the immediate crisis, some researchers [22] also point to tendencies that may ultimately promote gender equality in the labor market: businesses are rapidly adopting flexible work arrangements (e.g., teleworking), which could persist hereafter; there are also many fathers who now have to take or want the primary responsibility for child care, which may erode social norms that currently lead to an unequal distribution of the division of labor in housework and child care.
As far as the moderation hypotheses are concerned, the results do not demonstrate any significant gender moderation between boundary violations and work-family balance in either direction-family-to-work and work-to-family. This result may be due to the fact that work-family balance is an overall inter-role valuation of congruity between work and family domains [38] and, thus, once again, is in line with the role theory [18,23,24,41]. More specifically, as females socially accept the family role as being the most important, as do males their work role, violations do not interfere in the overall inter-role valuation of congruity between the work and family domains [38], in other words, in work-family balance. On the contrary, the same does not occur for females when the adoption of a strategy that allows them to protect their preponderant role, i.e., work-family segmentation, is questioned.

Limitations and Future Recommendations
Our study has limitations that must be acknowledged. First, its cross-sectional nature prevents the establishment of causal relationships, thus calling for longitudinal research to explore the causal directionality between the study variables. Second, our sample is based on only one nationality, and therefore it is important to test these relationships among individuals from other countries and cultures. Third, our study focuses on an analysis of individuals, and it would be interesting and relevant to study couples (dyads). Fourth, considering that the data collection was carried out in a pandemic context, during the first confinement, and that this fact compromised the reality of the individuals' boundary limits, the response to the scales (e.g., segmentation, WFB) could have been adjusted to what the individuals would consider the situation borderline ideal due to COVID-19. Thus, the reading of the data must be done with this precaution. Future research should also explore both directions of the concepts (i.e., boundary violations and boundary segmentation behavior from work to family and family to work), as our findings reflect different results according to the direction. In the same vein, future studies should include different dimensions of well-being, including positive and negative poles, as the different outcomes are evident from the results of this study. Additionally, future studies could test this model in work contexts other than that of teleworking in order to compare findings.

Practical Implications
Although this study was conducted during a period of lockdown, our results have several implications for the way teleworking beyond lockdown should be implemented to promote well-being. This study clearly shows that violations both from family to work and in the opposite direction are causing teleworkers discomfort. Thus, these violations should clearly be avoided. It should be noted that this research shows that segmentation is one of the means not only of reducing the negative effect of violations on well-being but also of attaining a direct positive effect on well-being, thus reducing burnout and promoting flourishing. Therefore, teleworkers should use the segmentation tactics identified in the four categories advanced by Kreiner et al. [8] (behavioral, temporal, physical and communication). For instance, teleworkers can turn off work notifications on their electronic devices when they are with their family (and vice versa); they can also establish and communicate their working hours and nonworking hours to those close to them (at work and in the family), establishing the permeable situations (e.g., an emergency); if possible, when working from home, they can choose an isolated place to perform their duties. Digital hygiene and turning off screens during nonworking hours is important, especially in a teleworking situation [58]. Kreiner et al. [8] suggested that, in the face of a violation, its occurrence should be verbalized in order to create awareness and prevent future violation situations. Accordingly, managers, co-workers and family members should also act in such a way as to avoid violations. Furthermore, this study suggests the importance of promoting work-family balance in this atypical situation of teleworking and lockdown.
Thus, it suggests that, during a crisis such as COVID-19, organizations should not fail to promote a culture of support for the work-family relationship. Moreover, as suggested by Sturges [59], individuals should also adopt balance strategies, referred to as "crafting" behaviors, namely proactive and goal-oriented behaviors, to achieve such balance. To this end, the author highlights the importance of each individual defining their work-family balance and using relationships (in the family and at work) to promote this balance. This strategy appears to be particularly important at a time when job and family demands, as a result of lockdown, have had to be readjusted. It is, therefore, important to readjust work-family balance to this reality.
Furthermore, this study has clearly highlighted the differences in gender roles. In the situation of teleworking during the pandemic, even more demands and difficulties in attaining segmentation and balance have fallen to females. These results reiterate the need to take gender-responsive economic and social measures. Furthermore, in a scenario such as that of lockdown, they stress the need for organizations to adjust workloads in order to maintain the well-being of employees, especially females.
The fact that the results are for all the teleworkers, regardless of their marital status or whether they have children or not is also noteworthy, thus pointing to the importance of considering these practical measures for all employees.