Smartphones and tablets are the newest communication tools, which have a relevant role in work dynamics today (Rennecker and Godwin, 2005), becoming a part of everyday working life (Derks et al., 2014). This issue is part of the wider debate on the technostress that is the stress due to information excess. In other words, technostress is the stress that workers experience as a result of their use of information systems in the organizational context; it may reduce job satisfaction, commitment, productivity and produce an invasion into personal life (Ayyagari et al., 2011; Tarafdar et al., 2015). In recent years, several studies have examined the role of the use of technology for professional purposes in relation with well-being and work-family balance (Derks et al., 2015). TASW refers to the fulfillment of work tasks using information technology and telecommunications, inside or outside the home (Fenner and Renn, 2004; Derks et al., 2014), mainly through the support of smartphones and tablets.

The spread of new technologies has increased the non-standard work schedule, gradually involving evenings, nights and weekends for many workers (Derks and Bakker, 2014), in the case of formal agreements that formalize it or even without these kind of agreements. Today, the progressive sliding of work into personal life, facilitated by new technologies, not only involves workers with specific smart working agreements but also traditional workers who have defined working hours and are paid for hours worked in the office, not just for their achievements. Therefore, in this paper we focus on supplemental work during off-work hours assisted by technology (Fenner and Renn, 2004) and not on the practices that, based on formal agreements, imply that employees work out of the office for part of their expected working time (Peters et al., 2009).

This topic is particularly important in Italy since digital devices (e.g., smartphones, tablets, and laptops) ownership and their use are high in this country (We Are Social, 2017). Moreover, the huge use of the internet through many different devices is especially high among professionals: 100% of directors, managers and academic professors, 99.5% of entrepreneurs and self-employed workers, and 98.8% of office workers and teachers, according to a survey that involved a representative sample of the Italian population (Belluati, 2016). The spread of the compulsive behavior of text-message and email checking makes this behavior socially acceptable, thus making it even more widespread. Moreover, Italy is a country characterized by a weak labor market, high job insecurity and unemployment, frequent downsizing and business failures, where those people who have a job tend to intensify their workload in order to preserve it (a widespread phenomenon also in other countries, Derks and Bakker, 2014; Giunchi et al., 2016). Work intensification is both the direct consequence of the reduction of the workforce (people who stay work more) and a strategy aimed at securing work continuity (people work more to stay). Nevertheless, no systematic studies exist about the effectiveness of this strategy. Today’s organizations have high expectations of employee availability, and this factor leads employees to feel obliged to immediately answer work messages and emails, also during off-work time (Davis, 2002; Derks and Bakker, 2014).

Some studies have highlighted smartphone use advantages: interaction and collaboration between co-workers is simplified (Pica and Kakihara, 2003), work schedules are more flexible, and productivity may improve (Locke, 2005, unpublished). However, among the problematic issues, TASW is associated with constant control and supervision by others, and lack of autonomy, which generate problems particularly if people perceive the intrusion of work into the rest of their life through new technologies, which are outside of their control (Derks et al., 2014). The perception of organizations’ expectations of being constantly connected in order to respond to working requests is another problematic aspect (Bakker and Derks, 2010). Smartphone use during off-work time is associated with information overload and loss of control over information flow (Derks et al., 2014), which may be a source of stress as literature on technostress highlighted (Ayyagari et al., 2011; Tarafdar et al., 2015).

Although in certain cases the use of technology is associated with work-life balance strategies (e.g., leaving the office earlier in order to solve a family problem and conclude working activities remotely, through a phone call, text or email in the evening), it can lead to difficulties in managing the balance between work and life (e.g., Davis, 2002; Higgins and Duxbury, 2005; Jarvenpaa and Lang, 2005). Finally, for those who generally use smartphones to stay connected to their work, it can be very difficult to detach psychologically from their work (Derks and Bakker, 2014; Derks et al., 2014). This can lead to negative consequences for their well-being (Meijman and Mulder, 1998) and work-family balance (van Hooff et al., 2006). In the present study, we take into account a specific aspect of the supplemental work assisted by technology. In the framework of the JD-R Model (Bakker and Demerouti, 2007, 2014, 2016), we want to deepen the role of a specific job demand: the off-work hours technology assisted job demand (off-TAJD), that is the perceived request to use technological devices to work during off-work time.

According to previously introduced references and in line with the boundary theory, permeable boundaries increase the spillover from work to the private domain (Ashforth et al., 2000). Problems in work-home balance are usual today and the presence of smartphones and tablets “always on,” the attitude toward continuously monitoring notifications, the tendency to react immediately to emails may have negative consequences on work-family balance (Derks et al., 2015).

Research on work-family boundaries emphasized that changes in demographics (more dual-earner or dual-career couples and single-parent families) and working conditions (more job insecurity and work intensification, blurred boundaries between work and life) has led to a lot of transformations in the work and family interaction (Ghislieri and Colombo, 2014). These changes have contributed to making workers less able to reach a satisfying work-life balance. The issues of WFC and WFE are the central constructs in this area of studies: to simplifying, several studies show that WFC is related to lower work satisfaction, higher turnover intentions, emotional exhaustion and a health impairment process (Colombo and Ghislieri, 2010; Cortese et al., 2010; Molino et al., 2016a; Ghislieri et al., 2017) while WFE is associated to high professional commitment and lower turnover intentions (Russo and Buonocore, 2012; Tummers and Den Dulk, 2013; Ghislieri et al., 2017). These findings indicate that research on WFC and WFE, intertwined with the study of the use of new technologies, can be considered an in-depth exploration of the technostress topics cited above (Ayyagari et al., 2011; Tarafdar et al., 2015).

The work-family interface is described as the process of influence between pressures and resources from the work (or family) domain and the individual’s behavior in the family (or work) domain (Bakker et al., 2011). Among the different theoretical models used to explain this process, the conflict perspective (WFC) dominated studies in the 1980 and 1990s, while the enrichment perspective (WFE) has become increasingly important in the last two decades (Ghislieri et al., 2011; Russo and Buonocore, 2012).

The WFC approach is based on the role theory (Merton, 1957) and on the role strain hypothesis (Goode, 1960). WFC has been defined as: “a form of inter-role conflict in which the role pressures from the work and family domains are mutually incompatible in some respect. That is, participation in the work (family) role is made more difficult by virtue of participation in the family (work) role” (Greenhaus and Beutell, 1985, p. 77). WFC may be time based, strain based or based on incompatible behavioral demands. Research on behavioral role conflict is infrequent and, in the present study, we only focused on time- and strain-based work to family conflict (Netemeyer et al., 1996; Derks and Bakker, 2014).

As mentioned above, in the last years, scholars have started to consider, in studies, not only WFC, but also WFE. Several authors, considering theoretical and empirical studies, points out that WFC and WFE (and related concepts like enhancement, facilitation or positive spillover) are distinct constructs (Grzywacz and Butler, 2005; Carlson et al., 2006; Ghislieri et al., 2011). About WFE, when one role enhances the quality of the other, we have an enrichment process: “WFE occurs when work experiences improve the quality of family life, and family-to-work enrichment occurs when family experiences improve the quality of work life” (Greenhaus and Powell, 2006, p. 73). The enrichment process may be instrumental, when resources move from one role to the other with a direct effect, or affective, when resources from one role create a positive affective state within this role that, consequently, fosters the performance in the other role (Greenhaus and Powell, 2006).

Scholars consider WFE a central issue for employers, employees, and organizations (Molino et al., 2013; Ghislieri et al., 2017) because research evidence shows that WFE is positively associated with job-related (Shockley and Singla, 2011), family-related (van Steenbergen et al., 2007) and health-related results (van Steenbergen and Ellemers, 2009). To enhance individual and organizational well-being, it is important to understand not only how to reduce WFC but also how to promote WFE (Bakker et al., 2011; Biggio and Cortese, 2013; Molino et al., 2013). The main determinants of WFE are job resources (Carlson et al., 2006; Bakker et al., 2010), and the role of the job demands in work-family enrichment is unclear. Despite the fact that some studies report low negative relationship between job demands and WFE (Carvalho and Chambel, 2014, there are few studies that include job demands like determinants of the WFE (Ghislieri et al., 2011). Furthermore, some authors who considered similar constructs point out that highly demanding jobs may be compatible with enrichment experiences in the presence of adequate support and autonomy (Grzywacz and Butler, 2005).

Among the different theories used to understand work-family interaction, the JD-R Theory is largely considered (Bakker and Demerouti, 2014, 2016; Schaufeli and Taris, 2014; Tement and Korunka, 2015). This theory explains health impairment and motivational processes considering two different types of working conditions as the main determinants of the processes: demands and resources. It is flexible and adaptable to a very different kind of organization and is useful not only to define implications for human resources management but also health promotion initiatives, starting from a specific identification of job demands and job resources that are relevant in a particular context.

According to the JD-R Theory, “Job demands refer to those physical, psychological, social or organizational aspects of the job that require sustained physical and/or psychological (cognitive and emotional) effort or skill and are therefore associated with certain physiological and/or psychological costs” (Bakker and Demerouti, 2007, p. 312). Job demands are not harmful by definition but they may be job stressors when great effort is required, which is not followed by adequate recovery (Meijman and Mulder, 1998). In this study, considering the research previously mentioned, we included off-TAJD in the category of job demands and formulated the following main study hypotheses:

We tested the relationship between off-TAJD and both WFC and WFE beyond and above the effects of other traditional job demands and job resources. With regard to other job demands, we included in this study both workload and emotional dissonance. Workload is a job demand that represents the amount of tasks and activities that workers have to carry out (Schaufeli and Taris, 2014). To date, several studies have investigated this (Bakker and Demerouti, 2007; Giunchi et al., 2016; Molino et al., 2016a) and showed that it is one of the main determinants of WFC (Cortese et al., 2010; Ghislieri et al., 2012). The second job demand considered is emotional dissonance, which refers to a discrepancy between felt and displayed emotions (Zapf et al., 1999). Studies underlined that its effects can “expand” in other life domains and influence the work-family interface (Cheung and Tang, 2009; Yanchus et al., 2010; Ghislieri et al., 2012). As mentioned above, the relationship between job demands and WFE has been investigated less in literature to date, since enrichment is mainly linked to resources (Carlson et al., 2006; Bakker et al., 2010). Therefore, in this study we hypothesized a low negative relationship between job demands and WFE, mainly for those demands characterized by a strong spillover effect on the rest of life, namely emotional dissonance (Ghislieri et al., 2017).

Job resources represent the second set of job characteristics and “refer to those physical, psychological, social, or organizational aspects of the job that are either/or functional in achieving work goals; reduce job demands and the associated physiological and psychological costs; stimulate personal growth, learning, and development” (Bakker and Demerouti, 2007, p. 312). Job resources are the main determinants of WFE and may reduce WFC (e.g., Voydanoff, 2004; Molino et al., 2013; Ghislieri et al., 2017). In this study, we considered supervisory coaching as a job resource; it refers to coaching behavior on the part of the supervisor that indicates respect, concern about employee needs and feelings, and help in routine tasks and skills development (Ellinger et al., 2003; Rafferty and Griffin, 2004). Previous studies found for both supervisory coaching and support a positive relationship with WFE and a negative one with WFC (e.g., O’Driscoll et al., 2003; Frye and Breaugh, 2004; Molino et al., 2013; Liu and Cheung, 2015) since supportive behavior from supervisors may generate a positive working environment (Totterdell and Holman, 2003; Emanuel et al., 2016a,b; Zito et al., 2016) and improve work-family interactions (Molino et al., 2013). In the study we hypothesized also that:

As mentioned in the introduction, the topic of work-family interface is deeply associated with studies about gender differences (Carr, 2002). In fact, the interest about work-life interface appeared when the participation of women in the labor market gradually increased, questioning the traditional gender role division between the male breadwinner and female homemaker (Cunningham, 2008). The transformation of the attitudes of men and women in the active participation and involvement in the two major life domains (work and family) seems to vary in relation to the cohort (linked to gender role socialization processes) and to the prevalence, in different countries, of traditionalist or egalitarian gender culture (Giunchi et al., 2016), although data concerning these differences are not always convergent (Rajadhyaksha et al., 2015). In Italy, while the gender roles of men and women have become more balanced (Ghislieri and Colombo, 2014), as in other contexts (Carr, 2002), the major load still remains higher for women with regard to care work (family, children, housework), even when the occupation is challenging (Naldini and Saraceno, 2008).

Despite the work-family balance burden being especially on women, several studies have revealed similar levels of WFC and WFE perceptions between men and women in Italy (Ghislieri and Colombo, 2014), as comparably observed in other studies (Byron, 2005). In Italy, women’s participation in the labor market is a well-established phenomenon and the job search is an important opportunity for the expression of women’s identity, especially those with a high education level; the job role remains central, especially for men (Bertolini, 2013).

The role of gender as a moderator of the relationship between demands-resources and WFC and WFE has been studied but results are conflicting (Rajadhyaksha et al., 2015). In fact, some studies have reported significant differences while others have not identified specific dynamics for the two genres. Some meta-analyses have shown different relationships between WFC and potential consequences. For example, the study of Shockley and Singla (2011) noted that the negative relation between WFC and satisfaction with family life was stronger in men than in women. However, few studies investigated the gender differences in-depth in the relationship between potential antecedents and work-family interference (WFC and WFE): this aspect is central to our study. Particularly, this study investigates if the introduction of new technologies may interact with work-family balance differently for women and men. Several authors emphasized the importance to understand, in studies on work-family balance, the effect of changing in work models and practices (Amstad et al., 2011), mainly nowadays when the introduction of new technologies are influencing the transformation of gender role models (Halpern, 2005; Ford et al., 2007; Ghislieri and Colombo, 2014). The perspective from which we study this dimension is explorative; therefore, we do not define specific hypotheses with regard to this.

This study involved human beings through the administration of a self-report questionnaire. The Bioethical Committee of the University of Turin examined and approved the research project (14/7/2016). Since there was no medical treatment or other procedures that could cause psychological or social discomfort to participants, additional ethical approval was not required. The research was conducted in line with the Helsinki Declaration (World Medical Association, 2001), as well as the data protection regulation of Italy (Legislative Decree No. 196/2003). Participation in the research was voluntary, without receiving any reward; data collection and analysis were anonymous.

The study involved a convenience sample of 671 Italian workers who filled in an on-line self-report questionnaire. The reason to use the convenience sampling is due to the study’s topic of interest and to the confidentiality and the degree of anonymity that it affords its participants. In fact, the present study asked participants if their Company demands them to use technology to work during off-work hours – a demand that is not formalized in their employment contract. Using another method of survey sampling (such as choosing a particular organization to sample) employees could have been hesitant and cautious in revealing their perception about the demands from the Company to use technology to work during off-work hours. In this study, all participants were anonymous; we knew the occupational sector and job profile, not the specific organization they worked for. The cover sheet of the questionnaire explained to participants the anonymity, confidentiality and voluntariness of their participation, and the aims of the research. The informed consent was present and integrated in the on-line administration procedure.

Participants were recruited by Master degree students in Work and Organizational Psychology at the University of Turin who received training in methodological and ethical issues about questionnaire administration. Seven students volunteered to distribute the questionnaire packages to employees they know; they received instructions to collect data from employees differentiated by occupational sectors and gender. From the 700 distributed questionnaires (100 per student), 671 usable questionnaires were returned.

Among the participants, 319 were male (47.5%) and 352 were female (52.5%).

In the Male sample, 52.0% were unmarried, 42.9% married or cohabited, 5.0% separated, divorced or widowed, and 32.6% had children. Among them, 49.2% had finished high school, 36.7% had a bachelor’s or master’s degree, and 9.7% had finished elementary school. Their average age was 36.29 years (SD = 11.47; minimum = 19; maximum = 62). Most of the male participants had a full-time job (88.4%) and a permanent contract (72.4%). The job profile was office workers for 55.8% of male participants, blue-collar workers 23.8% and middle managers 20.4%. Participants were from different occupational sectors: 23.8% industry, 14.7% commerce, 12.9% private services, 6.3% public health, 6.3% tourism, 5.0% education and research, and 5.0% public services; the remaining participants were from other sectors. Weekly working hours were, on average, 40.02 (SD = 9.78; minimum = 7; maximum = 60). Mean seniority on the profession was 10.33 years (SD = 10.48; minimum = 1; maximum = 46).

In the Female sample, 49.7% were unmarried, 44.9% married or cohabited, 5.1% separated, divorced or widowed, and 33.5% had children. Among them, 48.9% had a bachelor’s or master’s degrees, 43.2% had finished high school, and 4.0% had finished elementary school. Their average age was 36.01 years (SD = 11.78; minimum = 20; maximum = 64). Most of the female participants had a full-time job (74.7%) and a permanent contract (70.5%). The job profile was office workers for 66.8% of female participants, blue-collar workers 16.5% and middle managers 16.8%. Participants were from different occupational sectors: 16.8% commerce, 13.1% private services, 12.5% public health, 10.2% education and research, 8.2% public services, 8.0% tourism, 6.8% industry, and 6.8% social sector; the remaining participants were from other sectors. Weekly working hours were, on average, 36.39 (SD = 9.98; minimum = 7; maximum = 60). Mean seniority on the profession was 9.38 years (SD = 10.58; minimum = 1; maximum = 45).

Work-family conflict (WFC) was assessed with five items of the Netemeyer et al. (1996) scale (Italian version by Colombo and Ghislieri, 2008). All items were scored on a five-point scale, ranging from 1 = never to 5 = always. An example item is “Things you want to do at home do not get done because of the demands your job puts on you.” Cronbach’s alpha for the scale in this study was 0.89.

Work-family enrichment (WFE) was measured by three items (Ghislieri et al., 2011) scored on a Likert scale from 1 = strongly disagree to 5 = strongly agree. An example item is “At work you feel positive emotions and this helps you to be a better family member.” Cronbach’s alpha was 0.85.

Off-work hours Technology Assisted Job Demand (off-TAJD) was assessed with three ad hoc items, which asked how often, in the employee’s perception, the Company demands her/him to use technology to work during off-work hours. All items were scored on a five-point scale, ranging from 1 = never to 5 = always. An example item is “How often does your organization require you to answer phone calls and emails during off-hours?” Cronbach’s alpha was 0.94. An exploratory factor analysis (EFA) performed on a random half sample (N = 336) showed a one factor solution which explained 85.66% of the variance, with factor loadings equal to 0.90 for Item 1, 0.98 for Item 2 and 0.89 for Item 3. A confirmatory factor analysis (CFA) on the other half of the sample (N = 335) confirmed results of EFA: χ²(0, N = 335) = 0.00, p = 0.00, RMSEA = 0.00 (0.00, 0.00), CFI = 1.00, TLI = 1.00, SRMR = 0.00. The standardized factor loadings were 0.93 for Item 1, 0.94 for Item 2 and 0.91 for Item 3. We found that the level of off-TAJD is significantly higher for workers who receive more than 10 working e-mails per day (M = 6.74; SD = 3.93) compared with those who receive from 0 to 10 e-mails per day (M = 6.18; SD = 3.57) [t(664) = -1.85, p < 0.05].

Workload was assessed by four items (Bakker et al., 2004). All items were scored on a five-point scale, ranging from 1 = never to 5 = always. An example item is “You have too much work to do.” Cronbach’s alpha was 0.81.

Emotional dissonance was assessed with four items developed by Zapf et al. (1999). All items were scored on a five-point scale, ranging from 1 = never to 5 = always. An example item is “How often during your work do you have to display emotions which do not correspond to your inner feelings?” Cronbach’s alpha was 0.89.

Supervisory coaching was assessed by five items adapted from Graen and Uhl-Bien’s (1991) work. Items scored on a Likert scale from 1 = never to 5 = always. An example item is “Your supervisor informs you whether he/she is satisfied with your work.” Cronbach’s alpha was 0.91.

The statistics software SPSS 23 was used to perform descriptive data analysis in each sample separately (Males and Females). Moreover, Pearson correlations were tested in order to examine the relationships among variables, and Cronbach’s alpha coefficient was calculated to test the reliability of each scale. The analysis of variance (t-test for independent samples) was used to examine differences in the variables’ means between the two samples.

The psychometric characteristics of the off-TAJD scale were examined both through an EFA (Maximum Likelihood extraction) performed with SPSS Statistics 23 and through a CFA performed by Mplus 7 (Muthén and Muthén, 1998–2012). The CFA method of estimation was Maximum Likelihood (ML). The overall sample (N = 671) was randomly split into two subsamples homogeneous for some demographic characteristics (gender, job profile, and occupational sector): the EFA was performed on the first subsample (N = 336) and the CFA on the second subsample (N = 335).

A multi-group full structural equation model (SEM) was performed using Mplus 7 in order to test the hypothesized model and the measurement model across both Male and Female samples. The method of estimation was Maximum Likelihood (ML). According to the literature (Bollen and Long, 1993), the model was assessed by several goodness-of-fit criteria: the χ² goodness-of-fit statistic; the Root Mean Square Error of Approximation (RMSEA); the Comparative Fit Index (CFI); the Tucker Lewis Index (TLI); and the Standardized Root Mean Square Residual (SRMR). Non-significant values of χ² indicate that the hypothesized model fits the data. Values of RMSEA smaller than 0.05 indicate a good fit, values smaller than 0.08 indicate an acceptable fit and values greater than 1 should lead to model rejection. CFI and TLI values greater than 0.90 indicate an acceptable fit, and values greater than 0.95 indicate a good fit. The SRMR has a range from 0 to 1, with a cut-off criterion of 0.08, with higher values indicating poorer fit to the empirical data, and values lower than 0.05 indicating an excellent fit.

In order to address the common method variance issue, we conducted Harman’s single-factor test (Podsakoff and Organ, 1986; Podsakoff et al., 2003) and examined the unrotated factor solution involving all variables of interest (24 items) in an EFA. Results of this analysis showed six factors with an eigenvalue greater than one. No single factor explained a great amount of the variance (variances ranged from 5.05 to 22.93%) and the six factors combined explained 67.09%. This indicates that common method variance was not a major problem in the present study.

In order to assess the discriminant validity, using Mplus 7 we first test the measurement model, which fitted the data well [χ²(237, N = 671) = 724.88, p < 0.001, RMSEA = 0.06 (0.05, 0.06), CFI = 0.95, TLI = 0.95, SRMR = 0.04]. Then, we compared it with a model where the estimated correlation parameter between two of the six estimated constructs was constrained to 1.0. The chi-square difference test showed that each of the 15 constrained models (one pair of factors constrained at a time) had a significantly higher χ² than the measurement model, confirming that factors are not perfectly correlated and that discriminant validity is achieved (Anderson and Gerbing, 1988). Finally, we tested multicollinearity by running the variance inflation factor (VIF) and the tolerance levels using SPSS Statistics 23. Results showed that VIF values ranged from 1.06 to 1.32 and tolerance values ranged from 0.68 to 0.94, indicating that multicollinearity problem among the variables did not exist.

As all research with a cross-sectional approach, this study did not permit establishing causality relations between variables (Podsakoff et al., 2012). Further studies, using longitudinal design, should examine the causal relationship between off-TAJD, workload, emotional dissonance and supervisory coaching as determinants, and WFC and WFE as consequences. Moreover, the sampling procedure (convenience sampling) adopted in this study poses constraints on the extent to which the study’s findings can be generalized.

A second limitation is the use of self-reported questionnaires that can potentially contaminate results, since observed relationships may be artificially overestimated because of the respondents’ tendency to answer in a coherent way. Nevertheless, self-reported data seemed to be the most appropriate approach in our study since it evaluated workers’ subjective perceptions of job demands, job resources, WFC and WFE.

Another limitation is the questionnaire online administration procedure. The use of Internet represents the opportunity of having a large group of individuals and lower costs associated with collecting data. At the same time, it is a limitation (Kraut et al., 2004) because it is difficult to control the study environments and Web users have different types of hardware, software, and Internet connections. Furthermore, extraneous or temporary factors could influence responses.

Moreover, according to the JD-R model (Bakker and Demerouti, 2014), future studies should investigate the buffering role of some resources in the relationship between demands and WFC and WFE (Molino et al., 2015). In particular, as for the relationship between off-TAJD and WFC, the moderating role of resources such as job autonomy and transformational leadership could be investigated.

A further study limitation is the absence of objective data, including the number of emails and phone calls received during off-work time. This kind of data could be detected better in studies within specific working contexts. The present study, considering a heterogeneous sample with different job occupations, organizations and sectors, intended to observe the relationships among variables in a general and cross way. In the future, more specific studies could focus on the specificity and distinctiveness of particular working contexts, also considering the influence of the organizational culture (Schein, 2010).

Finally, future studies could integrate the quantitative research method with qualitative ones, in order to understand better the meaning of off-TAJD in employee opinions, what mechanisms can amplify its influence on the rest of life, and what processes of perception and meaning construction around these topics men and women at work apply, taking into account both the organizational culture and national culture.