An examination of remote e-working and flow experience: The role of technostress and loneliness

Since early 2020, the Covid-19 pandemic has led to numerous businesses around the world making use of information and communication technologies (ICT) more frequently than ever to help transition operations to remote e-working. As a result, using multiple technologies on a daily basis has become the norm for many employees across the world. While it is evident that working remotely may trigger higher ICT demands and reduced face-to-face interaction, less is known about how this exposure may influence employees ’ subjective mental experiences related to concentration and satisfaction at work (i.e., flow). The aim of this present study is to gain insights and to explore the relationship between remote e-working and employee flow experiences by introducing two key stressors; technostress and loneliness. Data were collected from a survey of 202 employees from the financial services sector in Turkey. The results revealed that remote e-working experience had a significant and positive effect on the flow levels of employees. Technostress and loneliness serially mediated the relationship between remote e-working and flow. The findings contribute to remote e-working research by exploring the consequences of such experiences and introducing two important key stressors, which result in lower levels of flow at work. Practical implications are provided for improving remote e-working conditions and employee well-being.


Introduction
In recent years, working remotely has become increasingly prevalent across the majority of working sectors. In particular, the Covid-19 global pandemic, which started in early 2020, has somewhat accelerated the shift to working from home using communication and information technologies. In order to implement social distancing during the pandemic, an overwhelming majority of individuals have started to work from home (Kniffin, Narayanan, Anseel, Antonakis, Ashford, Bakker, Bamberger, Bapuji, Bhave, Choi, Creary, Demerouti, Flynn, Gelfand, Greer, Johns, Kesebir, Klein, & Lee, 2021). In many cases, organizations are considering whether to continue the working from home regime in the aftermath of the pandemic (Molino, Ingusci, Signore, Manuti, Giancaspro, Russo, & Cortese 2020). Thus, in the future, remote e-working will most likely be the potential norm for work. The digitalization of working and the prevalent use of Information and Communication Technologies (ICT) has led to a substantial growth in research into its impact from a psychological as well as physical standpoint (Stadin et al., 2021).
While remote e-working has several benefits, such as flexibility (Mann & Holdsworth, 2003), it has also created various challenges and stressors on employees for reasons such as boundarylessness between work and non-work and lack of space to attend to work. In particular, exposure to technostress, which is caused by the use of ICT technology, has been associated with negative feelings such as anxiety (Salanova, Llorens, & Cifre, 2013), reduced user satisfaction (Jena, 2015;Tarafdar, Darcy, Turel, & Gupta, 2015), as well as health-related outcomes such as symptoms of burnout (Barber & Santuzzi, 2015;Hennington, Janz, & Poston, 2011).
Although there has been a plethora of studies related to the remote eworking experience, these studies have focused on a rather narrow definition of the concept, i.e., teleworkers (those workers who usually work from home), which are not necessarily applicable to all types of remote e-workers. Clearly, it is important to identify and reveal the possible synergies across different dimensions of remote e-working; yet, very few studies have made the attempt to provide a holistic view of eworking in terms of an assessment of aspects related to well-being, worklife balance, and job effectiveness (e.g., see Charalampous, Grant, Tramontano, & Michailidis, 2019, for exceptions). Existing studies have not yet studied flow as an outcome of remote e-working, despite its importance to employees particularly during challenging times, of which the current pandemic is clearly one. Informed by job-demands-resources (JD-R) theory, we explore the relationship between remote e-working and employee flow experience by introducing to key stressors (loneliness and technostress). While previous research has shown that lonely individuals are more likely to develop problematic Internet use behaviour (Caplan, 2002), the impact of remote e-working on loneliness remains unclear. Thus, research to date provides a limited understanding of the technology-induced stressors related to remote e-working that are likely to influence employees' psychological outcomes (Tams, 2015). Nevertheless, it is imperative to examine the effects of remote e-working because when individuals feel stressed due to perceived loneliness and increased use of ICT technology, this may have an impact on their flow levels, a key positive mental experience related to enhanced concentration at work (Ozkara, Ozmen, & Kim, 2016).
In light of the explanations above, the aim of the present study is to fill an important gap in the literature by gaining new insights into ICTinduced stressors on employee flow experience in the context of remote e-working during lockdown. We posit that when employees have a negative remote e-working experience, they will experience higher levels of technostress and loneliness, resulting in lower levels of flow at work. The focus is on technostress and loneliness for two reasons: first, spending long hours using ICT may impact employees' engagement in a work activity and optimal experience (i.e., flow) (Suh & Lee, 2017); second, although loneliness is considered as an important work-related stressor and distraction (Mann & Holdsworth, 2003), the mechanisms through which it may relate to flowa positive state of full concentrationhas not to date been explored. Thus, the study explores the relationship between remote e-working and flow using two key stressors: technostress and loneliness. The research model is shown in Fig. 1.

Job-demands-resources theory
The JD-R framework is one the most widely used stress models that includes a broad range of job demands as well as resources (Demerouti, Bakker, Nachreiner, & Schaufeli, 2001). The model was first developed to explain the antecedents of burnout and was later revised, on various occasions, to include a variety of individual and organisational level factors (Schaufeli & Taris, 2014). According to this framework, job resources refer to "those physical, social or organizational aspects of the job that may do any of the following: (a) be functional in achieving work goals, (b) reduce job demands and the associate physiological and psychological costs, (c) stimulate personal growth and development" while job demands refer to "those physical, social or organisational aspects of the job that require sustained mental effort and therefore associated with certain physiological and psychological costs" (Demerouti et al., 2001). These resources help employees to deal with job demands when attempting to achieve work-related goals (Mäkikangas, Leiter, Kinnunen, & Feldt, 2020).
In the current study, remote e-working is considered a job resource. Work-related resources mainly have supportive functions in the psychological regulation of work demands (Glaser, Seubert, Hornung, & Herbig, 2015). Organizations offering remote e-work without ensuring formal processes and procedures are often subsequently confronted with negative outcomes (Wheatley, 2012). On the other hand, researchers have shown that remote e-working may cause poor well-being and workplace pressure, which in turn will affect job effectiveness and performance (Barber & Santuzzi, 2015;Mann & Holdsworth, 2003).
Due to the prevalent diffusion of remote e-working there has been a growing interest in developing a measure to assess the quality and complexity of this experience. In order to contribute to this line of research, we adopt the E-Work Life (EWL) scale of Grant, Wallace, Spurgeon, Tramontano, and Charalampous (2019) to evaluate remote e-working as it assesses a range of theoretically relevant aspects of the remote e-working experience including work/life interference, organizational trust, effectiveness/productivity, and flexibility, all of which can be considered the support provided by the organization. According to the JD-R framework, job resources are critical to conducting tasks and achieving objectives by acting as a buffer against the negative impact of job stress (Adamovic, 2018;Halbesleben, Neveu, Paustian-Underdahl, & Westman, 2014;Hobfoll, 1989). Thus, it can be argued that if an organization provides support to the employee remote e-work experience, this can improve the employees' resources and their ability to handle job stresses.
Job demands are divided into challenge and hindrance demands (Cavanaugh, Boswell, Roehling, & Boudreau, 2000;Crawford, LePine, & Rich, 2010). While both kinds of job demands have adverse effects, challenge demands have the potential to advance personal growth and future gain, whereas hindrance demands restrain personal growth and achievement of goals (Cavanaugh et al., 2000;Schaufeli & Taris, 2014). According to the JD-R model, work circumstances can be demands/stressors or resources (Demerouti et al., 2001). The JD-R model does not limit the specific types of job demands or resources that one might consider; any demand or resource can have an effect on employees' well-being (Schaufeli & Taris, 2014).
Stressors can be considered hindrance job demands, which damage optimal behaviour (Wei, Zhu, & Chen, 2020). Relatedly, technology-related stressors are job demands that result from the specific characteristics of a given technology (Ayyagari, Grover, & Purvis, 2011). Research has shown that technostress directly affects employees' individual and organizational productivity (Suh & Lee, 2017). In particular, in the absence of situational coping mechanisms, technostress gradually exhausts individuals, leading to burnout (Mahapatra & Pati, 2018). To contribute to this line of research, we contend that technostress is a hindrance demand that is negatively associated with employees' state of flow. Despite the importance of technostress and its impact on employees, little research exists that reveals the effects of hindrance demands in the context of technology use (Wei et al., 2020).
In the current study, loneliness is also introduced as a hindrance demand, as loneliness is considered a work-related circumstance and stressor. The underlying reason is that the nature of remote e-working includes being isolated because of working away from the office (Mann & Holdsworth, 2003). Previous research suggests that work stress demands affect work outcomes (Cavanaugh et al., 2000;Podsakoff, LePine, & LePine, 2007). Thus, loneliness, together with technostress, may negatively affect flow as an outcome.
Flow is defined as being in a state of full concentration on one's work and all other distractions being removed (Slavec Gomezel & Aleksić, 2020), and is very useful for the growth and improvement of organizations and employees (Sharma, Misra, & Gupta, 2020). Studies show that flow is a crucial work-related outcome associated with job resources such as social support (Fagerlind, Gustavsson, Johansson, & Ekberg, Fig. 1. Research model. 2013;Mäkikangas, Bakker, Aunola, & Demerouti, 2010;Salanova, Bakker, & Llorens, 2006). Organizations benefit when their employees experience flow at work, as it is a desired experience that helps them to perform better (Engeser & Rheinberg, 2008). It is important to study the antecedents of flow, as these need more empirical investigation due to their complex nature (Knight & Waples, 2017), and as research that specifically focuses on the antecedents of flow at work are still scarce in the literature (van Oortmerssen, Caniëls, & van Assen, 2020). To date, challenge-skill balance (Csikszentmihalyi, 1990), daily recovery (Demerouti, Bakker, Sonnentag, & Fullagar, 2012), leadership (Sosik, Kahai, & Avolio, 1999), job characteristics (Demerouti, 2006), and job resources (Fagerlind et al., 2013;Kasa & Hassan, 2013) have been studied as the antecedents of flow within the literature.
In line with JD-R theory, research has found that flow can be extensively disrupted in the presence of hindrance demands (van Oortmerssen et al., 2020), due to the strong relationship between flow and engagement (Medhurst & Albrecht, 2016). To illustrate, if employees feel threatened by job (hindrance) demands, they may feel less motivated and this may result in a lack of the challenge-skill balance that is essential for flow (van Oortmerssen et al., 2020). Thus, we investigate flow as the outcome affected by job resources such as remote e-working, and job (hindrance) demands as technostress and loneliness.

The relationship between remote e-working and flow at work
The global remote e-working trend due to the current pandemic has paved the way for a sudden reconfiguration of work, resulting in a number of implications for both organizations and employees (Drabek & McEntire, 2003). Due to the sudden shift, some organizations were, quite understandably, not well prepared. Organizations offering remote e-work without ensuring formal processes and procedures were subsequently confronted with negative outcomes (Wheatley, 2012). As suggested by Grant et al. (2019), the complexity of remote e-working experience requires an assessment based on work effectiveness, work-life balance, and well-being. In line with JD-R theory, such necessary conditions provided by the organizations can be regarded as resources to explain flow at work. This is because job resources initiate motivation and strongly predict employee engagement and commitment (Bakker, 2015).
A fully experienced flow occurs when an individual has the impression of control and when their concentration is completely on the task at hand (Nakamura, Csikszentmihalyi, Snyder, & Lopez, 2002). The main condition for the development of a flow during an activity is that the individual perceives a balance between challenges and skills (van Oortmerssen et al., 2020). According to Csikszentmihalyi (1997), if employees are clearly informed about the objectives and are fully aware of how to perform the assigned tasks, they are more likely to experience flow. In turn, this may reflect on the employees' perception of remote e-working as a positive experience provided by the organization.
In this current study, we explore what happens when employees experience remote e-working and the mechanisms that operate behind experiences of flow as an outcome. Until now, flow has not been studied within the context of remote e-working. The focus is on flow as the outcome, as it is a crucial work-related state that can successfully increase emotional well-being during stressful times of uncertainty (Rankin, Walsh, & Sweeny, 2019). For these reasons, flow is suggested as an outcome of a positive remote e-work experience and the following hypothesis accordingly proposed:

H1.
Remote e-working is positively associated with flow at work.

The relationship between remote e-working and technostress
The proliferation of ICTs has impacted work in all sectors. These technological advancements have allowed work environments to be flexible, distributed, and distanced from conventional workplaces, such as home offices (Vasconcelos, Furtado, & Pinheiro, 2015). Consequently, empirical studies have shown that ICTs have resulted in increased stress levels amongst employees due to increased expectations of being constantly available, renewing their technical skills and working at a faster pace (Ayyagari et al., 2011;Ragu-Nathan, Tarafdar, & Ragu-Nathan, 2008;Wang, Shu, & Tu, 2008). However, there has been limited research in the literature on stress due to technology and, particularly, its antecedents (Tams, 2015).
A widely accepted definition of technostress is the "stress experienced by end users in organizations as a result of their use of ICTs" (Ragu-Nathan et al., 2008). According to this classification, the term includes five dimensions, (i) techno-overload (related to ICTs potential to require users to work faster and longer), (ii) techno-invasion (related to ICTs ability to invade users' personal lives), (iii) techno-complexity (related to ICT's complex features' potential to make users feel inadequate with their current skills), (iv) techno-insecurity (related to users feeling threatened with replacement due to automation or others with better ICT knowledge), and (v) techno-uncertainty (related to the disturbance users experience due to continuous upgrade and changes in ICTs).
We draw on the JD-R framework to contend the negative association between remote e-working experiences experienced by employees and technostress. The JD-R framework is one the most widely used stress models that includes a broad range of job demands as well as resources (Demerouti et al., 2001). The model was first developed to explain the antecedents of burnout and was later revised multiple times to include a variety of individual and organisational levels factors (Schaufeli & Taris, 2014). As previous studies have shown the positive relationship with remote e-working and the well-being of individuals (e.g., Grant et al., 2019) and in line with the JD-R framework, we argue that employees' remote e-working experience is a key resource that stimulates employee growth as well as achievement of work goals. Therefore, it is expected that employees who have positive remote e-working experiences will have reduced experience of technostress. Similarly, postulating from the perspective of JD-R, technostress is considered a key job demand that may result in significant psychological and physiological costs to employees. From the perspectives of employees, ICT intensifies pressures on individuals as it creates expectations for constant availability (Ayyagari et al., 2011;Hung, Chen, & Lin, 2015).
While recent years have witnessed a surge in studies that focus on demands related to workload, time pressure, and responsibility (LePine, Podsakoff, & LePine, 2005;Webster, Beehr, & Love, 2011), research on technostress, especially in the context of JD-R, is still in its infancy. For example, Mahapatra and Pati (2018) demonstrated that, in the absence of individual and situational coping mechanisms, technostress gradually exhausts individuals, leading to burnout. Thus, due to the widespread use of ICT in today's organizations, technostress may act as an essential component of the existing JD-R models. Hence, we propose that remote e-working is a key resource job resource and, due to its characteristics such as work-life balance, productivity, organisational trust, and flexibility, may induce a motivational process that reduces the effects of technostress on employees, and thus hypothesize that: H2. Remote e-working is negatively associated with technostress.

The relationship between technostress and loneliness
Loneliness is one of the most devastating problems people might suffer in social life (Gierveld, Tilburg, & Dykstra, 2006;Jeong & Kim, 2021;Singh, 1991). Described as a modern 'epidemic', loneliness has many physical, psychological, and social effects (Alberti, 2019). While loneliness has already been a public health issue within society, it has become one of the biggest struggles during the COVID-19 crisis (Luchetti et al., 2020). Mandatory social distancing and 'stay-at-home' orders have increased feelings of loneliness during the coronavirus outbreak (Luchetti et al., 2020). When the compelling circumstances of the pandemic and the proliferation of working from home are banded together, loneliness seems to be an inevitable consequence for employees.
Loneliness is defined as a subjective feeling about the deprivation of social relations, whereas social isolation is accepted as an objective lack of social companionships, especially where the quantity of social contact is important (Gierveld et al., 2006;Valtorta & Hanratty, 2012). In other words, loneliness is a subjective feeling that includes limited social skills and not having the desired social relations in terms of their number and quality. The state of loneliness has two generally accepted characteristics (Peplau & Perlman, 1982). First, it is a negative emotional state that occurs when individuals feel estranged from social interactions and emotional intimacy (Hazer & Boylu, 2010). Second, it is distinct from social isolation as individuals can feel lonely even if there is no social isolation, experience both of them together, or be socially isolated without feeling lonely (Valtorta & Hanratty, 2012). Previous research has shown that loneliness is associated with many negative attitudes and behaviours in the organizational context. For example, loneliness has an impact on various outcomes such as organizational commitment (Ayazlar & Güzel, 2014), employee performance (Ozcelik & Barsade, 2011), and intention to leave (Ertosun & Erdil, 2012).
Due to the nature of remote e-working and the mandatory circumstances of coronavirus pandemic, it is argued that people have frequently experienced the feeling of loneliness during this period (Kniffin et al., 2021). Relatedly, spending a long time using ICTs results in deteriorating social skills and leads to spending more time alone, being isolated from others, and introversion (Jaradat, Jibreel, & Shaik, 2020). Together, we contend that technostress and loneliness might be associated in the context of remote e-working. Remote e-working employees could suffer from technostress because of their excessive dependency on technology (Suh & Lee, 2017), whilst at the same time this context may hinder employees from establishing and maintaining social relationships, which results in loneliness.

The relationship between loneliness and flow
Remote e-workers may develop feelings of loneliness due to having fewer interactions and a diminished relationship with their co-workers (Fonner & Roloff, 2010;Ozcelik, Beetz, & Barsade, 2020). Mostly, remote e-workers do not see other colleagues very often compared to those performing conventional office work practices. The recent pandemic, however, has exacerbated this effect as mandatory work from home practices were put in place.
In the present study, loneliness is expected to prevent employees from being in flow when working, as flow is assumed to be an optimal state related to positive emotional and motivational experiences (Hektner, Schmidt, & Csikszentmihalyi, 2007). Since hindrance demands can have both psychological and social consequences (Bakker & Demerouti, 2007), loneliness can be considered a hindrance demand, which is regarded as work circumstances constraining employees' ability to achieve goals (Li, Taris, & Peeters, 2020). Hindrance demands negatively affect performance both directly and indirectly through strains and motivation (Bakker & Sanz-Vergel, 2013;Korunka, Kubicek, Paškvan, & Ulferts, 2015;LePine et al., 2005). Several studies have found that high-quality social interactions, such as daily discussions among colleagues, are critical to mental health (Mogilner, Whillans, & Norton, 2018). Yet, remote e-working decreases such high-quality interactions, including social networking (Klopotek, 2017). Particularly when the work is complex and ambiguous, remote e-working, during a global pandemic in our case, results in the inability to communicate, get support, and learn from others that can negatively affect employees. As employees cannot have the social interactions that they had in an office setting, they may suffer from loneliness during remote e-working (Larson, Vroman, & Makarius, 2020).
When people feel lonely, they become less committed to their organization and consequently perform worse (Ozcelik & Barsade, 2018). Thus, when e-workers experience increased feeling of isolation, they are less likely to maintain their productiveness or even feel that they are working effectively (Gajendran & Harrison, 2007). Isolated e-workers are less prone to rely on their own abilities and this can impair their job performance (Golden, Veiga, & Dino, 2008). Flow is perceived as a high-performance experience that is incompatible with negative emotions (Quinn, 2005;Rankin et al., 2019). In the presence of hindrance demands, flow is largely destroyed (van Oortmerssen et al., 2020). Thus, loneliness can be considered a hindrance demand that prevents efficient flow at work. Hence, the following hypothesis is suggested: H4. Loneliness is negatively associated with flow at work.

Remote e-working, flow, technostress and loneliness
According to the JD-R model, employee well-being results from a balance between job resources and demands (Schaufeli & Taris, 2014). In line with the JD-R framework, employees' remote e-working experiences are a key resource that stimulates well-being as well as achievement of work goals. It has been suggested that people look for the retention and protection of key resources (e.g., social support, self-esteem, knowledge) and are less motivated when they lose them, as resources hold employees responsible for work (Salanova et al., 2006). On the other hand, some work-related demands that constrain an employee's achievements and potential gains may emerge, which are described as hindrance demands (Cavanaugh et al., 2000). Accordingly, an imbalance between job resources and demands can lead to negative outcomes (i.e., loss of flow). Hence, for the purposes of our study, remote e-working is considered a job resource, technostress and loneliness as hindrance job demands, and flow as an outcome.
The relationship between e-work and flow is explored via our serial mediation model. Empirical evidence for the serial mediation process of technostress and loneliness remains unexplored. It is important to investigate this serial mediation because such an explanation would add to a detailed understanding of flow and its underlying mechanisms. We propose that, in the presence of a negative remote e-working experience, employees might well be expected to experience high levels of technostress and loneliness in turn, together resulting in lower levels of flow at work.

H5.
Changes in technostress and loneliness serially mediate the relationship between remote e-working and flow.

Sample
The questionnaire prepared to collect data was applied to professionals from the financial sector in Turkey who started working from home as of the outbreak of the pandemic. With regard to the respondents, the convenience sampling method was used. Participants were specifically financial sector professionals, as research into the management of current pandemic in European and Asian companies revealed that service industries doing knowledge work such as insurance businesses, banking services, and technological companies were those implementing higher rates of telework (Belzunegui-Eraso & Erro-Garcés, 2020; Charalampous et al., 2019). In addition, the results of a recent PWC report (2020) on remote working practices during the pandemic indicated that financial services executives are determined to make remote e-work more manageable for the employees. Participants were informed that participation was voluntary and anonymous. To provide better insight into the sample, participant demographic information, including sex, age, and education were requested in the questionnaires. Table 1 presents the demographic characteristics of the sample.

Data collection
The survey method was used to collect the data, with an online questionnaire used as the instrument. Data were collected via e-mail with the consent of the respondents by explaining the objective of the questionnaire. The process netted a total of 202 questionnaires, out of 511 initially sent via e-mail, giving a response rate of 40%. Our research model includes four constructs, and thus the questionnaire contained four scales: e-work life, technostress, loneliness, and flow at work.

Measurement
All the original scales used were first translated from English to Turkish and then back translated to English by experts, as suggested by Brislin (1980). The e-work life scale was derived from the measure developed by Grant et al. (2019). Respondents were asked to choose, on a five-point Likert scale, their degree of agreement or disagreement with the items (1 = strongly disagree, 5 = strongly agree). The scale contains 17 survey items (Cronbach's α = 0.88) and has a mean of 3.42 (SD = 0.77). Sample items include "I trust my organisation to provide good e-working facilities to allow me to e-work effectively" and "E-working makes me more effective to deliver against my key objectives and deliverables".
First, exploratory factor analysis (EFA) was conducted to test the variables and to define them with regard to their underlying factors (Hair, Black, Babin, Anderson, & Tatham, 2006). Accordingly, for e-work life, EFA explained 68% of the variance and the KMO measure showed a high sampling adequacy (KMO = 0.85), meaning that the data were suitable for factor analysis. Next, confirmatory factor analysis (CFA) was conducted to test the construct validity and to evaluate how well the data fitted the measurement model. CFA revealed that the original four-factor e-work life model showed good and acceptable fit indices (χ 2 /sd = 2.16, CFI = 0.94, GFI = 0.90, RMSEA = 0.076). All item loadings on the theorized construct were significant at the p < 0.001 level and no modifications were conducted.
The technostress scale was derived from the measure developed by Tarafdar, Tu, and Ragu-Nathan (2010). Respondents were asked to choose, on a five-point Likert scale, their degree of agreement or disagreement with the items (1 = strongly disagree, 5 = strongly agree).
The scale consists of 23 survey items (Cronbach's α = 0.85) and has a mean of 2.49 (SD = 0.57). Sample items include "I am forced to change my work habits to adapt to new technologies" and "I often find it too complex for me to understand and use new technologies".
For technostress, EFA explained 66% of the variance and the KMO measure showed a high sampling adequacy (KMO = 0.82). In addition, CFA indicated that the original five-factor model showed good and acceptable fit indices (χ 2 /sd = 1.97, CFI = 0.94, GFI = 0.90, RMSEA = 0.070). All item loadings on the theorized construct were significant at the p < 0.000 level and no modifications were conducted.
The loneliness scale was derived from Russell's (1996) measure. Respondents were asked to choose, on a four-point Likert scale, their degree of agreement or disagreement with the items (1 = never, 4 = always). The scale contains 20 survey items (Cronbach's α = 0.93) and has a mean of 1.95 (SD = 0.61). Sample items include "How often do you feel that there is no one you can turn to?" and "How often do you feel left out?".
For loneliness, the analyses showed that EFA explained 65% of the variance and the KMO measure showed a high sampling adequacy (KMO = 0.92). CFA indicated that the original one-factor model showed good and acceptable fit indices (χ 2 /sd = 2.21, CFI = 0.95, GFI = 0.92, RMSEA = 0.078). All item loadings on the theorized construct were significant at the p < 0.000 level and no modifications were conducted.
The flow scale was derived from Bakker's (2008) measure. Respondents were asked to choose, on a seven-point Likert scale, their degree of agreement or disagreement with the items (1 = never, 7 = always). The scale consists of 13 survey items (Cronbach's α = 0.92) and has a mean of 5.23 (SD = 1.09). Sample items include "When I am working, I forget everything else around me and "When I am working on something, I am doing it for myself".
According to the analyses results for flow, EFA explained 71% of the variance and the KMO measure showed a high sampling adequacy (KMO = 0.91). CFA indicated that the original three-factor model showed good and acceptable fit indices (χ 2 /sd = 1.61, CFI = 0.98, GFI = 0.97, RMSEA = 0.056). All item loadings on the theorized construct were significant at the p < 0.001 level and no modifications were conducted.

Data analysis
The PROCESS macro was used to test the direct and indirect relations (Hayes, 2013;model 6). This statistical tool is employed to conduct serial mediation analyses (Diehl, Weeks, & Gil de Zuniga, 2016). It relies on bootstrapping methods which draw subsamples from the posterior sample distribution and aggregates estimates across the samples. This method has been regarded as more efficient and less biased with regard to indirect effects (Hayes, 2009). The PROCESS macro computes the direct effects with a least-squares regression and tests the indirect effects with bootstrap confidence intervals at the same time. The number of bootstrap samples for the bias-corrected bootstrap confidence interval was set at 5,000 and the confidence level at 95% to test the indirect effects.

Common method bias test
Using common method factors may lead to bias such as halo effects, social desirability, or leniency effects (Bagozzi & Yi, 1991;Lindell & Whitney, 2001). In particular, data obtained from the same respondents with a self-reporting survey may bring about concerns regarding common methods bias . To uncover whether common method bias would have any effects on our results, a number of statistical approaches were employed, as suggested by Podsakoff, MacKenzie, Lee, and Podsakoff (2003). First, Harman's single-factor test was conducted to examine whether a single factor would emerge from the factor analysis and explain the majority of the variance (Podsakoff et al., 2003). The results showed that the first main factor only explained 18.5% of the total variance. Since this is below the recommended threshold value (<50%) (Saris & Gallhofer, 2014), the single factor result did not explain the majority of the variance. In addition, Bagozzi and Yi (2012) contended that using a single method to control common variance bias is not adequate for disentangling true variation from measurement error and method bias. Thus, the common latent factor was also tested in the current study to check for common methods bias following Podsakoff et al. (2003). To perform this approach, we inserted a common latent factor (CLF) into the structural model using AMOS. For models with and without CLF, we examined and compared the standardized regression weights of all items. To confirm that common method bias is not a major issue, the differences among regression weights should be below the recommended cut-off of 0.2 (Archimi, Reynaud, Yasin, & Bhatti, 2018;Gaudioso, Turel, & Galimberti, 2017). The regression weight differences of our two models were less than 0.2. Moreover, to reinforce this test result, we also follow the recommendation by Podsakoff et al. (2003) to control the effects of the unmeasured latent method factor. To perform this approach, we firstly conducted a CFA with the unconstrained common latent factor (CLF), and then conducted a CFA with the zero constrains common latent factor (CLF). We compared the chi-square test and the results show that there is no significant difference in the chi-square test. Based on these the above, common method bias was not considered a serious concern in the current study. As a procedural remedy for our CFA results, we followed recommendations to minimize common method bias in the design of our study (e.g. Podsakoff et al., 2003), assuring participants that their responses would be treated confidentially, using randomized items within question blocks, separating independent and moderator variables in the survey and using different response scales for different variables.
As a last step, in line with suggestions  and recent research (e.g., Bal, De Jong, Jansen, & Bakker, 2012) we conducted a marker-variable analysis (Lindell & Whitney, 2001). We did this by subtracting the lowest positive correlation between self-report variables which can be considered a proxy for common method bias, from each correlation value. Each of these values was then divided by 1 -the lowest positive correlation between self-report variables. The resulting correlation values reflect common method bias adjusted correlations. Large differences between the unadjusted and common method bias adjusted correlations suggest that common method bias is a problem. The absolute differences were relatively minimal in our sample, ranging between 0.002 and 0.001. Hence, from this perspective, it can be concluded that CMB was not an issue in our analyses.

Descriptive analyses
Means, standard deviations, and correlation numbers in relation to each variable appear in Table 2 below. The results showed that remote eworking was positively correlated with flow (r = 0.22, p < 0.01), but it was negatively correlated with technostress (r = − 0.17, p < 0.05) and loneliness (r = − 0.30, p < 0.01). Technostress was positively associated with loneliness (r = 0.27, p < 0.01), but there was no significant association found between technostress and flow (r = − 0.06, p > 0.01). Loneliness was negatively related with flow (r = − 0.26, p < 0.01). For technostress, the R 2 was 0.030 and p < 0.010. For loneliness, R 2 = 0.137 and p < 0.000. For flow, the total effects model had R 2 = 0.090 and p < 0.000. All R 2 values were found to be significant.

Hypotheses testing
The first hypothesis predicted that remote e-working was positively associated with flow. The results showed that those who had a good remote e-working experience were more likely to experience flow at work (b = 0.16, SD = 0.07, p < 0.05). Thus, Hypothesis 1 was supported. Looking at the relationships between remote e-working and technostress, as stated in the second hypothesis, those who had a good remote e-working experience tended to have lower levels of technostress (b = − 0.17, SD = 0.06, p < 0.01). This result supported Hypothesis 2. The third hypothesis suggested that technostress was negatively related to loneliness. The results showed that those experiencing technostress were likely to feel lonely (b = 0.23, SD = 0.06, p < 0.001). This supported Hypothesis 3. The fourth hypothesis contended that loneliness was be negatively related with the flow at work. The results revealed that the more employees felt lonely, the less they experienced flow (b = − 0.22, SD = 0.07, p < 0.002). Thus, Hypothesis 4 was supported. The fifth hypothesis stated that technostress and loneliness serially mediated the relationship between remote e-working and flow at work. The indirect effect of remote e-working on flow at work through the negative mediation of technostress and loneliness was significant (b = 0.009, SE = 0.005, p < 0.000). Table 3 below shows the indirect effects of remote e-working on flow as mediated by technostress and loneliness.

Key findings
In the current study, we integrated JD-R theory to develop our understanding of the influences of remote e-working on employees' flow experiences and posited that technostress and loneliness were two key stressors that had a negative impact on this key association.
Our results provided supporting evidence for our hypotheses. Firstly, our results revealed that remote e-working is positively associated with flow at work (H1). This finding expands research which explores the consequences of remote e-working which revealed that remote eworking negatively affects concentration (Vander Elst et al., 2017;Vittersø, Akselsen, Evjem, Julsrud, Yttri, & Bergvik, 2003). Our study extends this line of research by identifying flow as an important yet overlooked source that is positively associated with remote e-working. Studies have revealed flow is a crucial work-related state that can successfully determine emotional well-being during stressful and uncertain times (Rankin et al., 2019).
Second, our results revealed that remote e-working is negatively associated with technostress (H2). Employees who were satisfied with their remote e-working experience could be associated with lower levels of technostress. This finding is consistent with the evidence offered by Grant et al. (2019) suggesting that a positive remote e-working experience is associated with improved well-being and the state of one's mental health. Thus, considering technostress to be a sign of deteriorating well-being, our study extends this line of research by investigating how positive remote e-working experience may be linked with reduced levels of technostress.
Third, the results indicated that technostress is positively associated with loneliness (H3). This finding underscores the idea that employees who experience technostress will be more likely to feel lonely as Table 2 Means, standard deviations, and correlations.

Variables
Mean   (Sarabadani, Compeau, & Carter, 2020) that identifies the outcomes of emotions associated with technostress. Fourth, the present study makes an important contribution that reveals that feelings of loneliness affect employees' flow levels when working remotely (H4). A recurring finding from the stream of research on remote e-working shows that a lack of social interaction within the work environment can cause employees to suffer from loneliness during remote e-working (Larson et al., 2020;Sardeshmukh, Sharma, & Golden, 2012) and that loneliness can result in reduced employee performance (Ozcelik & Barsade, 2018). Our findings contribute to and expand on these conversations by exploring not only the consequences of remote e-work experiences but also by exploring two important stressors that result in lower levels of flow at work.

Theoretical implications
The results of the present study indicate that the JD-R framework can provide a useful understanding of the relevance of remote e-working experiences to employees' wellbeing at work. Also, the results provide valuable contributions to the existing literature. First, it utilizes the JD-R framework of resources and demands in the context of remote e-working and flow. Due to the nature of remote e-working and the mandatory circumstances of the coronavirus pandemic, it is argued that employees have had to be exposed to multiple technologies, creating anxiety and stress (Molino et al., 2020), and have frequently experienced feelings of loneliness during this period (Kniffin et al., 2021). Despite the proven negative impact of technostress and loneliness on employees, these two components have previously been neglected in the JD-R framework. Thus, the current study extends the current understanding of JD-R by identifying remote e-working as a personal resource and loneliness and technostress as job demands. By doing so, we respond to existing calls for further knowledge on the interaction of alternative job demands and personal resources (Bakker & Demerouti, 2017).
Second, the present study advances our understanding of remote eworking. Rather than treating remote e-working as a 'black box', we extend previous research by delineating the different aspects of remote e-working and their impacts on technostress. By utilizing the EWL scale , the current work represents the first study to assess the quality of the remote e-working experience and identifies three key related aspects (i.e., work effectiveness, well-being, and work-life balance) to explores how this relates to individuals' flow experience. To the best of our knowledge, such a holistic understanding of remote eworking has not previously been sought in the context of the work environment. Our results suggest that all three quality-related aspects of remote e-working are important sources of reducing technostress.
Third, the study advances existing studies on the cognitive and emotional states of employees by exploring the exact mechanism of how technostress influences loneliness. Studies in the psychology literature show that emotions are strongly associated with stress (Lazarus, 2006), which means that people have emotional reactions when they go through stressful situations. Furthermore, since the usage of ICTs is so prevalent, it was important to investigate the effects of technostress, the feeling of individual stress triggered by the use of ICT technology. While prior research into technostress has studied its effects on organizational and behavioural outcomes (Ragu-Nathan et al., 2008;Srivastava, Chandra, & Shirish, 2015), the emotional effect of technostress has mostly been ignored (Sarabadani et al., 2020). To fill this gap, the effect of technostress on lonelinessa prominent emotion which individuals have frequently experienced during the pandemicwas investigated. By doing so, this complements previous research by revealing technostress to be an antecedent of loneliness and contributes to a more clear and comprehensive understanding of the negative outcomes of stress caused by being exposed to multiple technologies.

Practical implications
The current pandemic and its consequences are still prevalent and expected to continue for a long time. The study provides important practical implications for management. Organizations, managers, and employees could well find the results useful when attempting to improve their responses to the current remote e-working trend.
First, our discussion on employees' remote e-working experiences reveals that the technological skills and capabilities of employees should be improved as remote e-working is almost certainly inevitable in the long term. Organizations should provide support and training to their employees to facilitate positive remote e-working experiences.
Second, employee work hours and means of communication should be flexible and based on mutual agreement with supervisors. Furthermore, since remote e-working inevitably leads to a culture that is 'always switched on', management should provide additional coping strategies to help employees navigate such challenges.
Third, organizations need to focus on providing a more sustainable remote e-working life with a strong emphasis on employees' well-being. Improved communication, such as personal contact with employees, may have an important impact when dealing with employees' loneliness problems. It should be noted that the remote e-working trend is likely to become the norm in the workplace and, therefore, associated managerial and organizational support is critical. To conclude, it is apparent that remote e-working practices are neither inherently good or bad, and their success depends on the ways in which the proliferation of technology are managed and experienced (Anderson & Kelliher, 2020).

Limitations and future research directions
The present study has a number of limitations that may also provide useful ideas for further advancement of research in this field. The crosssectional nature and limited sample prevent us from generalizing our results. In addition, participant characteristics and social desirability limitations should be taken into consideration in the evaluation of the results. It is recommended that the same research design be used to compare the findings with other industries to enhance the explanatory power of the model.
Future studies can further extend this research by addressing several of the limitations to our research. First, future research could expand the current scope by including other psychological variables such as detachment, resilience, and burnout, and can adopt a broader perspective on remote e-working and its consequences. Finally, future longitudinal research can be conducted to increase the generalizability of the results. The longitudinal impact of the suggested variables on employees' flow experiences would contribute to a further understanding of a positive remote e-working experience.

Conclusion
The present study has contributed new knowledge about remote eworking and its effects on employee flow experiences. The aim of the study was to extend remote e-working research by introducing two key stressors: technostress and loneliness. The findings of this research reveal the varying impacts of the remote e-working experience perceived by employees on their psychological outcomes. Accordingly, remote e-working was found to be positively associated with flow at work (H1). The explanation for this association was provided via two key variables: technostress and loneliness. The findings suggested that remote e-working arrangements by organizations are key variables that influence the reduction of technology-induced employee stress (H2). Also, it revealed that high levels of technostress can be associated with employees feeling greater levels of loneliness (H3) and these higher levels of loneliness result in reduced extent of flow (H4).
While five of the four (H1, H2, H3, H4) hypotheses correspond to the direct effects among remote e-working, technostress, loneliness, and flow; the findings point that technostress and loneliness serially mediate the relationship between remote e-working and flow at work (H5). Buttressed by decreased technostress and pillared by reduced loneliness, remote e-working practices by organizations culminate in increased flow experience of employees.
The findings have contributed to the related literature by enhancing the understanding of remote e-working experiences. Given the swift and extensive transition to working from home during the pandemic, it seems that remote e-working will remain a critical issue on the agenda of organizations. Therefore, organizations need to create opportunities to improve the technological knowledge and abilities of their employees to adopt ICTs and overcome the technostress that can be associated with loneliness and low levels of flow.

Declaration of competing interest
We declare no conflict of interest. This manuscript is only submitted to Computers in Human Behavior and is not under review anywhere else.