The following sections summarize previous work related to our research questions such as studies on wellbeing and social interaction in telework (RQ 1), potential and challenges of digital systems for remote interaction (RQ 1, RQ 3) and social norms in digital collaboration (RQ 2).
3.1 Wellbeing and social interaction in telework
Besides effects of telework on performance and productivity (e.g., Sánchez et al.
2007), researchers have shed light on the relationships between working from home and the facets of physical and psychological wellbeing related to a broad range of issues. This includes, for example, a study of mental health and stress symptoms among teleworkers compared to office-workers (Mann and Holdsworth
2003), consequences for organizational commitment, identification, and job satisfaction, team dynamics such as knowledge-sharing, or implications for work-family issues (Allen et al.
2015; Vittersø et al.
2003). Overall, the detected correlations and areas of risk can be considered highly dependent on the extent and specific design of telework. For example, Gajendran and Harrison (
2007) found rather negative effects on relationships with coworkers when more than half of working hours were spent teleworking. Baruch (
2000) reported that teleworking did not change conceptions of oneself as an employee as long as teleworking was balanced out by work in the ‘standard mode’. Later studies focused on specific mediators, identifying increased role ambiguity and reduced support and feedback as crucial for the negative effects of telework. Conversely, positive effects are related to increased autonomy (Gajendran and Harrison
2007; Sardeshmukh et al.
2012). Moreover, the quality of communication between team members may change in the context of telework. For example, contrasts of face-to-face and computer-mediated collaboration have revealed advantages for computer-mediated contexts regarding productivity, but a loss of socio-emotional content (Sauer et al.
2000). In sum, along the spectrum of qualities of work, from productivity and effectiveness on the one hand to experiencing team spirit, pleasure and work satisfaction on the other, the potential drawbacks of telework tend to concern the latter group of more experiential qualities (e.g., Baruch
2000; Mann and Holdsworth
2003; Sardeshmukh et al.
2012).
Going beyond direct effects of teleworking, researchers have also examined relevant contextual factors. In recent years, great emphasis has been placed on workplace culture as a crucial determinant of the success of flexible working arrangements (Abendroth and Reimann
2018). Based on linked employer–employee data for 4898 employees in large organizations in different industries, the authors conclude “how important it is to implement telework in a way that not only accommodates employers’ interest in flexibilization, but that it also makes it possible to reconcile work with a family life that involves high levels of responsibility” (Abendroth and Reimann
2018, p. 324). During the Covid-19 pandemic, integrating telework and private life became even more challenging, and digital culture, digital leadership concepts and team were found to be an essential element of responsible leadership. While earlier research on teleworking often focused on individuals working remotely or the effects of teleworking a few days a week (e.g., Baruch
2000), during the current period, working-from-home and existing as a team solely via digital structures has become standard. At the same time, the psychological side effects of telework (and digital collaboration in general) are becoming increasingly complex, requiring personal skills and resources far beyond basic skills such as time management (e.g., Baruch
2000; Beigi et al.
2018). As such, and although the many types of advanced videoconferencing software with numerous functionalities offer a multitude of opportunities, they also provide new space for misunderstandings and not-yet-established norms.
With regards to the present study focus, these previous findings on wellbeing and social interaction in telework are of relevance in several respects. First, previous findings show that employees’ wellbeing is dependent on the specific design of telework, highlighting the responsibility of employers and leadership to provide the best possible frame. This may include the used technology and introduced communication structure as well as aspects of team culture. As an advancement to previous research, the present study focuses on the particular role of social norms. Second, previous studies highlight the new challenges for the individual referring to the psychological side effects of telework such as dealing with (undetected) social norm conflicts among colleagues. Again, the pandemic context and sudden shift to a much higher share of telework makes research on how to support individuals’ wellbeing in the context of telework more timely than ever.
3.2 Potential and challenges of digital systems for remote collaboration
From the first days of telework until now, researchers have explored to what degree technologies such as videoconferencing actually support effective, satisfactory collaboration between people communicating remotely (e.g., Biehl et al.
2015; Egido
1988; Olson and Olson
2000; Whittaker
1995). Typically, conversation quality over mediated channels is compared to face-to-face meetings, including studies of objective conversation characteristics, such as fewer contributions by individuals in remote compared to face-to-face conversations, difficulties noting peripheral cues, pointing to things or manipulating real-world objects, or having side conversations, (Isaacs and Tang
1993), as well as subjective measures, such as remote team members feeling less connected than their co-located peers (Biehl et al.
2015).
Other qualitative differences between remote and co-located collaboration may be harder for people to detect and verbalize. Even if a technology theoretically provides a satisfactory substitute for the information transfer required for a given task, in practice, people may still circumvent remote collaboration if possible—presumably because they do not experience the technology as a satisfactory substitute on a psychological level. For example, Olson and Olson (
2000) reported that instead of using technology-mediated collaboration channels, people instead reorganized work so that they did not have to engage in close collaboration with a remote team member (Olson and Olson
2000). Similarly, Egido (
1988) assumes that the main reasons why the uptake of videoconferencing in the 1980s remained below expectations lie beyond the scope of technological and economic analysis, and instead concern psychological and sociological factors, such as reduced opportunities for informal interactions. Indeed, many of the envisioned use cases and associated reductions in traveling costs were not realized. As Egido (
1988, pp. 18/19) argues, the fact that “it is often over informal chats outside of official meeting rooms that important information is transmitted and real decisions are made” (e.g., Mintzberg
1973) “makes teleconferencing unattractive to politically savvy employees”.
Since then, many additions to standard videoconferencing, such as chat, polling tools, and interactive (online) whiteboards have been developed and have become a routine form of communication for teleworkers. Simultaneously HCI researchers have tried to address the shortcomings of traditional remote technology and have developed design solutions to meet specific requirements. One such path is designing different types of embodied telepresence devices that seek to mitigate the social disadvantages of video-based communication by providing remote users with a local embodiment, ranging from fully mobile robots to smaller devices that sit on a meeting room table (Biehl et al.
2015). Ideally, such an embodied representation of a remote person will establish a reciprocal sense of “being there” and harmonize interaction between remote and co-located team members. Indeed, Biehl et al. (
2015) found that local participants interacted with a remote participant in a fashion more similar to face-to-face when using an embodied technology—potentially due to an increased perception of the remote person’s agency.
A technically even more technologically advanced variant of telepresence is the Room2Room system and its life-size virtual projections of the remote person (Pejsa et al.
2016), which is intended to create an illusion of the remote person’s physical presence in the local space and support a shared understanding of verbal and non-verbal cues (e.g., gaze, pointing) In their evaluation study, Pejsa et al. (
2016) compared their Room2Room system to the established teleconferencing system Skype. While using the Room2Room system led to an enhanced feeling of the communication partner’s presence, participants did not have an opportunity to see what their communication partner was seeing. While the Skype interface shows the view from both the current user’s camera and their conversation partner’s camera, the Room2Room system lacks such a feature. This in turn makes it more difficult to gauge projection quality on the other end and assess whether one’s communication partner can see one’s nonverbal cues (Pejsa et al.
2016, p. 1723). This example shows that the limitations and also specific psychological benefits of such systems often first become apparent in the actual usage situation, and that the requirements for satisfactory communication can be difficult to foresee.
In sum, despite recent technological advances as virtual reality projections, communication and collaboration over mediated channels will always have some limitations, and distance still matters (Olson and Olson
2000). As outlined by Olson and Olson (
2000), this becomes visible when considering many basic characteristics of co-located interaction that affect communication quality. These include “shared local context”, i.e., participants experiencing similar conditions regarding the time of day or local events, which provides a framework a frame for easy socializing and a mutual understanding of what is on the other person’s mind. “Informal hall time before and after” is a basis for opportunistic information exchange and social bonding. “Personal information” means that contributors’ identities are usually known and can be taken into account. In sum, the full list of subtle but meaningful qualities of co-located interactions demonstrates how difficult it is to simulate such communication such communication via mediated channels and to establish what Olson and Olson (
2000) refer to as “common ground” (Clark
1996). Common ground refers to the knowledge that communication partners have in common and are aware that they have in common. As laid out by Olson and Olson (
2000, p. 157), we typically have specific assumptions about what different people know and use these to frame what we say: “People describe the same event or idea quite differently talking to a spouse, a coworker, a distant relative, a neighbor, a stranger from across the country, and a stranger from overseas […] We establish common ground not just from some general knowledge about the person’s background but also through specific knowledge gleaned from the person’s appearance and behavior during the conversational interaction itself. If we say something based on an assumption about what someone knows, but their facial expression or verbal reply indicates that they did not understand us, we will revise our assumptions about what common ground we share and say something to repair the misunderstanding”.
In remote communication, this quality of common ground, and hence opportunity to identify and repair misunderstanding, is limited. Various communication cues are missing depending on the channel. The fewer cues there are, the more likely misinterpretations become—and the conversation might move on before the conversation partners are able to reach correspondence (Olson and Olson
2000, p. 158).
Although various approaches to enriching digital communication with more communication cues exist (e.g., emoticons in chat conversation), such design “solutions” can also lead to more misunderstandings (e.g., Miller et al.
2016). Often, there is no common ground on how to use and interpret such communication cues in the digital realm, as they are usually expressed spontaneously and unconsciously in the face-to-face context. Already in the 1990s, Isaac and Tang (
1993) described the problems that arise if a videoconferencing system forces people to take explicit actions to engage in behaviors that normally take place unconsciously. Moreover, artificial behaviors may be interpreted differently by other participants: “For instance, a person who would have been seen as enthusiastic might be perceived as dominating if she uses an explicit mechanism [here: a visual mechanism] rather than a socially negotiated one to manage floor control” (Isaacs and Tang
1993, pp. 204/205). Due to the limited opportunities to apply social norms from face-to-face conversation, technology-mediated meetings often include a lot of organizational speech rather than real, content-related communication, people develop behaviors to compensate for shortcomings (e.g., always identifying oneself before speaking, more formal protocols for turn-taking), and more explicit discourse rules are needed (Olson and Olson
2000, p. 153).
In sum, the studies cited above reveal a strive for ever better and more advanced technical tools for remote collaboration, but also still existing pitfalls and potential for interpersonal conflict. Though not focused explicitly, the lack of shared social norms and discourse rules appears to be a recurring issue in many of these studies. Our research picks up this subject for deeper analysis and further connects it to interventions and technology design. In the following paragraphs, we summarize previous studies on social norms in digital collaboration and other fields of HCI beyond the work context.
3.3 Social norms in digital collaboration and other fields of HCI
Social norms are informal understandings that govern the behavior of societal members, i.e., shared beliefs regarding appropriate ways to feel, think and behave (Turner
1991). Social norms help us to interpret others’ behavior and react appropriately (Horne
2001). In contrast to legal norms (e.g., laws), which are created in a deliberate process, set down in written tests, and linked to specific sanctions and enforcement mechanisms, social norms occur spontaneously rather than being planned deliberately, usually remain unwritten, and are enforced informally (Hechter and Opp
2001). Therefore, we are usually not aware of the many unwritten laws underlying our behavior, which only become evident when conflict arises: if someone behaves in a way that contradicts our informal understanding of what is appropriate, such as cutting in line, entering an office without knocking, or starting to eat before everyone is seated at the table. Since the beginning of mankind, such norms have naturally developed from interpersonal interactions (Horne
2001; Popa et al.
2014), although the typical contexts of social interaction have changed over time—from gathering around the hearth to today’s meetings in chatrooms and other digital spaces. Understandably, whenever a new social interaction context emerges, conflicts may occur in the beginning, since established norms do not yet exist and people are unsure which norms may be transferrable from other contexts (Diefenbach and Ullrich
2018).
Accordingly, as communicating via email and the Internet became increasingly common in the work environment, many conflicts and confusion around social norms for communication technologies arose, and researchers debated the need for “digital etiquette” (e.g., Preece
2004). Indeed, no shared norms seem to have been established yet even with respect to basic questions such as the preferred salutation in email communication. This was illustrated by Diefenbach and Ullrich (
2018, p. 51 ff.), who highlighted contrasting recommendations about appropriate email salutations: An interviewee in the study by Preece (
2004, p. 58) clearly stated that “… not addressing me by my name and ending without a farewell greeting and the sender’s name—that’s rude and unfriendly”. Later on, an online discussion around the question “How do you address someone in your reply to their emails?” (Ars Technica Open Forum
2014) showed the norm of addressing people by name in email is quite controversial, with vastly different reasons given for (not) addressing others by name in email communication. For example, one user sees addressing someone by name as an anachronism, stating that “No name whatsoever. I just reply without any kind of anachronistic salutation line. Death to the letter and everything it stood for.” Another user emphasizes on how not addressing one’s counterpart by name might even be a sign of valuation and intimacy: “In personal email that’s a reply, I often don’t even use the name. You’ve emailed me. You know me. I know you.” In a similar vein, another user draws a parallel to face-to-face conversation: “I don’t think I have ever put a formal greeting in an email, even back in olden times. Of course, I don’t use someone’s name when talking to them in person either, they know who they are, no need to remind them.” Finally, another forum user comments that “It depends entirely upon the relationship and what kind of email”, i.e., there are no general rules. Hence, what one person experiences as rudeness could actually be a sign of friendship for another (Diefenbach and Ullrich
2018, p. 51).
Beyond the email context, numerous studies have explored the evolution of social rules in digital spaces and newly emerging communication channels such as social media (e.g., Moncur et al.
2016; Sabra
2017; Voggeser et al.
2018), videoconferencing and systems for mobile remote presence in the workplace (Lee and Takayama
2011), chat tools in online libraries (Radford
2006), online gaming (Martey and Stromer-Galley
2007), or online communities (Kim
2000). Also, researchers have examined to what degree social norms from interpersonal interaction (e.g., eye contact) are transferred to digital environments such as videoconferencing (Yee et al.
2007), as well as social dilemmas resulting from mixed contexts and dealing with social norms from the physical and virtual contexts at the same time (e.g., answering mobile calls in various circumstances (Inbar et al.
2014)). In sum, universally shared norms of social interaction in digital spaces are rare, with divergences in norms often associated with different sub-cultures or user groups, e.g., between online library employees and customers (Radford
2006).
Thus, in many digital contexts, there still is no consensus about what is right or wrong and appropriate or inappropriate behavior (Diefenbach and Ullrich
2018; Tene and Polonetsky
2014). According to Tene and Polonetsky (
2014, p. 60), with the new types of socialization resulting from technological innovation, we often have “nothing more than a fleeting intuition as to what is right or wrong” and “intuitions and perceptions of how our social values should align with our technological capabilities are highly subjective”, driving us into a “techno-social chaos” (Tene and Polonetsky
2014, p. 60). As described by Diefenbach and Ullrich (
2018, p. 44 ff.), the gradual development of routines around technology use, particularly in the field of communication technology, often breaks up such established social norms. A typical example is the commitment to fixed appointments (old norm) versus cancelling or delaying an appointment at the last minute, which can be easily done nowadays via text message. While the change and evolution of social norms over time is a natural process, the speed of this change has increased dramatically. In many cases, the speed of technological progress may exceed the tempo at which a society typically develops consensus about the informal rules governing societal life.
From this perspective, a lack of social norms could also be a risk for digital communication in the telework context, potentially harming relationships between colleagues, teams and customers due to the lack of a feeling of unity. While social norms and shared standards for social interaction have been acknowledged as crucial for team spirit and work motivation (e.g., Ehrhart and Naumann
2004; Ellemers et al.
2004), there are a lack of studies that consider this issue in light of our increasingly digital communication structures, such as in the context of telework. Shared norms can form a foundation for cohesion and positive social interaction experiences in the digital space as well (e.g., Diefenbach and Ullrich
2018; Hong et al.
2019). In this sense, social norms can be considered a potential source of conflict, but also an important resource in the context of telework. Accordingly, a particularly interesting question is to what degree technology itself could support positive or shared norms. First experiments on how technology could support shared communication rules and other behavioral norms in daily life include different fields, such as norms of punctuality when meeting friends (Diefenbach et al.
2017; Diefenbach and Ullrich
2018), norms of salutation in online forums (Mooseder
2018), or norms in mobile phone communication (Inbar et al.
2014).
With regards to the present study and its advancements to previous research, the above cited examples suggest social norms as a promising starting point to understand and improve digital interaction. However, so far, this has rarely been investigated in the telework context, where in fact the business software used could be an interesting resource and starting point for interventions. Therefore, in this survey of teleworkers, a particular focus was placed on social norms, providing a conceptual frame to understand the dynamics of conflict in digital communication in the telework context. The next sections present the survey and questions that were used to gain deeper insights into employees’ daily life experiences in the context of digital communication and the pandemic-induced working-from-home situation.