Connecting work teams in a remote workplace: an identity leadership perspective

ABSTRACT

Accelerated by the COVID-19 pandemic, a shift towards remote and hybrid working practices is currently taking place in organizations worldwide. This transition creates an important challenge for workplace management to ensure that remotely working employees feel connected to their team, and the organization at large, despite the loss of frequent in-person, face-to-face interactions. We propose that identity leadership behaviours from workplace supervisors are key to ensuring such feelings of connectedness among remotely working employees. We conducted two studies among working adults in the United Kingdom to demonstrate the importance of identity leadership in a remote working context. Our first study showed that employees preferred identity leadership behaviours from their supervisors more in a remote versus an on-site working context. Results of our second study showed that identity leadership behaviours mitigate against the detrimental consequences of remote working on team connectedness and on various work-related outcomes for employees, such as their work engagement, job satisfaction, work stress and absenteeism. We discuss the theoretical and practical implications of these findings as well as directions for future research.

KEYWORDS:

• Remote working
• identity leadership
• team connectedness
• work experiences

Introduction

Due to the digital transition of the workplace, and accelerated by COVID-19 pandemic, a shift towards remote working is currently taking place in organizations worldwide. Rather than working on-site in central offices, employees work from home more and collaborate with their co-workers through increasingly sophisticated digital connectivity technologies. Experts anticipate that by 2027, over 25% of all professional workers in the United States (Roubinson, 2022), Europe (Ahrendt et al., 2020), and Asia (Afuang et al., 2022) will work remotely, either by working-from-home or working-from-anywhere.

This shift towards remote working, or a hybrid version thereof (whereby employees partly work-from-home, and partly on-site), can have important benefits for both employees and organizations. Employees who work remotely, for example, benefit from having less rigid work schedules and reduced commuting times (e.g., Aksoy et al., 2023; Allen et al., 2015), and organizations profit by having to make fewer costs in physical infrastructure (e.g., offices, parking space; Tursunbayeva et al., 2022). Remote working has therefore long been championed as a force for good (for reviews, see Allen et al., 2015; Gajendran & Harrison, 2007), and initial research, carried out before the COVID-19 pandemic, showed that working remotely is associated with increased productivity and job satisfaction and reduced stress, work-family conflict, and turnover intentions (e.g., Gajendran & Harrison, 2007; Harker Martin & MacDonnell, 2012; Shockley & Allen, 2007).

More recent studies, carried out during and after the COVID-19 pandemic, suggest, however, that there are also substantial downsides when organizations allow, or sometimes force, their employees to work remotely (Franken et al., 2021; Kniffin et al., 2021). In particular, a growing body of work suggests that working remotely is associated with increased feelings of isolation and loneliness, because the frequency of in-person, face-to-face contact between co-workers is significantly reduced (e.g., Brown & Leite, 2022; Van Zoonen & Sivunen, 2022; Wood et al., 2022). The transition towards a remote or hybrid workplace thus creates an important challenge for workplace management and leadership: How to ensure that remotely working employees feel connected to their teams?

Organizational researchers and practitioners have long recognized leadership as a critical condition for the success of remote working practices by ensuring smooth collaboration and communication between co-workers in the absence of regular face-to-face contact (e.g., Avolio et al., 2000; Bell et al., 2019; Larson & DeChurch, 2020). The social identity approach to leadership (Hogg, 2001; Hogg et al., 2012; Van Knippenberg & Hogg, 2003), in particular, suggests that identity leadership plays an important role in cultivating and strengthening the shared sense of togetherness among remote workers (Leonardelli, 2022; Van Bavel et al., 2020). Rooted in social identity theory (Tajfel et al., 1979) and self-categorization theory (Turner, 1991; Turner et al., 1987), the social identity approach to leadership was first introduced by Hogg (2001) and defines identity leadership as a social influence process in which the effectiveness and influence of leaders are determined by their ability to foster and enhance a shared social identity among team members (Haslam et al., 2011; Hogg, 2001; Steffens et al., 2014).

Leaders can create such a social identity, in their teams by (1) representing the team’s unique qualities (i.e., “being one of us”; cf. identity prototypicality), (2) promoting the team’s shared interests (i.e., “doing it for us”; cf. identity advancement), (3) defining core values, norms, and ideals of the team and thereby creating a shared sense of “we” and “us” within the team (i.e., “crafting a sense of us”; cf. identity entrepreneurship), and/or by (4) developing structures, events, and activities that give weight to the team’s existence and provides team members with an opportunity to live out their group membership (i.e., “making us matter”; cf. identity impresarioship; Haslam et al., 2011; Steffens et al., 2014). Similar to other styles of leadership (e.g., Gerpott et al., 2019; Lehmann-Willenbrock & Allen, 2018; Lehmann-Willenbrock et al., 2015), identity leaders can engage in these behaviours through both in-person events (e.g., office parties, plenary meetings) and in online activities (e.g., online social gatherings, team management interventions, and professional development workshops; cf. Blanchard & Allen, 2023; Haslam et al., 2023). A growing body of research has shown that these identity leadership behaviours reduce team members’ feelings of isolation (Krug et al., 2021) while promoting team identification (e.g., Fransen et al., 2020; Milesi, 2022; Steffens et al., 2014).

Building on this important line of work, we propose that identity leadership contributes to establishing and facilitating feelings of connectedness in work teams, especially when employees work remotely. We therefore argue that when employees work remotely, they will show a greater preference for an identity leadership style from their supervisor compared to when they work on-site (e.g., Fransen et al., 2015; Krug et al., 2020). Second, we hypothesize that identity leadership will help to buffer the negative impact of remote working with regard to enhancing the degree to which employees identifying with, and feeling connected to, their team. Third, we aim to show that identity leadership mitigates against a decline in overall work outcomes experienced by remotely working employees, such as employees’ job engagement, job satisfaction, work stress and absenteeism. Figure 1 describes our overall conceptual model.

Figure 1. The overall conceptual model.

Our research on identity leadership in remote work settings potentially makes important theoretical and practical contributions. First, our study contributes to implicit leadership theories (Lord et al., 1984, 2020; Offermann et al., 1994) by showing that employees favour different types of leadership behaviours when working remotely rather than on-site. Second, our research challenges the consensus that remote working is, in principle, a force of good (Allen et al., 2015; Gajendran & Harrison, 2007), and adds to an increasing body of literature that highlights that remote working has important downsides in nowadays workplaces where such arrangements have become increasingly prevalent, especially for the loss of team connectedness (e.g., Brown & Leite, 2022; Van Zoonen & Sivunen, 2022). Third, our study extends identity leadership research by showing the critical role of identity leadership in a remote, digital, workplace by buffering against some negative consequences of remote working arrangements on employees. Finally, our research provides valuable practical insights into how workplace supervisors can best manage remotely working employees within the team – suggesting that supervisors should pay much attention to identity leadership activities by representing, advancing, creating, and embedding a shared sense of social identity among team members. These leadership behaviours can prevent remotely working employees from experiencing a loss of connectedness, thereby making the transition towards more remote working a success.

Theory and hypotheses development

The social identity approach to leadership

The social identity theory, and its approach to leadership, provides an appropriate theoretical framework to understand the potential challenges of remote working for workplace teams. The core tenet of social identity theory contends that individuals not only define themselves in terms of their idiosyncratic personal characteristics (i.e., their personal identity, “I”) but also in terms of the groups that they belong to (i.e., their social identity, “we”; Hogg & Terry, 2000; Tajfel et al., 1979; Turner et al., 1987). Building on this, the social identity approach to leadership suggests that leaders and followers are connected through a shared social identity (Haslam et al., 2011; Hogg, 2001; Steffens et al., 2020), which offers a unique understanding of leadership that originates from social identity-based perceptions of the leader as a group member (Hogg, 2001; S. D. Reicher et al., 2018; Steffens et al., 2014).

At the core of the social identity approach to leadership is the assumption that leaders are able to influence team processes to the extent that followers perceive leaders as prototypical group members, in other words, as individuals who represent and embody the group’s attributes, norms and values (Hogg, 2001; Van Knippenberg & Hogg, 2003). To this end, leaders who establish themselves as being representatives, and acting on behalf, of their group, are able to develop social bonds with followers and motivate them to contribute to group goals.

Research further suggests that leaders can engage in different identity leadership activities to develop a stronger social identity in work teams. First, identity leadership of supervisors can offer a model of what it means to be “one of us” within the team, which they can do by being an ideal representative of the team vis-à-vis other teams within the organization (identity prototypicality; Steffens et al., 2014, 2020). Moreover, supervisors can stimulate the team’s identity advancement by acting for their team and working towards the common goals of the team rather than focusing on their narrow personal interests (identity advancement; Haslam et al., 2011; Hogg et al., 2012). Third, supervisors engaging in identity leadership shape a common understanding of what “we” stands for in a working team and develop a shared sense of “who we are” that promotes team cohesiveness and inclusiveness (identity entrepreneurship; S. Reicher et al., 2005). Finally, identity leadership of supervisors who serve as identity impresarios can exert their influence by organizing events and activities that give employees a sense that their team exists and that team membership matters (identity impresarioship; Haslam et al., 2011).

The context of remote working

Here, we focus on the impact of identity leadership in a remote working context. Remote working constitutes a relatively novel working arrangement, characterized by having fewer face-to-face interactions with physically distant co-workers, whereby the on-site interactions are replaced by digitally mediated communications via, for instance, emails or video meetings (DiMartino & Wirth, 1990). It is good to note the differences with how humans have traditionally worked. Throughout human evolution, humans have worked together in small groups of hunters and gatherers with whom they had frequent face-to-face interactions and felt emotionally strongly connected (Dunbar, 2018; Van Vugt & Schaller, 2008). Face-to-face contact is therefore an essential cue to social influence and social connectedness among co-workers (cf. Latané, 1981; Van Vugt, 2017).

The massive shift towards remote working in response to the COVID-19 pandemic aided by increasingly sophisticated digital work tools (e.g., video conferencing), has drastically reduced the amount of such face-to-face interactions at work and has reshaped how employees work together (Bell et al., 2019; Kniffin et al., 2021). Scholars have long argued that remote working makes it more difficult for employees to connect to their team as usual due to the loss of co-located, face-to-face interactions with team members (e.g., Ashforth, 2020; Thatcher & Zhu, 2006; Wiesenfeld et al., 2001). As a result, these changes towards remote working may have important drawbacks for employees’ feelings of team connectedness (Leonardelli, 2022; Rock & Pratt, 2002), which we define as the extent to which an employee identifies with, and feels connected to, their team (Peluso et al., 2019).

We build on implicit leadership theories (Lord et al., 1984, 2020, Lord & Maher, 1991) to argue that employees value identity leadership more in a remote work setting compared to an on-site work setting. Implicit leadership theories argue that people have strong (unconscious) beliefs about what traits, skills, and abilities their ideal leaders possess (e.g., Lord et al., 1984, 2020; Offermann et al., 1994). Moreover, this line of research suggests that such implicit mental models are context specific, because different situations require different leadership behaviours (Epitropaki et al., 2013; Lord et al., 1984, 2001; Van Vugt & Grabo, 2015). Scholars increasingly recognize that remote workplaces necessitate leaders to employ a unique set of skills and qualities to facilitate the coordination of the team’s work and build strong relationships with team members (e.g., Bell et al., 2019; Joshi et al., 2009; Liao, 2017). An increasing body of research demonstrates that supervisors who engage in such activities to connect their team members are especially effective in remote working conditions (e.g., Joshi et al., 2009; Purvanova & Bono, 2009). Following this line of reasoning, in remote working contexts we argue that the need to ensure and enhance a sense of social connectedness among distantly operating team members will result in the activation of a mental model of the ideal supervisor as someone who engages in identity leadership behaviours. This should result in a stronger endorsement of identity leadership, especially when employees are working remotely more often. We thus hypothesize:

Hypothesis 1:

Employees endorse identity leadership more in a remote work setting than in an on-site work setting.

Furthermore, our research attempts to replicate and extend prior research on the detrimental consequences of remote working – suggesting that remotely working employees are more likely to feel isolated and disengaged from their team (e.g., Golden et al., 2008; Leonardelli, 2022; Van Zoonen & Sivunen, 2022; Wiesenfeld et al., 2001) – by examining the effects of remote working on employees’ feelings of team connectedness. We thus hypothesize that the more employees work remotely the weaker will be their team connectedness, whereby we operationalize team connectedness in two ways (Peluso et al., 2019): (a) the extent to which employees identify with their team and (b) the extent to which they feel connected with (i.e., less isolated from) their team.

Hypothesis 2:

Remote working is negatively associated with employees’ team connectedness.

Identity leadership in remote work settings

So far, we have argued that the shift towards remote working comes with the potential loss of team connectedness. This thus poses a key challenge for supervisors on how to best lead and manage employees who might lack a sense of belongingness and experience social and professional isolation within the team. Building on the social identity approach to leadership (Hogg, 2001; Hogg et al., 2012; Van Knippenberg & Hogg, 2003), we propose that supervisors can overcome the downsides of remote working, by cultivating a sense of shared social identity within remote working teams, in other words, by engaging in identity leadership behaviours along the four facets we discussed earlier (identity prototypicality, advancement, entrepreneurship, and impresarioship).

The diverse range of identity leadership behaviours fostering the togetherness can influence how employees think, feel, and act (Haslam et al., 2011; Hogg et al., 2012). Research has found, for instance, that identity leadership enhances employees’ affiliation with their co-workers (e.g., Fransen et al., 2020; Steffens et al., 2014) as well as fosters a sense of intimacy and connectedness among work teams (S. D. Reicher et al., 2018). However, this has been shown in an on-site working environment. We argue that identity leadership will be especially effective in remote work settings as employees are seeking ways to connect themselves with their team in the absence of regular face-to-face interactions with co-workers (S. D. Reicher et al., 1995).

Specifically, with regard to identity prototypicality, when supervisors represent and embody the unique qualities that define their team, this may help remotely working employees perceive their team members as more similar to each other and therefore more trustworthy (Steffens et al., 2020). In addition, when supervisors in remote working contexts are seen to promote and advance their team’s interests, it can help foster greater team connectedness (Steffens et al., 2014). Moreover, when supervisors under remote working conditions act as identity entrepreneurs, this can help disseminate the relevant social norms, values, and practices within the remotely operating teams which may further promote team connectedness (Krug et al., 2021; cf.; Op ‘t Roodt et al., 2021). Lastly, research shows that supervisors who organize frequent team meetings online as well as online and in-person social events facilitate the connectedness of remotely working employees (Krug et al., 2021), which may be seen as examples of identity impresarioship.

Accordingly, we argue that identity leadership will be particularly effective in remote work settings as it has the potential to address one of the main drawbacks of working remotely, that is the feeling of being disconnected from co-workers (Leonardelli, 2022; Rock & Pratt, 2002). We thus propose that identity leadership behaviours will mitigate against the negative effects of remote working regarding employees’ feelings of team connectedness. In other words:

Hypothesis 3:

Identity leadership moderates the effect of remote working on employees’ team connectedness such that this negative association is weaker when identity leadership is higher instead of lower.

Implications for work experiences

Research further suggests that remote working, if not managed well, results in a range of negative work experiences and outcomes for employees, such as a lower job performance, decreased work satisfaction, as well as higher withdrawal and work-life conflict (e.g., Charalampous et al., 2019; Franken et al., 2021; Hill et al., 2022). For example, Galanti et al. (2021) found that employees who experience increased feelings of isolation and job stress are less engaged when working remotely compared to when working from home. Similarly, Andel et al. (2021) found that the lack of connection in remote working practices limits employees’ task performance and reduces their job satisfaction. Furthermore, McCloskey and Igbaria (1998) showed that remote work results in higher levels of job dissatisfaction and reduces organizational identification.

Following the argument we made about the supporting role of identity leadership, we propose that identity leadership behaviours exercised by supervisors mitigate against the negative effect of remote work on employees’ work experiences through helping to maintain team connectedness (cf. Op ‘t Roodt et al., 2021). A recent meta-analysis on identity leadership by Steffens et al. (2020) supports such reasoning as they show that managers who engage in identity leadership behaviours foster togetherness among team members, help employees to (a) achieve good task performance, (b) feel more satisfied with their work, (c) be less stressed with their work, and (d) feel more committed to their organization. Building on this work, we therefore hypothesize:

Hypothesis 4:

Identity leadership moderates the negative indirect effect of remote working on employees’ work experiences (i.e., decreased work engagement, job satisfaction, and increased work stress and absenteeism) via their team connectedness, such that this indirect effect is weaker when identity leadership is higher instead of lower.

Overview of study designs

We conducted two studies with different designs to test our hypotheses. Study 1 tested Hypothesis 1 by examining whether employees endorse identity leadership behaviours more in remote versus on-site work settings by means of a vignette design. In Study 2, drawing on the obtained results from Study 1, we further explored the role of identity leadership in navigating the impact of remote working on employees’ team connectedness and subsequent work experiences, specifically examining Hypotheses 2–4 in a field survey. Both studies were approved by the authors’ affiliated Institutional Review Board (Protocol VCWE-2020-053), and their relevant informed consent has been obtained. The materials used in both studies are included in the Supplemental Material.

Study 1

Method

Participants and procedure

We recruited participants from the United Kingdom in June 2022, through the online platform Prolific Academic (https://www.prolific.co). Participants could take part when they worked full-time, had a supervisor within their team, and had recent experiences with working remotely.¹ An a priori power analysis showed that we needed 176 participants to detect a medium correlation (d = .50), with 95% power and an alpha level of .05. We oversampled with a 15% addition of the proposed sample size and thus requested 202 participants to complete our study. After we excluded 19 participants due to incomplete responses and two participants due to failed attention checks, the final sample consisted of 181 full responses (response rate = 90%). Of the final participants, 50% were female, their average age was 37.38 years (SD = 8.07), and their average working tenure was 7.99 years (SD = 6.81). Participants had an average of 2.74 years (SD = 1.73) experience of working remotely. Participants held diverse jobs from a variety of industries, including education and culture (16%), information technology (15%), and financial services (13%).

We asked participants whether they found a list of leadership behaviours that their supervisors might engage in more or less important in an on-site versus a remote work setting. Notably, following previous research (e.g., Moore et al., 2023; Schoel et al., 2011; Van Vugt & De Cremer, 1999), we employed a bipolar scale to ask participants to compare leadership behaviours in remote versus on-site work settings. This approach aims to make the contrast between the different work settings salient and encourages participants to think about whether a particular leadership behaviour is relatively more important in a remote or on-site work context. In addition to identity leadership, we also included measures of task-oriented, prototypical leading, and managing behaviours to find out if these other types of leadership were more valued in remote versus on-site work settings as well. We focused on these leadership behaviours because task-oriented and managing behaviours have been recognized as especially important when leading teams in remote settings (Bell et al., 2019; Kniffin et al., 2020; Liao, 2017). Besides, we wanted to know if employees would value prototypical leading behaviours (such as inspiration and building confidence) more in remote work settings. Finally, participants completed a measure of their demographics.

Measures

We assessed participants’ endorsements of a list of leadership behaviours with the following instructions: “For each behaviour, please indicate the extent to which you find it less or more important that your supervisor acts in this manner when you and your team members are working on-site compared to working remotely. Please note that we ask for your ideals, and not the actual behaviour of your supervisor.” All leadership behaviours were rated on a 7-point Likert scale ranging from −3 (“this is more important when working on-site”) to 3 (“this is more important when working remotely”).² As such, the midpoint of this scale (i.e., a score of 0) indicates that participants think the focal leadership behaviour is equally important when working on-site versus remotely, a positive (above 0) score indicates that participants think the focal behaviour is more important when working remotely versus on-site, and a negative (below 0) score indicates that participants think the focal behaviour is more important when working on-site versus remotely.

Endorsement of identity leadership behaviours

We measured participants’ endorsement of identity leadership behaviours with the 15-item Identity Leadership Inventory (ILI; α = .89) developed by Steffens et al. (2014). The ILI offers a complete measure of the four-dimensional model for identity leadership on identity prototypicality (e.g., “Being a model member of the team”), advancement (e.g., “Promoting the interests of members of the team”), entrepreneurship (e.g., “Developing an understanding of what it means to be a member of the team”), and impresarioship (e.g., “Devising activities that bring the team together”).

Endorsement of other leadership behaviours

For exploratory purposes, we also measured participants’ endorsement of other prototypical leadership behaviours in remote versus on-site settings, which has been found that supervisors adopted a more directive, task-oriented leadership style when working remotely caused of the COVID-19 (e.g., Garretsen et al., 2022; Stoker et al., 2022). First, we selected 10 items from the leadership behaviour description questionnaire – Form XII (LBDQ XII), developed by Ohio State University (Stogdill, 1963) – to gauge various task-oriented behaviours (α = .77). A sample item is “Keeping the work moving at a rapid pace.” Second, we used the Leading-Managing Index (LMI; Kniffin & Van Vugt, 2023; Kniffin et al., 2020) to capture the importance of behaviours that are either more aligned with being a leader or a manager, with six items each (α = .73 for leading, α = .53 for managing). A sample item for leading behaviour is “Inspiring team members” and a sample item for managing behaviour is “Controlling team members.”

Attention checks

We utilized the following attention check questions: “I am totally immune to all infectious diseases” and “I am required to skip this item without any response”. We filtered out respondents who answered without reading the survey questions carefully by excluding those who answered “strongly agree” to be totally immune to all infectious diseases and did not skip questions when asked to do so.

Analytic strategy

We conducted one sample t-tests to compare whether the endorsement of each focal leadership behaviour was more or less important when working on-site versus working remotely. This was determined by assessing if the average score across participants was higher or lower than zero, the midpoint of the rating scale (meaning that the leadership behaviour is equally important when working on-site versus remotely).

Results

Table 1 shows the means, standard deviations, and bivariate correlations among the variables included in Study 1. Of all the demographic variables, only age was negatively associated with endorsement of identity leadership (r = −.18, p = .02), suggesting that especially younger employees favoured identity leadership in a remote (versus an on-site) work setting. To test Hypothesis 1, we conducted a one-sample t-test analysis (see Table 2). In line with Hypothesis 1, employees valued identity leadership behaviours more in a remote work setting compared to an on-site work setting (M = .55, SD = .85, t(180) = 8.75, p < .001, d = .65, 95% CI [.43, .67]). Results further showed that employees valued task-oriented behaviours (M = .47, SD = .59, t(180) = 10.60, p < .001, d = .79, 95% CI [.38, .56]) as well as leading behaviours more (M = .45, SD = .79, t(180) = 7.73, p < .001, d = .57, 95% CI [.34, .57]) in a remote, compared to an on-site, work setting. We observed no differences in preferences for managing behaviours between remote and on-site work settings (M = −.02, SD =.56, t(180) = −.40, p = .69, d = −.03, 95% CI [−.10, .07]). In addition, we further compared identity leadership with the other leadership behaviours through a set of paired sample t-tests with Bonferroni corrections (see Table 3). Results indicated that employees valued identity leadership behaviours more than managing behaviours in remote versus on-site work settings (t(180) = 9.48, p < .001, d = .71, 95% CI [45, .68]). However, there was no significant difference in preferences for identity leadership compared to task-oriented (t(180) = 1.29, p = .20, d = .10, 95% CI [−.04, .21]) or leading behaviours (t(180) = 1.98, p = .05, d = .15, 95% CI [00, .19]).

Table 1. Means, standard deviations, and correlations among variables in Study 1.

Variable	Mean	SD	1	2	3	4	5	6	7	8	9
1. Gender	.50	.50	—
2. Age	37.38	8.07	−.02	—
3. Tenure (in years)	7.33	6.81	−.14	.56**	—
4. Tenure with supervisor (in years)	3.43	3.25	−.04	.25**	.43**	—
5. Remote work experience (in years)	2.74	1.73	−.07	.06	.07	.03	—
6. Endorsement of identity leadership behaviours	.56	.85	.04	−.18*	−.10	−.08	.02	(.89)
7. Endorsement of task-oriented behaviours	.45	.60	−.13	−.15	−.13	−.04	.06	.33**	(.77)
8. Endorsement of leading behaviours	.45	.79	−.07	−.16*	−.12	−.08	.02	.67**	.47**	(.73)
9. Endorsement of managing behaviours	−.02	.56	−0.05	−.20**	−.09	−.04	.10	.41**	.23**	.50**	(.53)

N = 181. Gender: 0 = male, 1 = female. The endorsement of leadership behaviours was rated on a 7-point Likert scale ranging from −3 (“this is more important when working on-site”) to 3 (“this is more important when working remotely”), a score of 0 indicates that participants think the behaviour is equally important when working on-site versus remotely, a positive (above 0) score indicates that participants think the behaviour is more important when working remotely versus on-site, and a negative (below 0) score indicates that participants think the focal behaviour is more important when working on-site versus remotely. Cronbach’s values are reported on the diagonal in parentheses where applicable.

*p < .05. **p < .01.

Table 2. Results for hypotheses testing in study 1.

	Mean	SD	t	df	p	Cohen’s d	95% CI
1. Endorsement of identity leadership behaviours	.55	.85	8.75	180	<.001	.65	[.49, .81]
2. Endorsement of task-oriented behaviours	.47	.59	10.60	180	<.001	.79	[.62, .95]
3. Endorsement of leading behaviours	.45	.79	7.73	180	<.001	.57	[.42, .73]
4. Endorsement of managing behaviours	−.02	.56	−.40	180	.69	−.03	[−.18, .12]

N = 181. A one-sample t-test was conducted.

Table 3. Results for multiple comparison tests among leadership behaviour endorsements in study 1.

	Mean	SD	t	df	p	Cohen’s d	95% CI
1. Endorsement of identity leadership behaviours	.55	.85	1.29	180	.20	.10	[−.04, .21]
2. Endorsement of task-oriented behaviours	.47	.59
1. Endorsement of identity leadership behaviours	.55	.85	1.98	180	.05	.15	[.00, .19]
3. Endorsement of leading behaviours	.45	.79
1. Endorsement of identity leadership behaviours	.55	.85	9.48	180	<.001	.71	[.45, .68]
4. Endorsement of managing behaviours	−.02	.56

N = 181. Paired-sample t-test were conducted. p_adjusted = .05/3 = .008.

Our first study provided support for the importance of identity leadership in a remote working context by showing that employees endorse identity leadership more in a remote versus an on-site work setting. The aim of our second study is to build on these findings as we examine whether identity leadership behaviours can mitigate the negative consequences of remote working using a field survey.

Study 2

Method

Participants and procedure

We recruited participants from the United Kingdom through the online platform Prolific Academic (https://www.prolific.co) in August 2022. Participants could take part in our study if they worked full-time and had a supervisor within their team. An a priori Monte Carlo power analysis indicated that we required at least 186 participants to detect a large correlation (r = .50), with 95% power and an alpha level of .05. We oversampled with a 40% addition, resulting in targeting a sample of 260 participants. We excluded 58 invalid responses for several reasons: 13 participants for providing incomplete responses, 22 for failing to pass one or both attention checks, and 23 for giving an inconsistent response to their remote working time (i.e., ratios of self-reported time spent working remotely compared to total working time were above 1). The final sample therefore included 202 participants (response ratio = 78%). Of the final participants, 53% were male, their average age was 38.45 (SD = 9.32), their average organizational tenure was 7.99 years (SD = 6.74), and the average tenure with their current supervisor was 3.94 years (SD = 4.04). Participants had an average of 1.49 years (SD = 1.76) experience of working remotely and worked remotely, on average, 52% of their working time. Participants worked in a variety of occupations and worked in a wide range of industries (e.g., 22% from education, 19% from healthcare, 10% from technology, and 8% from financial services).

Measures

Unless specified otherwise, all items were rated on a 7-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree).

Remote working

Following the operationalization of remote working in prior research (Barsness et al., 2005; Wiesenfeld et al., 1999), we assessed remote working using the proportion of total time employees spent working remotely during a typical work week. Unlike previous research adopting self-reported proportions directly offered by participants, we estimated the proportions based on employees’ answers to the questions about how many hours they worked in total and how many hours they had spent working remotely. This allowed us to assess the proportion of remote working in a more objective and accurate way. Hence, the scale of remote working ranges from 0 (i.e., no remote working) to 1 (i.e., completely working remotely).

Team connectedness

We measured participants’ team connectedness with two measures: team identification and team isolation. We measured team identification with a six-item scale (α = .91) developed by Doosje et al. (1995) together with Mael and Ashforth (1992), which was adapted and validated by Milesi (2022). A sample item is “This team’s successes are my successes.” We measured team isolation with a four-item isolation scale (α = .91) developed by Connaughton and Daly (2004). A sample item is “I often feel disconnected from fellow team members.”

Identity leadership

We assessed participants’ perceptions of identity leadership with the same scale as in Study 1 (α = .97).

Work experiences

We captured participants’ work experiences with work engagement, job satisfaction, stress, and absenteeism. We assessed their work engagement with the nine-item Utrecht Work Engagement Scale (UWES-9; α = .93) developed by Schaufeli et al. (2006). A sample item is “I feel bursting with energy at my work.” We assessed job satisfaction with the three-item scale (α = .92) from Cammann et al. (1983). A sample item is “I am satisfied with my job.” Furthermore, we measured work-related stress with the four-item scale (α = .89) from Dubinsky et al. (1994). A sample item is “I feel a great deal of stress because of my job.” Last, we assessed absenteeism with a single item by asking participants “how many workdays per week, on average, have you missed for any reason other than an approved vacation?”

Control variables

We considered participants’ age (in years), gender (0 = male and 1 = female), and tenure with their supervisor (in years) as potential control variables because previous research has shown that they are associated with our outcome variables (e.g., Bedeian et al., 1992; Christian et al., 2011).

Attention checks

We utilized the following attention check questions: “I have never used digital devices such as smartphones or computers” and “Please click the answer option strongly disagree”. We excluded participants who stated they had never used a digital device and/or who did not follow the instruction to click the option “strongly disagree”.

Analytic strategy

We conducted structural equation modelling (SEM) using Mplus 8 (Muthén & Muthén, 1998-2017) to examine our hypothesized model. Following the two-step analytic strategy proposed by Anderson and Gerbing (1988), we first confirmed the measurement model using confirmatory factor analysis (CFA), and then performed SEM based on the measurement model to estimate the fit of the hypothesized model to the data. Furthermore, we employed 1000 bootstrapped samples to calculate 95% bias-corrected confidence intervals (CIs) to estimate robustness of the results.

Results

Descriptive statistics

Means, standard deviations, and bivariate correlations of the main variables of Study 2 are presented in Table 4. Results showed that remote working was negatively related to team identification (r = −.14, p = .02) but not significantly related to team isolation (r = .11, p = .11). Furthermore, remote working was negatively associated with work engagement (r = −.24, p < .001) and job satisfaction (r = −.15, p = .03), while it was neither significantly associated with stress (r = −.08, p = .25) nor absenteeism (r = .03, p = .69). In addition, team identification was positively related to work engagement (r = .66, p < .001) and job satisfaction (r = .63, p < .001), but not significantly related to work stress (r = −.11, p = .12) or absenteeism (r = −.10, p = .15). However, as expected, team isolation was negatively associated with work engagement (r = −.48, p < .001) and job satisfaction (r = −.52, p < .001) and positively associated with work stress (r =.15, p = .03) and absenteeism (r =.25, p < .001).

Table 4. Overall means, standard deviations, and correlations among variables in Study 2.

Variable	Mean	SD	1	2	3	4	5	6	7	8	9	10	11	12	13
1. Gender	.49	.53	—
2. Age	38.45	9.32	−.08	—
3. Tenure (in years)	7.99	6.74	−.05	.57**	—
4. Tenure with the supervisor (in years)	3.94	4.04	−.03	.32**	.54**	—
5. Remote work experience (in years)	1.49	1.76	−.09	.19**	.17*	.07	—
6. Remote working	.52	.40	.09	.06	−.05	−.14*	.50**	—
7. Identity leadership	5.09	1.30	.05	−.10	−.06	.05	.12	.16*	(.97)
8. Team identification	5.43	1.13	.12	.04	.11	.11	−.03	−.14*	.52**	(.91)
9. Team isolation	2.59	1.43	.01	−.10	−.09	−.10	.12	.11	−.41**	−.66**	(.91)
10. Work engagement	4.82	1.20	.08	.10	−.01	.01	−.05	−.24**	.52**	.66**	−.48**	(.93)
11. Job satisfaction	5.16	1.56	.09	.06	−.05	.04	−.05	−.15*	.54**	.63**	−.52**	.80**	(.92)
12. Work stress	4.24	1.53	.12	−.03	.09	.05	−.05	−.08	−.25**	−.11	.15*	−.20**	−.29**	(.89)
13. Absenteeism	.49	1.19	.05	.09	.06	.08	.12	.03	−.11	−.10	.25**	−.09	−.22*	.12	—

N = 202. Gender: 0 = male, 1 = female. Cronbach’s values are reported on the diagonal in parentheses where applicable.

^**p < .01. ^*p < .05..

Factor structure

Prior to hypotheses testing, we conducted a set of CFAs to ensure that our key constructs (i.e., identity leadership, team connectedness, and work experiences) had satisfactory discriminant validity. We employed single indicator and item parcelling strategies for these constructs (e.g., Hayduk & Littvay, 2012; Little et al., 2002), which allowed for achieving an optimal ratio of sample size to the number of estimated indicators in CFAs. This method is suitable for our study because we primarily focused on the distinctiveness of the measured constructs in the measurement model (Little et al., 2002, 2013). Following the recommended parcelling and single indicator techniques from previous literature, we used the dimensional scores to generate four parcels for identity leadership based on its validated multi-dimensional structure (Little et al., 2002; Williams et al., 2009). We also employed mean scores to generate single-item indicators for team connectedness and work experiences, allowing us to identify our complex measurement model and improve the theoretical precision of our model (Hayduk & Littvay, 2012). Specifically, team connectedness was measured by mean scores of team identification and team isolation, while work experiences were measured by mean scores of work engagement, job satisfaction, work stress, and absenteeism. Following suggestions by Garver and Mentzer (1999) as well as Williams and O’Boyle (2008), we reversely scored the items of team isolation, work stress, and absenteeism to ensure that loadings of the items on their corresponding latent variables are in the same direction as other items measuring the same variable. The results indicated that the hypothesized three-factor model (i.e., identity leadership, team connectedness, and work experiences) fitted the data well, χ²(32) = 67.54, p < .01; RMSEA = .07, SRMR = .04, CFI = .97, TLI = .96. Moreover, against this baseline model, the best alternative model (with team connectedness and work experiences loading onto one factor) had a poorer fit (χ²(34) = 107.85, p < .01; RMSEA = .10, SRMR = .05, CFI = .94, TLI = .93) (Δ χ² [2] = 40.31, p < . 01). Accordingly, the focal variables in the three-factor model displayed adequate discriminant validity among variables.

Hypotheses testing

We tested the full structural model with combined moderation and mediation hypotheses using SEM (see Figure 2). Following the suggestions of the latent moderated structural equation modelling approach by Maslowsky et al. (2015) as well as Sardeshmukh and Vandenberg (2017), we estimated the model fit of the full model by employing the information criteria using the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) to compare it with its baseline model where the interaction term is not included. The results of the structural model suggested that the baseline model fitted the data well, χ²(41) = 97.08, p < .01; RMSEA = .08, SRMR = .07, CFI = .96, TLI = .95, AIC = 5728.38, BIC = 5840.84. Furthermore, results of the full model with the interaction term included showed that the AIC was 5724.93 and BIC was 5840.72. Accordingly, the difference in the AIC (ΔAIC = AIC_F – AIC_B = 5724.93–5728.38) indicated that the full model with a smaller AIC was considerably better fitted over the baseline model.

Figure 2. Structural equation modelling with moderation results.

The items of team isolation, work stress, and absenteeism were reverse-scored in the structural equation modelling analyses. Parameter estimates are from the unstandardized solution.

*p < .05, **p < .01, ***p < .001.

Hypothesis 2 stated a negative relationship between remote working and team connectedness. As expected, results revealed that remote working was negatively associated with team connectedness (estimate = −.61, SE = .17, 95% CI [−.93, −.29], p < .001). In line with Hypothesis 3, that identity leadership moderated the relationship between remote working and team connectedness, results showed that the interaction of remote working and identity leadership was significant and positively related to team connectedness (estimate = .25, SE = .12, 95% CI [.01, .48], p = .04). Further simple slope tests indicated that, in line with expectations, remote working was negatively associated with team connectedness when identity leadership was lower (−1 SD; estimate = −.85, SE = .23, 95% CI [−1.32, −.42], p < .001), but not when identity leadership was higher (+1 SD; estimate = −.36, SE = .19, 95% CI [−.75, −.03], p = .05). This interaction is plotted in Figure 3. Last, in support of Hypothesis 4, that identity leadership moderated the indirect effect of remote working on employees’ work experiences via team connectedness, results showed that remote working had a significant negative indirect effect on work experiences when identity leadership was lower (−1 SD; estimate = −.78, SE = .22, 95% CI [−1.24, −.36], p < .001), but not when identity leadership was higher (+1 SD; estimate = −.33, SE = .18, 95% CI [−.72, −.02], p = .07). Hence, Hypothesis 4 was supported. We further note that including the potential control variables did not meaningfully change any of the reported findings, and results showed strong evidence that the significant negative indirect effect of remote working on employees’ work experiences was weaker when identity leadership was higher instead of lower (See Table S1 in the Supplemental Material for additional analyses including the control variables).

Figure 3. Remote working interacting with identity leadership to predict team connectedness.

General discussion

The present research set out to examine the importance of identity leadership within a remote working context. Our studies showed that employees valued and preferred identity leadership behaviours more in a remote work setting compared to an on-site work setting. Moreover, we found that identity leadership mitigated the detrimental consequences of remote working on a number of important work experiences. In particular, our results showed that remote working was negatively associated with employees’ team connectedness and subsequent work experiences, but only when they rated their supervisors as engaging less in identity leadership behaviours.

Theoretical implications

Our research makes important contributions to the literature. First, our research contributes to research on implicit leadership theories (Lord et al., 1984, 2020; Offermann et al., 1994) by showing that endorsements of particular ideal leadership traits and behaviours from supervisors are contingent on the working context. Moving beyond existing evidence of context dependence of implicit leadership theories in workplaces (e.g., Joshi et al., 2009; Offermann & Coats, 2018), our findings show that in remote work settings, the mental model of an ideal supervisor switches to someone who engages in identity leadership activities.

Second, our research contributes to the vastly growing remote work literature by establishing that when employees work more remotely and have fewer face-to-face interactions with their co-workers, they are at a risk of losing connectedness, particularly in experiencing isolation and disidentification, to their co-workers and teams. In doing so, our research shows the costs of remote working arrangements at a time when such arrangements are getting increasingly common and are even turning into the “new normal” in workplaces around the world (Franken et al., 2021; Kniffin et al., 2021; Nyberg et al., 2021). In contrast to earlier literature that looked at remote work as a special working arrangement between an employee and employer that was accompanied by privilege and the benefits of flexibility and autonomy (e.g., Allen et al., 2015; Gajendran & Harrison, 2007), our research highlights that working remotely nowadays also comes with burdens. In doing so, our research adds to a growing body of work that portrays working remotely as a more nuanced phenomenon that has specific downsides (e.g., Golden et al., 2008; Hill et al., 2022; Thatcher & Zhu, 2006).

Third, our research contributes to identity leadership research by extending its significance to remote working contexts. Our findings suggest that supervisors who engage in identity leadership activities (such as acting as a champion for the team, and making team members feel that their team matters) can foster a sense of shared identity among remotely working teams, and promote a mentality of “we are all in this together” in physically distal, digital work contexts. Thus, our research explicitly links identity leadership to remote working in a post-pandemic workplace (Ashforth, 2020), and aligns with recent calls to promote identity leadership as an effective management strategy in the workplace (e.g., Leonardelli, 2022; Van Bavel et al., 2020). Importantly, our findings demonstrate the crucial functions of identity leadership for how remotely working employees experience their work, which goes beyond the existing narrow understanding of relationship-oriented leadership in navigating remote working, such as inspirational leadership (Joshi et al., 2009). Through a variety of different strategies aimed at promoting a strong team identity (i.e., identity prototypicality, advancement, entrepreneurship, impresarioship), identity leadership can form a buffer against some of the negative aspects of remote working arrangements, such as a lower work engagement and satisfaction, as well as, increased work stress and absenteeism from work.

Practical implications

Our research also offers important practical contributions to the effective management of remote working. A growing number of organizations worldwide are shifting to remote or hybrid work settings. Despite obvious advantages for both employees and organizations, our research highlights crucial challenges associated with this major transition.

More importantly, our findings also provide specific practical tools that supervisors can use to minimize these damaging consequences of remote work. In fact, our findings suggest that leaders should represent, advance, create, and embed a shared sense of social identity among team members. According to the recently developed and tested online 5 R leadership development programme (Haslam et al., 2023), supervisors can, for instance, enhance their prototypicality by individually mapping out their own identities, together with their teams mapping out identities that matter to them, and finally, with their team, develop a shared identity. These activities can be done in in-person meetings, but also in online sessions (Haslam et al., 2023).

Furthermore, leaders can advance a sense of “we-ness” by promoting the team’s shared interest to the rest of the organization. Supervisors can, for instance, advocate for the interests of the team concerning budget and personnel decisions or when implementing new policies regarding working-from-home. This is especially important in increasingly remote workplaces in which the team supervisors often serve as the main liaison between the team and senior management of the organization (Haslam et al., 2020). Moreover, especially in such remote working conditions, leaders should actively communicate and provide evidence of their identity-enhancing attempts to their team so that these do not go unnoticed.

Leaders can further craft a sense of togetherness by routinely acting like identity entrepreneurs in the sense that their words and behaviours foster a sense of a shared identity (S. Reicher & Hopkins, 2001). Supervisors can do this by explicitly defining and communicating the core values, norms, and ideals of their team, especially by promoting open disclosure and actively listening to the insights on these cores of the team shared by team members, which would be especially important to foster the togetherness in an impoverished online environment. Moreover, supervisors should reinforce the notion of collective identity inside the team, for instance, by talking about “we” and “us” instead of “I” and “me” when communicating with their team members (Leonardelli, 2022).

Finally, leaders can embed a shared social identity by organizing online and/or in-person events in which team members can live out their group membership. For instance, as exemplified by the study conducted by Haslam et al. (2023) on online interventions, supervisors can involve team members in collective goal setting to strengthen the collective membership in the geometrically separated team (Wegge & Haslam, 2003). Furthermore, supervisors can set up events, such as remote office parties and online afternoon teas, as a means of fostering connectivity and cohesion within virtual teams and enhancing collaborations among their members (Leonardelli, 2022). Although these activities should not overburden the workload of remote employees, they have the potential to greatly improve team connectedness when timed appropriately. In addition, supervisors are encouraged to initiate the team-supporting structure in remote work settings (Haslam et al., 2023), for example, by regularly checking in on team members and by listening to their potential struggles at work.

Limitations and future research directions

The findings of our research should be interpreted with limitations in mind. First, we acknowledge that the cross-sectional and correlational nature of our data may cause common method bias and prevents us from establishing causal claims regarding our hypothesized relationships (Podsakoff et al., 2003). At the same time, we note that our focal variables of feelings of connectedness and work experiences reflect employees’ personal and internal experiences at work, and therefore, self-report measures seem like an appropriate way to get to these (Chan, 2010; Spector et al., 2019). Furthermore, our focal variables have already been well established in the literature, and our studies show satisfactory construct validity, as evidenced by their high levels of reliability (α > .90) and sufficient discriminant validity. We further note that common method bias is unlikely to account for the interactive effects between remote working and identity leadership, as such interaction terms are said to be less influenced by common method bias (Podsakoff et al., 2012).

Following recommendations from Conway and Lance (2010), Podsakoff et al. (2003), and Spector (2019), we further proactively took several steps to reduce potential common method bias. First, in Study 1, we randomized the order of the items within the scales, as well as, the scales within the survey. Moreover, we counterbalanced the response anchors, such that participants were randomly assigned to a condition that either included answer options ranging from “this is more important when working on-site” to “this is more important when working remotely” or vice versa. In Study 2, we further employed different scale formats for the predictor (i.e., remote working) and outcome variables (e.g., work engagement, absenteeism). Second, our measure of remote working was more objective as it was calculated from self-report measures regarding the number of total hours employees work and the number of hours they work remotely. Third, we assured participants of the anonymity and confidentiality of their responses and emphasized that there were no negative consequences for not responding or quitting the studies. Lastly, the Harman’s single-factor test revealed that the total variance extracted by one factor was 46%, which is below the threshold of 50%, suggesting that our data was limitedly affected by common method bias.

Nevertheless, future research should try to corroborate our findings through applying longitudinal or (quasi-) experimental designs, to establish a causal link between remote working and loss of team connectedness. It would be ideal to conduct an intervention study with a pre-/post-test measurement design, capturing leaders’ engagement in identity-enhancing leadership activities (e.g., organizing online tea/coffee mornings, working together in real-time via Google Docs). In addition, we encourage researchers to use more objective data to gauge remote working frequencies and patterns, such as through reports from supervisors and HR managers on how often people work from home or on-site, or even digital data obtained through so-called wearable devices to find out how much people are actually spending on work activities remotely versus on-site and how much face-to-face contact they have with co-workers.

Second, the measure we adopted to assess the preference for leadership behaviours in the first study, known as comparative scaling, was cognitively demanding, as it required to evaluate the leadership behaviours in contrasting work settings. We therefore encourage future research to assess the endorsement of identity leadership across work settings separately and independently, for example, by asking participants’ individual endorsement of identity leadership for working on-site and for working remotely, and then comparing their endorsements.

Furthermore, we compared, for exploratory purposes, the preference for identity leadership behaviours with various other prototypical leadership behaviours (such as task-oriented, leading, and managing behaviours). We did, however, not compare identity leadership with the full range of leadership behaviours, including more transformational, servant, or empowering behaviours (e.g., Kniffin et al., 2021). In addition, for these other leadership behaviours included, the findings further demonstrated that remote workers valued, not identity leadership only, task-oriented and leading behaviours more in a remote versus an on-site work setting. We did not follow up on the findings to further investigate the role and the impact of those leadership behaviours in this study. Accordingly, future research can focus more on additional prototypical leadership behaviours and discover their possibly significant roles in remote or hybrid workplaces, as employees may expect their leaders to engage in specific behaviours more in remote than on-site work settings.

Third, the employee samples in both studies were recruited from the United Kingdom. As a result, our findings may not be easily generalizable to other countries and work organizations (Maner, 2016; Pitesa & Gelfand, 2022), although the trend towards remote or hybrid working is observed across many nations. Therefore, it would be a significant extension to replicate our findings in other countries in Europe and elsewhere in the world to see if our conclusions can be generalized.

Fourth, our work merely concerns the employee-centric perspective on identity leadership behaviours in remote work settings. While the theoretical concept of identity leadership primarily captures the follower viewpoint, it would be an interesting and important extension of our research to examine whether identity leadership would be more or less difficult to execute in remote work settings. Existing research has increasingly realized that constructing one’s and others’ identity within an organization is more costly in a remote compared to a typical on-site workplace (e.g., Bartel et al., 2007), but we know little about the possible challenges faced by leaders especially in crafting and fostering a shared identity within remote working teams.

Finally, the present research primarily examined the mitigating role of identity leadership for team connectedness. Building on our findings, it would be interesting to find out if identity leadership could also play a role in strengthening the ties between remote employees and their organization as a whole. Recent studies showed that during COVID-19, employees felt less committed to their organizations (e.g., Mihalache & Mihalache, 2022), and the recent trend of quiet quitting is generally seen as a manifestation of a weaker organizational identification among employees (Harter, 2022). Future research could examine whether identity-enhancing activities also create a stronger sense of connectedness with the whole work organization, thereby reducing employees’ turnover intentions.

Conclusion

Overall, the present research demonstrated the importance of identity leadership in remote working contexts. First, our results showed that employees value identity leadership activities from their supervisors more in a remote work setting than in an on-site work setting. Second, our findings indicated that working remotely can come at the cost of losing identities and connections. Most importantly, we showed that identity leadership can mitigate these negative consequences of remote working by protecting feelings of connectedness among remotely working employees, thereby reducing a range of negative work experiences. Overall, our research thus demonstrates the importance of identity leadership in remote work settings and for making the transition towards increased remote or hybrid working practices a success.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request. The data are not publicly available due to privacy restrictions for participants.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/1359432X.2024.2362460

Notes

1. We checked the characteristics of the included sample and found that most participants worked both remotely and on-site over the last year (67.4%), while only 59 participants worked completely remote (32.6%). To increase the confidence of our reported findings, we further tested Hypothesis 1 excluding these 59 completely remote working employees. Results of this analysis showed that the exclusion of these complete remote workers did not result in any meaningful changes regarding reported preferences for identity leadership in a remote versus an on-site work context, M = .51, SD = .89, t(121) = 6.33, p < .001, d = .57, 95% CI [.35, .67].

2. To attenuate potential common method bias regarding these measures of leadership behaviour endorsements, two technical steps were implemented. First, we counterbalanced items within and outside their sets by randomizing the order of items sets and the order of items within each set. Second, we employed two opposite anchoring scales in a randomly assigned approach for participants. Specifically, half of the participants were assigned to a scale that ranged from “−3 = more important when working on-site” to “3 = more important when working remotely”, while the other half were assigned to a scale that ranged from “−3 = more important when working remotely” to “3 = more important when working on-site”. For the statistical analysis, we used the first option and thus reversely scored the responses from the second group of participants.