Abstract
Managers are increasingly required to lead individuals they rarely, if ever, see, but most leadership research still assumes frequent face-to-face interaction. To explore these relationships, the authors used a web-based survey (n = 402) and structural equation modeling techniques to determine the fit of a proposed model of leadership in both remote and proximal environments. The authors used four previously identified contextual elements: perceptions of control, prior knowledge of leader, unplanned communication, and regularly scheduled communication with leader. The findings suggest that in the remote environment, perceptions of transformational leadership mediate the relationship between these elements and individual outcomes. The same model does not fit the proximal environment, suggesting that being managed proximally differs significantly from being managed remotely and that this difference resides in the context.
Information technology is increasingly used by organizations to enable employees to work at a distance, yet there has been minimal investigation into the nature of working in an environment in which the leader and follower are separated by physical distance and the majority of one’s interaction with a manager is conducted through technology (Martins, Gilson, & Maynard, 2004; Zigurs, 2003). According to Kayworth and Leidner (2002), little research exists that examines the effect of this context, identifies, and explains the factors that contribute to its increased complexity, or explores the specific relationships and processes through which it influences individual outcomes. The few extant studies have focused primarily on the performance and communication patterns of virtual teams with little attention to the individual experience and/or the contextual elements of this environment (Powell, Piccoli, & Ives, 2004). Leadership in the remote environment has been particularly underexplored (Collinson, 2005), even though research on virtual teams has demonstrated that it is vitally important (Zimmermann, Wit, & Gill, 2008).
Extensive research has shown that leadership is an important predictor of both individual and team outcomes (Barling, Weber, & Kelloway, 1996; Dvir, Kass, & Shamir, 2004), and much of this research assumes that individuals interact with their leaders and group members on a “face-to-face” basis (Kelloway, Barling, Kelley, Comtois, & Gatien, 2003). As organizations become increasingly reliant on electronic communication, the extent to which this model adequately explains leader behaviors and their outcomes is questionable: Leaders cannot simply lead virtually exactly the same as they do in face-to-face situations (Hambley, O’Neill, & Kline, 2007a; Zaccaro, Ardison, & Orvis, 2004). Accordingly, we investigated the relationships between leadership factors/behaviors and perceptions of leadership and the relationship of these to individual outcomes in an environment characterized by physical distance and reduced face-to-face interaction between employees and their leaders. We refer to this situation as “remote leadership.” In particular, we considered the possibility that leader behaviors are related to individual outcomes in a relationship mediated by perceptions of transformational leadership and that these relationships are differentially related to the context in which they occur—that is, proximal versus remote.
The Remote Context
It is impossible to understand behavior in organizations without an explicit consideration of the organizational context, and yet context has been overlooked in many areas of organizational behavior (Pawar & Eastman, 1997), specifically, the remote environment, which differs from the proximal one in many ways. For the most part, these differences have been neither inventoried nor explored. Obviously, the remote environment is characterized by physical distance and often, reduced face-to-face interaction; specifically, the individuals and their leaders work in different locations (Zimmermann et al., 2008). In our mixed method study of remotely led workers, we identified, tested, and refined several factors that may act as antecedents and consequences of perceptions of transformational leadership in a remote leader–member relationship. Using these antecedents (perceived control, unplanned communication, regularly scheduled communication, and prior knowledge of leader), we developed a preliminary model of remote leadership (Figure 1), which we then tested on both remotely and proximally managed individuals, with the objective of determining whether the data fit the model and whether that fit varied significantly between the two groups.
Preliminary model of remote leadership
Theoretical Background and Hypotheses
Perceived Control
Perceived control has previously been directly associated with individual and organizational outcomes, such as performance, job satisfaction, organizational commitment, stress, and organizational climate in face-to-face environments (Aiello & Douthitt, 2001; Schat & Kelloway, 2000; Yagil, 2002). It is considered especially important in service organizations where employees frequently operate on their own and are difficult to monitor (Yagil, 2002). Using this logic, perceived control may be particularly important in the remote context, where similar conditions exist. Purvanova and Bono (2009) observed that virtual communication creates a sense of uncertainty and ambiguity. Moreover, feelings of isolation may result from being physically separate from one’s leader and/or work group. Lack of understanding of organizational drivers, which often is gained through casual face-to-face interaction, may lead members to feel particularly powerless. Control, as an element of empowerment, has been significantly associated with transformational leadership in numerous studies in the proximal environment (Kark, Shamir, & Chen, 2003; Ozaralli, 2003). Several authors have suggested that transformational leadership is a reciprocal process in which both leader and follower are changed by one another; in effect, that follower characteristics have the potential to predict leadership (Dvir, Eden, Avolio, & Shamir, 2002). Therefore, we proposed the following:
Hypothesis 1a: Perceived control predicts perceptions of transformational leadership style in the remote leader–member relationship.
Hypothesis 1b: Perceived control predicts job satisfaction in the remote leader–member relationship.
Hypothesis 1c: Perceived control predicts organizational commitment in the remote leader–member relationship.
Hypothesis 1d: Perceived control predicts perceptions of manager trust in the employee in the remote leader–member relationship.
Communication Quality and Quantity
Quality
Researchers such as Purvanova and Bono (2009) suggest that electronic communication is inferior to face-to-face communication. Nonverbal cues, which may convey as much as two thirds of message content, are reduced or lost, potentially resulting in misinterpretation, increased role ambiguity, decreased trust, cue substitution, inaccurate perceptions of self and others, lowered leader influence, and reduced group cohesiveness (Antonakis & Atwater 2002; Avolio, Kahai, Dumdum, & Sivasubramaniam, 2001; Hambley, O’Neill, & Kline, 2007b; Hart & McLeod, 2002; Jarvenpaa & Leidner, 1998; Kayworth & Leidner, 2002). This may explain why virtual teams take longer to make decisions, and team members are less able to make inferences about other members’ knowledge or anticipate their responses (Cramton, 2001; Martins et al., 2004). Electronic communications are perceived as less “warm.” E-mail messages in particular contain higher levels of negativity than face-to-face communications (Berry, 2006; Kurtzberg, Naquin, & Belkin, 2005); consequently, the degree to which interpersonal relations may develop is reduced (Cramton, 2001; Kayworth & Leidner, 2002; Weisband & Atwater 1999). Furthermore, context cues may also be lost; head nods and eye contact facilitate and regulate interaction and indicate whether a message requires further explanation, thereby reducing ambiguity and error (Straus & McGrath, 1994). Social context cues, which contribute to perceptions of leader legitimacy, expertise, and status may also be reduced, affecting perceived power and leader influence and reducing the salience of leader behaviors (Hambley et al., 2007b; Sosik, Avolio, & Kahai, 1997).
Quantity
In virtual teams, frequent communication results in improved working relationships among team members, increased job satisfaction, and increased perceptions of trust in leaders (Hart & McLeod, 2002; Staples, 2001). Specifically, more frequent and higher quality communication has been found to strengthen the relationship between affective commitment and transformational leadership. Yet the remote environment carries significant risk of decreased communication frequency (Straus, 1997), partially because it is not conducive to chance encounters (Gluesing, 2000). Such serendipitous communication provides an opportunity for informal interaction, the importance of which has long been accepted (Kotter, 1972). Virtual communication may further constrain social interaction by the physical effort involved in typing what may be seen as nonessential words (Straus, 1996, 1997). Simulating the frequent regular interactions that comprise the proximal environment can be challenging for remote leaders. The importance of regularly scheduled communication is accepted in the practitioner literature on managing (Anshel, 1992). However, there has been little prior empirical investigation of its importance. Jarvenpaa and Leidner (1998) suggested that predictability of communication improves virtual team effectiveness. In an interview study, participants emphasized that effective virtual leaders need to encourage regular communication and to conduct regular one-on-one meetings (Hambley et al., 2007a).
Accordingly, we proposed the following:
Hypothesis 2a: Regularly scheduled communication predicts perceptions of transformational leadership style in the remote leader–member relationship.
The serendipitous encounters of a proximal relationship are important to relationship building and information transfer and are problematic in a remote context. It has been suggested that virtual team leaders should try to simulate an environment in which members are collocated with the leader (Hambley et al., 2007a). There has been little investigation of this specific issue, but related research demonstrates the importance of certain transformational leadership behaviors, such as individualized consideration, in both the traditional and the remote environments (Avolio, 1999; Bass, 1998; Kelloway et al., 2003). Unplanned, informal interaction may be a vehicle through which these consideration behaviors are enacted, affecting perceptions of leadership (Judge, Piccolo, & Ilies, 2004). Therefore, we proposed the following:
Hypothesis 2b: Unplanned communication predicts perceptions of transformational leadership style in the remote leader–member relationship.
Prior Knowledge
Researchers have long recognized the significance of trust to leadership and have identified a host of antecedents to its development (Dirks & Ferrin, 2002; Kramer & Tyler, 1996). However, there may be different antecedents in the case of a leader who does not frequently interact on a face-to-face basis with members (Jarvenpaa, Knoll, & Leidner, 1998; Kayworth & Leidner, 2002). Studies examining the effect of prior relationships among members of virtual and proximal teams suggest that prior knowledge significantly affects trust formation, subsequent performance, and member satisfaction (Alge, Wiethoff, & Klein, 2003). Members with strong personal relationships have been found to share a common understanding and require less clarification in their communication (Hart & McLeod, 2002). Even one inaugural face-to-face meeting of virtual teams, a form of prior knowledge, is associated with greater team effectiveness, facilitating members’ use of leaner media to communicate effectively and allowing them to build a shared interpretive context (Alge et al., 2003; Maznevski & Chudoba, 2000; Suchan & Hayzak, 2001). How reputation as a form of prior knowledge affects trust development in a remote environment has not been explored, but McKnight et al. (1998) found that in face-to-face relationships, good reputations help create cognitive-based trust, which Staples (2001) suggests is important in the remote environment. Therefore, we chose to include prior knowledge of leader, personally or by reputation, as a contextual predictor in our preliminary model. We proposed the following hypothesis:
Hypothesis 3: Prior knowledge predicts perceptions of transformational leadership style in the remote leader–member relationship.
Individual Outcomes
Job satisfaction
There is little research on job satisfaction in the remote environment. In general, lower levels of job satisfaction have been reported in virtual teams (Martins et al., 2004; Straus & McGrath, 1994; Warkentin, Sayeed, & Hightower, 1997). Satisfaction with technology, use of multiple communication methods, and nature of task have all been positively related to the job satisfaction of virtual teams’ members (Kayworth & Leidner, 2002; Martins et al., 2004; Morris, Marshall, & Rainer, 2002). Both cognition-based and affect-based trust were found to significantly affect job satisfaction of remote workers (Staples, 2001). Studies of telecommuters have yielded conflicting findings (Igbaria & Guimaraes, 1999), with Golden and Veiga (2005) suggesting a curvilinear link between extent of telecommuting and job satisfaction.
Organizational commitment
Few studies of organizational commitment have been situated in the remote environment. The importance of subculture and satisfaction with control to organizational commitment in the proximal environment suggests that context and other aspects of the culture of the remote workplace may be more influential than the culture of the larger organization (Lok & Crawford, 1999), and that it is important to determine which factors affect the remote worker’s satisfaction with level of control. Lok and Crawford (1999) found that of the two leadership style variables, consideration and structure, the former was more strongly related to organizational commitment. In a similar vein, Boshoff and Mels (1995) found that leadership styles incorporating participation in decision making and goal setting increased individuals’ organizational commitment. However, a meta-analysis by Mathieu and Zajac (1990) demonstrated that, in the traditional environment, the relationships between various leader behaviors and organizational commitment are contingent on other factors in the work environment. In the proximal environment, a strong, positive relationship between job satisfaction and organizational commitment has been consistently found (Lok & Crawford, 1999; Mathieu & Zajac, 1990). However, findings related to the direct impact of organizational commitment on individual performance have been conflicting, leading researchers to question whether mediational processes exist (Mathieu & Zajac, 1990). Research in the proximal environment suggests that organizational commitment may also be an important predictor and consequence in the remote environment; however, the related factors, the strength of the relationships, and the processes of influence have yet to be determined.
Manager trust in employee
In all leader–member relationships, reciprocal trust is important. Typically, follower trust in leader is included in perceptions of transformational leadership, both by construct definition and in this study, statistically. However, leader trust in remote follower has not been previously investigated, other than minimally in the literature on telecommuting (Harrington & Ruppel, 1999). In remote relationships, both the reality and the perception are important. In the proposed model, however, it is the perception of leader trust in follower that is posited as an outcome of perceptions of transformational leadership. As Cramton (2001) demonstrated, electronic communication can lead to significant errors in attribution; silence, for example, is often attributed to dispositional effects, when it may be the result of innocuous causes, such as wrongly addressed e-mails. When the follower believes that he/she is trusted by the leader, these attributions differ. In the remote leader–member relationship, there is an increased opportunity for third parties to influence a leader’s perception of and trust in distant members (Atkinson & Butcher, 2003; Burt & Knez, 1996; Cramton, 2001). Burt and Knez’s (1996) work suggests that there is a predilection among organizational members for gossip and negative attributions, and that the leader–member relationship is inherently political. It is important, then, that members believe they are trusted by their leader and that decisions affecting them are not biased by such political behavior. Prior studies have already demonstrated the pragmatic effect of reduced face-to-face interaction in terms of decreased ratings for distant employees (Judge & Ferris, 1993). In order for a remote environment to be perceived as supportive, members must feel that their leaders trust them enough to give them the benefit of the doubt when unable to ask about something in person. Because this as yet unexplored trust process clearly functions differently in the remote environment, we chose to include it as an outcome of leader behavior in our model of remote leadership
Transformational Leadership Style
Transformational leadership as conceptualized by Bass (1985) enables followers to transcend their own self-interests for a higher collective vision and thereby, exceed performance expectations. The behaviors cluster into four factors: idealized influence, inspirational motivation, intellectual stimulation, and individualized consideration (Howell & Hall-Merenda, 1999). It has been found to have differential effects on followers’ performance, both directly and indirectly (Boerner, Eisenbeiss, & Griesser, 2007; Judge & Piccolo, 2004). Despite a plethora of research on transformational leadership, the question of how organizational context influences its emergence and effectiveness in the remote environment is still a relatively unexplored area (Purvanova & Bono, 2009). In a study of sales managers in an environment characterized by physical distance between leaders and members, Dubinsky, Yammarino, Jolson, and Spangler (1995) found no incremental effect of transformational leadership in follower performance. They suggested that the remote context in their study may have affected the findings. In a subsequent study of proximal sales managers, Humphreys (2002) did find a significant relationship between transformational and follower performance, and called for comparative investigations of the two environments. Kelloway et al. (2003) demonstrated that individuals could perceive differences in leadership styles in electronic communications and other researchers have suggested that distance actually strengthens the effects of transformational leadership behaviors on both individual and group performance (Antonakis, Avolio, & Sivasubramaniam, 2003). Hambley et al. (2007b) concluded that as a communication medium becomes more anonymous, transformational leadership may be the more effective leadership style relative to transactional leadership. However, other studies suggest that leader proximity is required for the effective use of transformational leadership behaviors (Howell & Hall-Merenda, 1999; Yagil, 1998).
Specific transformational leader behaviors, such as coaching and frequent communication, have been associated with perceptions of higher ability and increased benevolence and trust among members of virtual teams (Avolio et al., 2001). Preliminary research suggests that remote leaders can employ individualized consideration to influence the moods and emotions of the group, by showing concern for each member’s needs (Jarvenpaa & Leidner, 1998). Also, by encouraging teams to consider and recognize the value of each member’s input, remote leaders can focus the team on collective goals (Avolio et al., 2001).
The remote transformational leader can engage in intellectual stimulation by increasing the frequency, level, and nature of information exchange (Kelloway et al., 2003). Previous studies support the importance of frequent task-related messages to perceptions of leader ability and influence (Hart & McLeod, 2002; Jarvenpaa et al., 1998; Jarvenpaa & Leidner, 1998). This behavior may also reveal information about members’ ability, benevolence, and integrity, contributing to the development of unconditional trust (Jarvenpaa & Leidner, 1998). Therefore, we proposed the following:
Hypothesis 4a: Transformational leadership style predicts job satisfaction in the remote leader–member relationship.
Hypothesis 4b: Transformational leadership style predicts organizational commitment in the remote leader–member relationship.
Hypothesis 4c: Transformational leadership style predicts perceptions of manager trust in the remote follower.
Method
The primary goal of this study was to test the boundary conditions of the proposed model of remote leadership by determining its fit in two different environments. A second goal was to determine whether the remote and proximal groups differ significantly on the context characteristics that have been suggested by previous research.
Participants and Procedures
Participants for this study were reached by advertising to members of targeted associations, e-mail lists, and by word of mouth. There were 402 valid responses to a web-based survey. The sample consisted mostly of individuals in either professional or management positions (33.8% and 49%, respectively), with the majority (67%) between the ages of 30 and 50 years. Most respondents (64%) had been in their current job more than 2 years. Male respondents comprised 48.8% and female respondents comprised 51.3% of the sample. The majority of respondents (56.9%) reported that they see their managers face-to-face at least two to three times per week, with a further 6.8% seeing their managers once per week. A large percentage (36.4%) spent face-time with their managers two to three times per month or less. More than 60% of respondents were located in the same building as their managers, with a further 10% in a different building but in the same city, whereas approximately 29% were located in a different city or province/state from their manager. Although it is certainly possible for remote workers to have significant levels of contact with their managers, more typically, physical distance results in lower levels of face-to-face contact. Indeed, Humphreys (2002) argued that physical distance is often accompanied by varying degrees of other types of distance (e.g., social and psychological), resulting in what Napier and Ferris (1993) refer to as “dyadic distance.” Therefore, we considered those respondents located in a different city or province/state from their managers to be remotely managed for the purposes of this study. The survey included two sets of scales: one set specifically assessed characteristics of the remote working context and the second set comprised scales developed for general organizational use. Items were rated on a seven-point Likert-type scale ranging from 1 (strongly disagree) to 7 (strongly agree).
Measures
We assessed the remote context using four unpublished scales that we had previously developed through interviews with remote workers, then refined and subsequently validated in a survey of remote workers (n = 701). These scales are included in the appendix.
Regularly scheduled communication
If the respondent and his/her manager established regular times for communication, we asked respondents to complete this four-item scale, which measured the importance of these interactions to the respondent (Cronbach’s α = .87).
Unplanned communication
We used this four-item scale to measure the extent to which the respondent’s manager initiated communication on an unplanned basis (Cronbach’s α = .76).
Prior knowledge
We used this six-item scale to measure the respondent’s familiarity with his/her manager and/or group before the current job assignment (Cronbach’s α = .84).
Manager trust
We used this three-item scale to measure the respondent’s perception of his/her manager’s trust in him/her (Cronbach’s α = .97).
The general organizational scales measured the following variables:
Control
We measured respondent perception of their control over their surroundings by the three-item Powerlessness Scale (Ashford, Lee, & Bobko, 1989; Cronbach’s α = .88).
Organizational commitment
We measured affective organizational commitment by the six-item version of the Organizational Commitment scale (Meyer, Allen, & Smith, 1993; Cronbach’s α = .86.)
Job satisfaction
We measured job satisfaction by the five-item General Satisfaction Scale from the Job Diagnostic Survey (Hackman & Oldham, 1975; Cronbach’s α = .82).
Transformational leadership
We measured leadership style by the seven-item Global Transformational Leadership scale (Carless & de Paola, 2000; Cronbach’s α = .95). Items were rated on a 5-point scale, ranging from 1 (rarely or never) to 5 (very frequently, if not always). In previous research, this measure was found to assess a single global construct of transformational leadership and to have acceptable discriminant and convergent validity.
Results
Descriptive statistics and correlations for all study variables are displayed in Table 1. Prior to testing the models, we formed two subsamples comprising proximally and remotely led employees. If a respondent and manager were located in the same building, the respondent was assigned to the proximally managed group (N = 241). All other respondents (i.e., all those who worked in a building separate from their managers), were assigned to the remotely managed group (N = 151). To assess group differences on the study variables, we conducted a multivariate analysis of covariance using age and gender as covariates, remote versus proximal management as the independent variable, and all 10 study variables as the dependent variables. To assess group differences, we used the multivariate test based on Wilks’s lambda, with subsequent univariate tests evaluated using a Bonferroni-adjusted alpha level. With the use of the Wilks’s criterion, the combined dependent variables were significantly related to group membership, F(10, 375) = 2.55, p < .01 and to each of the covariates; age, F(10, 375) = 3.31, p < .01 and gender, F(10, 375) = 2.61, p < .01. One variable, unplanned communication, F(1, 386) = 9.27, was significant at the adjusted level of .005, when age and gender were controlled. Examination of the estimated marginal means revealed that for the remotely managed group, the mean score for unplanned communication was 4.59, whereas for the proximally managed group, the mean score was 5.05, indicating that the respondents in the proximally managed group reported more frequent unplanned communication than did the remotely managed group.
Descriptive Statistics, Internal Consistency Values, and Intercorrelations for Variables Used
Note: Listwise N for correlations = 395; α = Cronbach’s index of internal consistency.
p < .05. **p < .01.
To assess the proposed structural model, we conducted a series of nested model comparisons on both groups, proximally and remotely managed. The starting point for the path analysis was a fully mediated model, using data from the remotely managed group, with the four contextual variables acting as exogenous variables that predict transformational leadership that in turn predicts job satisfaction, organizational commitment, and manager trust.
The partially mediated model, with all four contextual variables, provided a better fit to the data than did the fully mediated model,
Partially mediated model of remote leadership (remote sample)
All hypotheses were supported, using the remote sample. Job satisfaction (β = .29, p < .01; Hypothesis 4a), organizational commitment (β = .36, p < .01; Hypothesis 4b), and manager trust (β = .29, p < .01; Hypothesis 4c) were all predicted by transformational leadership. Job satisfaction (β = .36, p < .01; Hypothesis 1b), organizational commitment (β = .39, p < .01Hypothesis 1c), and manager trust (β = .20, p < .01; Hypothesis 1d) were also predicted by control. Transformational leadership in turn was predicted by control (β = .22, p < .01; Hypothesis 1a), regularly scheduled communication (β = .14, p < .01; Hypothesis 2a), unplanned communication (β = .38, p < .01; Hypothesis 2b), and prior knowledge (β = .14, p < .01; Hypothesis 3). For the remotely managed group, the model explained 17% of the variance in manager trust, 30% of the variance in job satisfaction, 39% of the variance in organizational commitment, and 35% of the variance in perceptions of transformational leadership.
The partially mediated model did not adequately fit the data for the proximally managed group, χ2(17) = 35.07, p < .01; GFI = .97; AGFI = .91; RMSEA = .07; NFI = .92; CFI = .96; PNFI = .35 (Table 2). As hypothesized, job satisfaction (β = .31, p < .01), organizational commitment (β = .36, p <.01), and manager trust (β = .17, p < .01) were all predicted by transformational leadership. Job satisfaction (β = .26, p < .01) and organizational commitment (β = .27, p < .01) were also predicted by control. In this sample, however, Hypothesis 1d was not supported; control did not significantly predict manager trust.
Fit Indices for Models
Note: GFI = goodness-of-fit index; AGFI = adjusted goodness-of-fit index; RMSEA = root mean square error of approximation; CFI = comparative fit index; NFI = normed fit index; PNFI = parsimony normed fit index.
Transformational leadership in turn was predicted by control (β = .26, p < .01) and unplanned communication (β = .26, p < .01). Regularly scheduled communication and prior knowledge did not significantly predict transformational leadership in this sample. For the proximal group, the model explained 12% of the variance in manager trust, 31% of the variance in job satisfaction, 29% of the variance in organizational commitment, and 16% of the variance in perceptions of transformational leadership (see Figure 3). The standardized parameter estimates and squared multiple correlations are generally lower than those in the remote sample, suggesting that these contextual characteristics are not as important in the proximal environment as in the remote one.
Partially mediated model of remote leadership (proximal sample)
Discussion
This study offers three major insights. First, it provides a preliminary model of the relationships between the salient characteristics of work context, perceptions of leadership, and individual outcomes in the remote environment. Second, it suggests that leadership style acts as a mediator between work context and individual outcomes for those managed from a distance; in other words, context predicts perceptions of leadership. Third, the study suggests that the experience of being managed proximally differs qualitatively from the experience of being managed remotely and that this difference resides in the context.
The finding of significant difference between the two samples contributes to understanding the experience of being managed remotely. There is a significant difference between the two groups, even when demographic factors are controlled, and this difference is more than a matter of degree in leader behaviors. The model is a very good fit to the data from the remote sample and includes four context variables as significant predictors of perceptions of transformational leadership, which in turn predicts job satisfaction, organizational commitment, and perceptions of manager trust. This same model does not adequately fit the data from the nonremote sample and in fact, includes nonsignificant paths from two of the context variables, regularly scheduled communication and prior knowledge, to transformational leadership, suggesting that in remote jobs, context appears to matter more than in proximal jobs.
These results also suggest that remote management relationships appear to be characterized by a lower level of unplanned communication. This in itself is not surprising. The lack of opportunity for serendipitous communication at a distance has previously been acknowledged (Gluesing, 2000). However, this variable also includes frequency of usage of e-mail and telephone—proximal managers use e-mail and telephone more than distant ones. Because increased face-to-face communication is usually a feature of proximal management, this suggests that more face-to-face communication encourages higher communication frequency using other channels as well. Although one might anticipate a higher level of usage of alternate media in a remote relationship, this does not appear to be the case. The relationship between proximity and increased attraction may explain the higher communication frequency in proximal relationships (Latane, Liu, Nowak, Bonavento, & Zheng, 1995; Moon, 1999). Unfortunately, this may have a negative impact on the outcomes in a remote relationship. Numerous studies in this environment have demonstrated the significant links between increased communication frequency and positive individual and group outcomes (Jarvenpaa & Leidner, 1998; Kayworth & Leidner, 2002; Staples, 2001). So, while existing research suggests that higher use of e-mail and telephone communication would be beneficial in the remote environment, these findings indicate that higher usage is occurring in situations in which face-to-face communication is also higher. This is consistent with Humphreys’ (2002) contention that physical distance gives rise to other manifestations of distance in the leader–member relationship.
Furthermore, the model suggests that unplanned communication is a significant predictor of perceptions of transformational leadership in both remote and proximal leader–member relationships. As noted, higher levels of unplanned communication reflect increased communication frequency. As well, unplanned communication may create a perception within the individual of being important to the leader. Consideration behaviors have long been a part of leadership theories and a recent meta-analysis suggests that they are strongly related to member satisfaction with leader and job, as well as leader effectiveness (Judge et al., 2004). In Humphreys’s (2002) study of transformational leadership in a proximal sales environment, individualized consideration accounted for 89% of the unique variance in follower performance. Frequent unplanned communication is consistent with the individualized consideration element of transformational leadership (Bass, 1985).
There has been little prior empirical investigation of the importance of regularly scheduled communication. Jarvenpaa and Leidner (1998) suggested that predictability of communication improves virtual team effectiveness and, as previously discussed, other researchers have noted the importance of communication frequency. These findings, however, suggest that in the remote environment, regularly scheduled communication between the leader and the member significantly influences the member’s perceptions of transformational leadership, which in turn has been extensively associated with positive outcomes (Kelloway et al., 2003). The path is not significant in the proximal environment. It may be that regularly scheduled communication achieves an effect beyond the increase in communication frequency; it may create a perception of being valued by the leader. This would partially explain the difference of this path in the two models; conveying a sense of consideration may be more easily achieved when leader and member are collocated. Furthermore, demonstrating reliability by adhering to a communication schedule may result in increased in trust in the leader. Trust and individualized consideration have previously been demonstrated to be important predictors and/or components of transformational leadership (Bass, 1985; Dirks & Ferrin, 2002).
Prior knowledge or “history” between group members may increase trust or mutual knowledge, both of which have been associated with increased communication effectiveness (Cramton, 2001). In an experimental investigation of proximal charismatic leadership, Puffer (1990) found that knowledge of the leader’s performance and decision style resulted in higher levels of attributed leader charisma. These findings suggest that, in the remote environment, prior knowledge of the leader and/or the group may contribute to prediction of leadership perceptions, but not so in the proximal model. It may be that prior knowledge of the leader, his or her reputation, or group provides a prior basis of trust, which has been significantly associated with perceptions of leadership (Dirks & Ferrin, 2002; Hart & McLeod, 2002; Mennecke & Valacich, 1998). It may also be that remotely managed workers feel more vulnerable than do their proximally managed counterparts, and therefore a preexisting basis for trust is significant.
Extensive investigation of perceived control in the proximal environment suggests that control beliefs positively and directly affect individual outcomes, such as job satisfaction (Furnham, 1992) and increased trust in manager (Brashear, Manolis, & Brooks, 2005). Analysis of our proximal sample partially supports these findings; however, the direct effect of control on individual outcomes was smaller than in the remote sample, and in the proximal sample, the path between control and the perception of manager trust was nonsignificant. This may reflect the potential for remote workers to feel excluded and out of touch with organizational politics; control beliefs may be translated in this environment into a perception of being trusted by their managers, a connection potentially not as relevant in the proximal settings.
Although control is typically conceptualized as being predicted by transformational leadership, it appears plausible that in the remote relationship, the causal flow may be reversed; the perception of control may invoke the trust component, enhancing overall leadership perception, with its resulting positive individual outcomes. In fact, previous related studies in the remote environment focused on the effect of control beliefs in a virtual team setting have linked perceptions of individual control with trust among team members (Piccoli & Ives, 2003). The indirect effect of control on individual outcomes through perceptions of transformational leadership may be related to trust in the same way. Furthermore, it may be that participants whose self-schema (Payne, Colletti, & Connor, 1987) incorporates feelings of control selectively receive, process, and interpret more clearly those leader behaviors that align with their beliefs; in other words, the behaviors that appear to be more empowering. Since an empowering style is central to transformational leadership (Avolio, 1999), followers would thus perceive their leaders to be transformational. However, causation is clearly a topic for further research.
The model of remote leadership developed and tested here suggests that transformational leadership is predicted by the four context variables discussed above and, in turn, predicts the individual outcomes, job satisfaction, organizational commitment, and perceived manager trust. The model explains 35% of the variance in perceptions of transformational leadership for the remote sample, and only 16% of the variance for the proximal sample. As discussed above, perceptions of transformational leadership are substantially influenced by specific elements in the remote context, whereas this is not the case in the proximal environment. However, the role of transformational leadership as a predictor of outcomes is quite similar in both samples. As noted earlier, positive relationships between transformational leader behaviors and follower outcomes have been extensively reported (Barling et al., 1996; Kirkpatrick & Locke, 1996). It is not surprising, therefore, that transformational leadership significantly predicts job satisfaction and organizational commitment in both samples. It appears that, although the perception of transformational leadership is strongly influenced by context in the remote environment and not to the same degree in the proximal environment, once a follower perceives a leader to be transformational, the outcomes are similar, regardless of how that perception was established. This is consistent with recent findings by Purvanova and Bono (2009) that suggest that the perception of transformational leadership is a prerequisite for job satisfaction in both proximal and remote teams.
In both samples, perceptions of transformational leadership significantly predict perceptions of manager trust. The remote model explains 17% of the variance in manager trust, while the proximal model explains 12%. The specific process by which perceptions of transformational leadership predict a higher perception of manager trust is as yet unclear, but it may be related to an increase in self-efficacy as a result of transformational leader behaviors. One of the goals of transformational leadership is to convey leader confidence in followers’ abilities, and this is enacted through behaviors such as open communication, intellectual stimulation, and participative decision making (Bass, 1985). A follower receiving such a message may feel an increased sense of self-efficacy, associated with a perception of being trusted by the manager. Although there are contradictory findings about the nature of the relationship between self-efficacy and leader effectiveness, there is evidence of a significant association between them (Felfe & Schyns, 2002; van Knippenberg, van Knippenberg, de Cremer, & Hogg, 2004). These findings suggest that this association is important in both environments, although to a greater extent in the remote leader–member relationship.
Conclusion and Implications
The main goal of this research was to advance our knowledge of the remote leader–member relationship; specifically, to identify if the contextual characteristics of that relationship influence the impact of leadership on individual outcomes. This study provides strong evidence for the argument that the remote environment requires a new model of leadership, different from those based on the premise of face-to-face interaction.
In a practical vein, the findings of this research can provide direction to individuals charged with managing individuals remotely. The “good” news is that it is possible to establish perceptions of transformational leadership from a distance and that once established, positive outcomes result. The “bad” news is that establishing this perception takes work. The remote leader–follower relationship should be characterized by individuals feeling free to initiate casual interaction through various media. To create perceptions of transformational leadership, frequent interaction of two types—unplanned and regularly scheduled—is required. Managers should telephone, e-mail, or visit remote followers on an ad hoc basis, and not merely in response to a problem situation or issue; that is, plan for “unplanned” communication. Managers should schedule regular times to communicate, through various media, and keep those commitments. The predictability and reliability of these interactions is what matters. Managers should ensure they do not have more frequent communication only with those with whom it is easy to communicate. Proximal followers naturally have more opportunities for interaction; managers must manufacture similar opportunities for remote followers. Managers can do little to create prior knowledge where none exists; however, they can still create that mutual basis of knowledge through initial face-to-face contact. Control beliefs are particularly important in the remote environment. Significant levels of decision and process control have been associated with increases in followers’ control beliefs, suggesting an empowering management style is important in remote contexts. Creating an atmosphere of trust in both directions also contributes to the formation of control beliefs. The varying components of trust are invoked by differing leader behaviors that convince followers of one’s reliability, ability, and benevolence.
In conclusion, these findings suggest that, in the remote environment, context matters by influencing perceptions of transformational leadership. It may be that in the remote environment, context is so omnipresent that it filters the way in which individuals perceive and interpret leader behaviors. The “remote” relationship environment is, after all, defined by its context—and that context is fundamentally different from the environment in which the majority of leader–member, indeed most human relationships have been conducted since the dawn of history. The context of the proximal relationships is equated with “normal” and may therefore not be differentially influential in individuals’ perceptions and judgments. This possibly explains the lack of consideration of relationship context as the predictor of leadership style in previous research. This study suggests that these models are not applicable in an environment in which relationship context is not “normal.” To manage perceptions of leadership style in the remote environment, it is not sufficient to exhibit specific transformational behaviors; leaders must manage, consider, and adapt to the characteristics of the context in which the relationships are conducted. We live in a world where electronic communication is rapidly replacing face-to-face contact, and we have little understanding of how this affects human interaction, especially in the workplace. This research suggests that it may require a different way of managing.
Footnotes
Appendix
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.