Moderation and Mediation Effects of Distance on the Leader-Follower Relationship
Physical distance negatively moderates the influence of leadership behavior on follower self-leadership strategies.
Distance leadership has recently attracted the attention of researchers as early investigations provided evidence of effects of distance on the leader-follower relationship. Leadership at distance changed work in organizations and traditional leadership behaviors are becoming inadequate (Bradner & Mark, 2008; Hertel, Geister & Konradt, 2005). In early stages, physical distance was expected to neutralize leadership behaviors (Kerr & Jermier, 1978, p. 396). Antonakis and Atwater (2002, p. 685) postulated that physical distance may be linked to negative work- related outcomes. Various research attempts discovered repeatedly that physical distance impacted the leader-follower relationship (e.g., Howell & Hall-Merenda, 1999; Howell et al., 2005).
Correlations of physical distance with other variables are discussed prior to the analysis of the hypothesis. An indication of negative relation between active leadership behavior and physical distance is visible in the correlation matrix. Transformational leadership showed significant negative interrelation with physical distance (r = -.17, p < .001). The more leader and follower were geographically separated, the lower were perceptions of transformational leadership. This finding could be an indicator of how difficult it is to transmit transformational aspects while not being physically present. Physically distant leaders could practice other leadership behaviors instead, based on what they know about the challenges of being inspiring, motivational, and visionary while using technology media. No correlation could be detected for transactional leadership and physical distance. For passive leadership, the correlation showed significant positive effects with physical distance (r = .15, p < .01). Followers who were led from a large physical distance perceived passive leadership characteristics more distinctly. The reason for this could be that physical distance by nature brings elements of laissez-faire and suggests a “free rein.” Physical distance was further found to correlate weak but significantly positive with followers’ gender. Hence the majority of followers led at a distance were male. Interpretation of this is vague as no evidence of causality can be drawn. As seen in the descriptive statistics, nearly 80% of followers that work in large international corporations in the technology industry were males.
The subject of the hypothesis was to determine statistically whether physical distance moderated the influence of leadership behavior on follower self-leadership strategies. Calculations display that physical distance indeed has moderating effects for the influence of transformational leadership on follower self-leadership (P = .09, t = 1.67, p < .05). Looking at the correlations at various physical distances, transformational leadership was only correlated to self-leadership at either no distance (0 km; very close) (P = .19, t = 2.87, p < .01) or at very large distance (> 1,000 km; very distant) (P = .49, t = 5.15, p < .001). For all other distance categories, transformational leadership did not predict self-leadership. For followers who were very distant from leaders, transformational leadership even predicted follower selfleadership more strongly than when they were very close. An explanation for transformational leadership better leading to self-leadership at no distance could be that transformational leaders can execute their role and act as role-model right in front of their followers. This behavior can be found particularly in the idealized influence (behavior) subfacet. Supporting this explanation, idealized influence (behavior) reports the strongest correlation with self-leadership (r = .29, p < .001). This is valid for both, the very close (r = .25, p < .001) and the very distant (r = .53, p < .001) group. An explanation why this could be true for the close group was provided, but why does idealized influence (behavior) predict self-leadership in followers, even when they are far away most of the time? Recalling the interaction patterns in very distant leader-follower relationships, face-to-face encounters are relatively rare. When meeting only sporadically, leaders are likely to exhibit their best behavior, striving to seem determined, dynamic, and equipped with extraordinary capabilities. This picture would then remain in the minds of the followers.
In general, the findings indicate that only leadership at no distance or at very large distance allow for the development of self-leadership. Those followers - neither very close nor very distant - are located in an interval state, where leaders are unable to permanently role model yet they cannot grant full autonomy either. For the other two dimensions of Full Range Leadership, effects were differentiated. Whereas no moderation effect of physical distance could be detected for the influence of transactional leadership on follower self-leadership, moderation occurred for passive leadership. A direct effect of passive leadership on follower selfleadership could not be verified, yet the significance of the interaction term bears potential for contemplation. Splitting the file into distance categories, physical distance did not have an influence on the relation between passive leadership and selfleadership for any of the groups, except for the very distant one. For this group, passive leadership suddenly predicted follower self-leadership negatively (P = -.24, t = -2.26, p < .05). This finding leads to the conclusion that passive leadership in combination with large physical distance might even enhance the negative effects of this counterproductive leadership behavior on the development of followers’ self-leadership.
Physical distance negatively moderates the influence of leadership behavior on follower performance.
Correlations of physical distance with other parameters have been discussed in previous sections. The second hypothesis concerning potential moderating influences of physical distance on the leader-follower relationship was investigated next. Physical distance was expected to negatively influence effects of active leadership behaviors on followers’ performance, and to positively impact the influence of passive leadership on follower performance. Moderation tests revealed that physical distance did not moderate any of those relationships. It was assumed that the effects were too small to display, hence this outcome led to more in-depth analysis of the effects of leadership behavior at certain physical distances. The reason for this procedure is found in the outcomes of the previous hypothesis. It was presumed that the influence of transformational and passive leadership (even if very small) could vary with physical distance. Therefore, correlations were computed for the different distance groups. At very close condition, transformational leadership predicted follower performance (r = .17, p < .05). This was also true for the very distant group when transformational leadership predicted follower performance even more strongly (r = .33, p < .01). Transactional leadership did not predict performance at any physical distance. Passive leadership did project negative performance at very close (r = -.17, p < .05) and very distant levels (r = -.29, p < .01).
The outcomes are comparable with those of the prior hypothesis. Again, the very close and very distant groups seem to benefit from transformational leadership, whereas large physical distance increases the counterproductive impact of passive leadership. In fact, Kayworth and Leidner (2002) found that effective leadership in a virtual environment was mostly related to mentoring abilities of leaders, which in turn are an indication for transformational leadership behavior (Bass, 1985, 1990) and high quality relationships (Erdogan & Bauer, 2014; Law et al., 2000). Howell and Hall-Merenda (1999) revealed transformational leadership to be in particular effective under close conditions. More support for this outcome is provided by Howell et al. (2005). Yet, at that time, the researchers found that transformational leadership did not predict business unit performance under distant conditions.
Returning to the original hypothesis and the question whether moderation through physical distance occurs, it can be concluded that the effects are too small to discern meaningful effects. Summarizing these findings, the role of physical distance in the leader-follower relationship has long been exaggerated (at least considering effects on individual follower performance). Hence this implies that physical distance does not have to be a barrier for effective leadership (Neufeld et al., 2010) as other parameters in the leader-follower relationship seem far more influential (Eichenberg, 2007).
Physical distance does show negative effects on the quality of relationship.
The roles of physical distance and the quality of relationship have gained interest in recent organizational research (e.g., Eichenberg, 2007; Howell & Hall-Merenda, 1999). According to Graen and Uhl-Bien (1995) relationship quality can be expressed by the extent of leader-member exchange. The intercorrelation matrix illustrated various significant relations between LMX and other variables. Some have been already explained in previous paragraphs (e.g., self-leadership).
The underlying research found leader-member exchange to be positively related to transactional (r = .54, p < .001) and transformational leadership (r = .81, p < .001). This finding can be explained as LMX can be considered both, transactional and transformational. Meanwhile the relation with transformational leadership is remarkably stronger. This is consequently due to the development of the LMX/leadership relation as, to begin with, LMX is more or less understood as social exchange process, whereas effective LMX relationships quite often result in transformational leadership (Graen & Uhl-Bien, 1995, p. 239). The strong positive association with the active leadership styles expresses, in turn, a strong negative correlation with passive leadership (r = -.63, p < .001). Leaders that are not present when needed, who fail to coach and to provide feedback, certainly have difficulties developing functioning relationships with their subordinates.
LMX further correlated significantly positively with individual follower performance (r = .19, p < .001). This finding supports prior research that found LMX and performance to be positively associated (Carter et al., 2009; Howell & Hall- Merenda, 1999; Kacmar et al., 2003; Wang et al., 2005). This correlation indicates that the quality of exchange between leader and followers does impact how leaders rate followers’ performance, and - confirmed by the positive correlation in this study - how followers rate themselves. LMX also showed a positive link to followers’ tenure, acknowledging Graen and Uhl-Bien’s (1995) postulation that relationship-building takes time. Investigating the relationship between LMX and communication channels, the only significant correlation was detected with face-to-face contact (r = .20, p < .001). The positive relation demonstrates that a higher degree of face-to-face contact is required in order to foster high quality relationships.
The underlying hypothesis articulated a negative link between physical distance and the quality of relationship. Results revealed that physical distance predicts relationship quality moderately but significantly negatively (r = -.22, p < .001). In other words, when the physical distance between leaders and followers expands, the quality of relationship decreases or is more difficult to establish. The findings support prior outcomes by Bass (1990) who declared that distance is likely to have negative effects on the quality of exchange between leaders and followers. Later, this was confirmed by Eichenberg (2007) who found relationship quality to diminish with spatial distance. In organizations where work is carried out regardless of location, establishing high quality relationships with the leader over physical distance is challenging. As the findings show, this is even more the case, the further the two parties are geographically separated and the less they meet face-to-face. This is not the only problem, however. If relationships are established, they need to be sustained.
Relationship quality mediates the influence of leadership behavior on follower performance.
Examining the leader-follower relationship more closely, the question was posed how active leadership behavior can achieve its full potential in close and distant environments. After it was learned that physical distance and relationship quality exhibited a negative correlation, it was tested whether relationship quality had mediating effects on the influence of leadership behavior on followers’ performance. The first tested model revealed the relationship between transformational leadership and follower performance to be fully mediated by relationship quality. For transactional leadership, an indirect effect was found, whereas for passive leadership the mediation turned out to be negative.
The findings can be interpreted as showing that high quality relationships are the bond between transformational leadership and followers’ performance. Relationship quality makes this link become relevant. Comparable results were detected by Wang et al. (2005). The researchers discovered LMX to be mediating the effects of transformational leadership on follower performance (task and organizational citizenship behavior). The finding reveals that transformational leadership cultivates high quality relationships (Wang et al., 2005) and followers are able to interpret relationships (Carter et al., 2009). Team members with a high quality LMX showed further higher organizational commitment when working virtually (Golden & Veiga, 2008). For the transactional leadership/performance relationship, the indirect effect expresses that the quality of relationship is an intensifier of the association, though limited. Relationship quality is less important for transactional leadership than it is for transformational leadership. This can be explained by the fact that the effort-reward relationship of transactional behavior might not fully allow for the development of high quality LMX. Transactional leadership depends on behaviors of awarding employees for exchange and thus followers would know what to achieve for a certain performance (Pearce & Sims, 2002), regardless of a low or high quality relationship. The negative outcomes for the influence of relationship quality on effects of passive leadership on follower performance can be understood to mean that relationship quality has the potential to reduce negative effects of passive leadership. Followers exposed to passive leadership could yet perform adequately as long as the relationship between leader and follower is established and they have the appropriate competences to perform the job.
Summarizing these findings, relationship quality seems to be the tying knot between leaders and followers. Relationship quality not only allows for the influence of transformational leadership on follower performance; with a working relationship between the two, even negative effects of passive leadership can be reduced. The study thus agrees with Eichenberg (2007) who manifested that relationship quality has the strongest effects among distance dimensions on the leader-follower relationship and may act as the bond between the two, especially in a distance work setting.
Interaction frequency positively moderates the influence of transformational leadership and transactional leadership behavior on follower performance.
The present investigation examined moderation effects of leader-follower interaction frequency on the influence of leadership behavior on followers’ performance. Taking this component into consideration is essential when studying distance leadership as the interrelation between virtual communication and distance can be explained by the fact that technology constitutes a necessary aspect and a prerequisite to executing distance leadership effectively (Eichenberg, 2007, p. 43).
The computed interaction frequency index had no interrelations with other variables. This could be due to the focus on frequency of interaction, neglecting media richness of the numerous channels. (In that sense, it was disregarded whether followers communicated, e.g., five times per week face-to-face or twice per chat and three times via e-mail. The frequency of interaction would, in that case, have been five.) Although prior research has shown that communication frequency can indeed be assessed only by the number of leader-follower interactions (e.g., Kacmar et al, 2003), the present study addressed the obstacles encountered with the measure, and it was soon decided to control for interaction channels. In the next paragraphs, results of the intercorrelation matrix are illustrated and described; afterwards, the outcomes of the hypothesis test are discussed.
Outcomes of the correlation reflect a poor picture with regards to the interaction frequency index and its interrelatedness. The missing correlation with other variables raised questions as to the adequacy of the computed interaction frequency score. Therefore, interaction frequency, together with specifications of certain media channels, was taken into consideration as deployed in prior studies (e.g., Kirk- man et al., 2004; Hambley et al., 2007b). Approximately 20% of respondents in the present research had face-to-face interaction with their leaders on a monthly basis or less. Predictably, the amount of face-to-face encounters decreased significantly with increased physical distance (r = -.61, p < .001). The further apart leader and follower were, the less they met personally. Face-to-face interaction was further discovered to correlate negatively with followers’ tenure (r = -.10, p < .05). In other words, followers that had a longer work relationship with their leader met him or her less frequently in person. This might be due to the fact that when leaders and followers know each other (for a longer period of time), personal encounters are not needed as frequently as in the early stages, as both parties know what makes the other person “tick.” In addition, female followers seemed to meet more often with their leaders than did their male counterparts (r = -. 14, p < .01). However, no conclusion can be provided as to whether this was initiated by the leaders or female followers. E-mail was the second most dominant channel for leader-follower interaction. More than 90% of study participants exchanged e-mails with supervisors at least once a week. Those followers using e-mails for interaction, communicated also frequently per telephone (r = .32, p < .001). More than 70% indicated talking to their leaders on the telephone.
A statistically significant positive intercorrelation was found between the usage of videoconferences and physical distance (r = .26, p < .01). Hence, the relation to the number of face-to-face meetings was negative (r = -.15, p < .01). Videoconferencing is most likely the communication device reflecting the highest degree of a real encounter. Other researchers reported that videoconferencing helped bridge large physical distances between leaders and followers (McGrath & Hollingshead, 1994). It is viewed as a potential alternative to face-to-face meetings (Baker, 2002; Brad- ner & Mark, 2008; Duarte & Snyder, 1999). The use of videoconferencing also reported positive relations with frequency of e-mail (r = .12, p < .05) and telephone interaction (r = .19, p < .001). Videoconferencing was still the communication mode least frequently applied in organizational settings. Interaction scores of videoconferencing and chat were found to be equally cohesive in a study by Hambley et al. (2007b). Yet, tasks were fulfilled more quickly using videoconferencing. Those study participants who used videoconferencing, also made use of chat software frequently (r = .47, p < .001). It was not surprising that the frequency of chat communication was higher in distance work settings (r = .23, p < .01). Chat occupies low boundaries and is applied rather informally. Similar to the application of videoconferencing, chat was more frequently used when face-to-face interaction is low (r = - .10, p < .05). Chat was further discovered to correlate positively with telephone (r = .22, p < .001) and e-mail conversations (r = .23, p < .001).
Findings of the study show that - although distance leadership is not a rare practice - communication habits do not yet fulfill the potential they actually offer. Those leader-follower pairings where interaction is dependent on telephone, will most likely exchange e-mails frequently as well. Those using chat will most likely use videoconferencing too, and vice versa. In any text-only interaction, leaders and followers use additional audio or audio-visual media. The findings show that leadership requires a personal touch which in turn can be supported by text-only software (e.g., for documentation).
Examining the correlations between leadership behavior and the use of media channels, a direct correlation of transformational leadership and communication device could not be detected. A negative relation was found between transactional leadership and the use of chat software (r = -.15, p < .01). This was unexpected, as text-only technology was projected to be better suited in situations where standardized routines are demanded (Huang et al., 2010). Where predominantly quantitative tasks need to be fulfilled, transactional leadership was found, in prior research, to be more appropriate (Hoyt & Blascovich, 2003). A potential explanation could be that chat - opposed to other text-only media - is less used for documentation purposes. With its limited input space, it might be used more often for informal talks or quick enquiries as opposed to setting goals and defining rewards for achievements. For passive leadership, the negative correlation with face-to-face interaction was expected (r = -.15, p < .01). At any rate, passive behaviors are characterized by limited interaction with followers and the negligence of leadership (Furtner & Baldegger, 2013).
In order to test the hypothesis whether interaction frequency intervened on the relation between leadership behavior and followers’ performance, two models were calculated and analyzed. Tests revealed that interaction frequency moderated the influence of transformational leadership on follower performance; it acted as a strong enhancer (r = .62, p < .001) of this relationship. Recent research did not produce concordant outcomes. Hambley et al. (2007b) argued that the type of media employed did not impact the influence of leadership behavior on team outcomes. Later findings, however, support the assumption that leadership effectiveness is related to communication media (Kahai et al., 2012) and media richness (Huang et al., 2010). Whereas quantitative performance was encouraged by transactional leadership (in both face-to-face teams and virtual teams), qualitative performance was determined by transformational leadership (Hoyt & Blascovich, 2003). For the frequency of interaction, research confirms that high levels of interaction lead to improved team performance (Weisband, 2002). Neufeld et al. (2010) argued that the key to leadership effectiveness might lie in communication. The researchers found communication effectiveness to play a mediating role in the relationship between transformational and contingent reward leadership and leaders’ performance. Communication frequency was further discovered to moderate the relationship between LMX and job performance ratings (Kacmar et al., 2003). LMX was more strongly related to positive performance outcomes when communication frequency was intense. As leader-member exchange inherits both, transformational and transactional elements Graen and Uhl-Bien (1995) provide support for the findings of present study; yet only for transformational leadership. Moderation was not confirmed for effects of transactional leadership on follower performance.
The findings suggest that leader-follower interaction frequency bears potential, in particular for transformational leadership, to enhance its positive effects on follower performance. Remembering that transformational leadership showed its best effects in very close and very distant settings, it is hypothesized that, especially at very large physical distance, transformational leadership in combination with ex?tensive interaction could bridge the geographical gap and lead to a clear improvement of followers’ performance.
This chapter combined results of present study and compared them with findings of previous research. After a general discussion of leadership behavior, self-leadership, and relationship quality outcomes of the hypotheses were discussed and argumentation was provided to explain and justify outcomes of the present work. The first part of the discussion of the hypotheses concentrated on interpreting direct effects of leadership behaviors on follower work-related outcomes, while the second part was directed at explaining results of the moderation and mediation analyses of distance dimensions.