From a theoretical point of view, the outcomes of this study have several implications. During the analysis of literature, it became evident that, to date, no common definition of distance leadership existed as definitions often overlapped. Recognizing distinct dimensions of distance remains challenging and requires further evaluation. Future research could implicitly follow up on the definition provided in this paper. Empirical contributions still constitute the exception and thus, validated research instruments assessing objectively for distance are scarce. Therefore, academic literature should focus on the application of empirical studies in future investigations. In contrast to prior research, this work treated distance not solely as a contextual factor. Next to examining physical distance it also took relational elements from leader-member exchange theory into account. Future investigations might consider and elaborate on emotional, affective and/or cognitive aspects of distance.
Research should further emerge in the area of examining self-leadership and its influences on leadership behavior and follower work-related outcomes. This study provides evidence that self-leadership and leadership behavior are related. Selfleadership has furthermore been shown to impact the relationship between leaders and followers. Expanding on the present work, researchers may want to examine different work outcomes related to leadership and self-leadership behavior. Employee satisfaction has been often linked to transformational leadership (Judge & Piccolo, 2004), now it would be useful to test whether self-leadership would produce similar results. Andressen and colleagues (2012) recently observed that, in a virtual setting, self-leadership has a higher impact on employee motivation than in a co-located setting. Even if Full Range Leadership constitutes the most profound theory in current leadership research, there are more leadership facets that find only limited attention in organizational science thus far. Research on the impact of distance on effects of shared or empowering leadership lack empirical foundation.
In addition, the present work utilizes and tests a new scale for the evaluation of self-leadership. The SLSI has been applied to a larger organizational sample for the first time since its development by Furtner and Rauthmann (in prep.). High reliability values between .82 and .93 attest good internal consistency. Factor loadings varying between .79 and .95 suggest further good quality of the constructs, except for the first-order factor of self-reminding which shows considerably lower values. As similar results were retrieved in a principal analysis done by Furtner and Rauthmann (in prep.) it is recommended to exclude the first-order dimension in future investigations.
At one step, the influence of leaders’ self-leadership on employee satisfaction could be tested, and in a second step one could test for the impact of employees’ selfleadership behavior on satisfaction. Other work-related outcomes to test for might be related to creativity or innovativeness. DiLiello and Houghton (2006) hypothesize that strong self-leaders have more potential for creativity, creative problemsolving, and innovation than weak self-leaders. Carmeli et al. (2006) even claim self-leadership to play the critical role of enhancing the innovation process and the exhibition of innovative behavior. Hauschildt and Konradt (2012b) emphasize that self-leadership has thus far been applied sparsely in organizational contexts. Particularly, evaluating the effect of self-leadership on performance has potential. This could be assessed using a different measure, preferably one where follower performance is rated by a different source (e.g., leaders).
Graen and Uhl-Bien (1995) criticize approaches in leadership research that have often disregarded the levels of dyadic relationship between individuals. Research has formerly acknowledged one part of the model exclusively; either leaders’ or the followers’ perspective. The authors argue that leadership largely depends on three domains which are the follower, the leader, and their relationship (Graen & Uhl- Bien, 1995, p. 221). As such, they contemplate the leader-follower relationship as a multi-domain construct that may provide useful insights and practical implications for organizational research. The researchers postulate that research should ideally focus on all of these facets. A recent investigation by Erdogan and Bauer (2014) revealed that 83% of all studies assess LMX from followers’ perspective. The studies further showed low levels of convergence between leaders’ and followers’ ratings of LMX which might be caused by the reluctance of leaders to admit that they do not have a good relationship with their team members (Hiller, DeChurch, Mu- rase & Doty, 2011; Sin, Nahrgang & Morgeson, 2009).
This research contributes to the organizational literature in the field of leader- member exchange theory. In early work, Graen and Uhl-Bien (1995) suggest to further test LMX and its influence on organizational outcome variables. Particularly, LMX is recommended to be applied as a moderator. Discovering that physical distance did not moderate the relationship between LMX and performance in a study by Howell and Hall-Merenda (1999, p. 690) the finding highlights the great potential of LMX quality where leadership from a distance can still be executed effectively as long as a qualitative relationship is established.
Although followers’ tenure with the leader did not reveal any interference, a longitudinal investigation of dyadic relationships would be feasible to highlight the stage and the evolution of relationship quality over time. Graen and Uhl-Bien (1995) suggested that LMX quality traverses different stages during its development. In distance leadership research, the stages might differ as they probably depend on the timing of the first physical contact. If leader-follower contact is rare at the beginning, research could then answer if and how a working relationship with high quality LMX could be built up.
Despite the fact that the importance of AIT rises in organizational contexts, the results of this study demonstrate that communication frequency is not as important as previously outlined (Cummings, 2008, p. 46). This could be due to the measurement of frequency instead of looking more deeply into the power of different media channels that are applied in firms. Future studies may emphasize differences in channel usage such as communication using lean or rich media. This may help to identify the appropriate channels for each state of the leader-follower or project life-cycle, proposing rich media for the introductory phase and less rich media for a relationship that is established. Also little is known about the interaction between leadership styles and AIT usage (Avolio et al., 2014, p. 126).
Research projects in which leaders and followers are asked to provide ratings are complex. For this study, only followers’ cases were evaluated and calculations done accordingly. In that sense, making assumptions on the variance of responses is not feasible as dyadic relationships were not considered. Multi-source data should be gathered in future investigations, resulting in a multi-level analysis.