5. Discussion and conclusion
By focusing on “relational power,” this research expands the two-dimension taxonomy of power in organizations. The study also probes into the three theoretical issues of this emerging concept, namely, definition, measurement scale, and independence. First, after reviewing the classic power models, we point out that the relationship-based power source is neither contained in the original taxonomy nor subjected to full empirical tests. We then explore the possibility of the existence of relational power and its correlated characteristics by referring to previous research. Second, we design the relational power scale (see Appendix 2) by combining qualitative and quantitative methods with the goal of providing an effective tool for follow-up study. The results of the EFA indicate that the concept of relational power comprises two subtypes, namely, direct relational power and indirect relational power. This result is consistent with the theoretical discussion of Shang et al. (2012) on the classification of relational power. Direct relational power may be based on the strength of the relationship between leader and subordinate. This kind of relationship emphasizes affection and felt obligation between two parties. The leader and his/her subordinate may socialize with each other after work; discuss personal problems and exchange advice on personal issues (Shang et al., 2012). Indirect relational power may be related to the use of social or relational capital to exert influence on the subordinate's behavior. If target object wants to share the social or relational capital of the leader, he/she needs to become a core member of the leader's network first. Finally, we verify the correlation and independence among relational, position, and personal power. The empirical test results strongly prove that relational power can be regarded as the third major power source in organizations.
