In the theoretical model associated with Hypothesis 4, I argue that authentic leadership facilitates followers’ perceptions of PS value congruence. This, in turn, fosters followers to speak up their ideas and concerns. I further qualify this prediction by proposing that authentic
leadership strength can influence whether this mediating process strongly hold. Specifically, I argue that the mediating effect will be stronger when leadership strength is higher and that the mediating effect will be weaker when leadership strength is lower.
When assessing the emergent properties of group or unit-level phenomena such as leadership, prior research almost focused on the average of all group members’ perceptions of their leaders’ behaviors. However, this view may lose valuable information regarding the variability of these perceptions (Felfe & Heinitz, 2010). As Hannah, Walumbwa, et al. (2011) suggested, the effects of group composition can be fully represented by examining not only the mean level of a given phenomenon, but also the strength of that phenomenon across the team. Thus, it is necessary for researchers to simultaneously consider these two group composition phenomena. According to Chan’s (1998) typology of composition models, the above two group compositions are associated with additive, direct consensus, and dispersion models. Additive models suggest that the meaning of a group or unit-level construct is an average of the individual-level perceptions regardless of the variance among these perceptions.
Direct consensus models draws on “within-group agreement of scores to index consensus at the lower level and to justify aggregation of lower level scores to represent scores at the higher level” (Chan, 1998, p. 237). Based on Chan’s additive and direct consensus models, I treat unit-level authentic leadership in the current study as shared and within-group agreement of authentic leadership among all members within a unit. In contrast, dispersion models conceptualize within-group variance as a focal construct of theoretical significance rather than merely a statistical prerequisite for aggregation (Chan, 1998). Hence, leadership strength in this study represents the extent to which members agree on the favorability of authentic leadership within the unit. In addition, Leadership researchers have increasingly underscored that leadership behaviors may interact with leadership strength to predict workplace outcomes (e.g., Cole et al., 2011; Felfe & Heinitz, 2010). Thus, leadership strength should play a
prominent role in the study of leadership.
Drawing upon a social influence perspective (Festinger, 1950), Feinberg et al. (2005) showed that followers’ attributions of transformational leadership depend on both the extent to which a leaders engages in the appropriate behaviors and the extent to which followers perceive the leader similarly. They further pointed out that leaders who display
transformational leadership behaviors and create a consensus or a similar mindset among team members will be perceived as being the most transformational. Social influence theory posits that group members refer to one another to construe their social reality, and that there are strong pressures within a group to establish and maintain consensus. Given leaders play a crucial role in shaping the perceptions of followers regarding the features of the work
environment (Piccolo & Colquitt, 2006), Bliese and Britt (2001) echoed this assumption by proposing that the emergence of consensual leadership-climate perceptions represents the quality of a unit’s shared work environment. In a similar vein, Hannah, Walumbwa, et al.
(2011) also emphasized that authenticity strength can create a strong situational influence, which offer members with clearer norms for what is expected and how to behave
appropriately. Conversely, weak situations may produce ambiguity and friction and lead to different interpretations because of unclear behavioral norms.
In line with the above perspectives, I propose that strong leadership strength should augment the likelihood that authentic leaders will transfer their appropriate values and beliefs to their followers and followers will align with their leaders’ values and beliefs
correspondingly, which in turn facilitate followers to be willing to share their true ideas and concerns. Thus, by modeling authentic values and behaviors, high levels of leadership
strength amplify followers’ value congruence with their leaders and encourage their followers to speak up based on their values and beliefs. Conversely, when a unit is with a wider polarity of member high and low on perceptions of authentic leadership, I argue that authentic leaders
will hardly transfer their correct values and behaviors to their followers through the processes of positive modeling that make voice more likely because of a lack of clearer behavioral signals and guidelines. Therefore, weak leadership strength may not allow authentic leaders to transfer their values to their followers and subsequently speaking up.
The aforementioned arguments mean that the mediated effect captured in Hypothesis 4 varies over levels of leadership strength within a unit. The positive relationship between authentic leadership and follower voice through person-supervisor value congruence is stronger when leadership strength is higher. Predictions of this pattern are referred to as a first-stage moderated-indirect effect model in which a mediated effect that differs over levels of a moderator that operates at the first stage of the mediated relationship (Edwards &
Lambert, 2007). Therefore, I expect the following hypothesis:
Hypothesis 5. The strength of the mediated relationship between authentic leadership
and follower voice (via person-supervisor value congruence) will vary depending on the extent of leadership strength; the indirect effect of authentic leadership via person-supervisor value congruence on follower voice will be stronger when leadership strength is higher.