The present study has several strengths. First, I addressed a research gap in the authentic leadership literature by examining leader moral potency as an important antecedent
of authentic leadership. Second, I examined the effects of authentic leadership on an important organizational outcome (i.e., employee voice), capturing both promotive and prohibitive dimensions of voice. Third, I proposed PS value congruence as an intervening mechanism of the authentic leadership-behavioral response relationship. Fourth, I introduced two different stage (i.e., first- and second-stage) moderators in the process through which authentic leadership influences employee voice. Fifth, given the difficulty, stemming from possible social desirability bias, in assessing employee voice using self-reports, I choose to measure employee voice using supervisor reports. Sixth, I tested my theoretical model using data collected solely from the military. It can serve as a source of strength because supervisors and subordinates performed relatively similar duties and responsibilities across the units, allowing fairly equal comparisons.
Despite these strengths, several limitations of this study remain. First, although I employed a supervisor-subordinate-dyads design to address issues of common method
variance (Podsakoff et al., 2003), this approach does not completely rule out the possibility of the inflating effects of common-source bias. However, several factors can help reduce the threat of common method variance. First, I used procedural design remedies (e.g., the
assurances of complete confidentiality and anonymity, counterbalancing the measure order of questionnaires, and distinct questionnaire sections and instructions) to alleviate the
vulnerability of my data to common method variance. Second, the data of authentic leadership and leadership strength were calculated and aggregated from unit members’ perceptions.
These variables are less subject to subjective bias, and distinct from individual-level variables.
Third, the high-order interactions were not susceptible to common method variance (Duffy, Scott, Shaw, Tepper, & Aquino, In press). Thus, common method variance is not likely to be a plausible alternative explanation for the present set of findings.
Second, because the current study is cross-sectional by design, I cannot completely
rule out alternative causal mechanisms to the ones I proposed. For example, while this
research proposes that employees’ perceptions of value congruence with their leaders enhance their voice behaviors, it is plausible that an employee’s active opinion expression and
interaction with his/her leaders promotes his/her value congruence with his/her leaders. To make stronger causal inferences, future longitudinal and experimental research would clearly help strengthen the inferences drawn from this study.
A third issue, which is both a strength and limitation, pertains to the generalizability of my findings. As mentioned earlier, I examined my hypotheses by collecting a sample from military units in Taiwan. The benefit of having data from a single organization is the ability to control for organizational confounding variables. However, this restriction of range might have constricted the relationship observed in the present study. Therefore, future research may want to collect data from other different organizational contexts, such as healthcare, high-tech companies, and insurance institutions where issues of authentic leadership can be equally important drivers of voice behaviors.
Fourth, aside from leader-member exchange (LMX), it is plausible that other forms of leadership constructs can influence follower outcomes. Future research could benefit by collecting measures of other related leadership styles that have been found to positively relate to authentic leadership to examine whether authentic leadership explains additional unique variance above and beyond other positive forms of leadership theories such as
transformational and ethical leadership.
Fifth, expanding my model to include a broader array of antecedents of authentic leadership would be worthwhile endeavor. For example, an individual being perceived as an authentic leader among subordinates can be predicted using his/her personal characteristics such as conscientiousness and neuroticism. Further, future research could also examine how the psychological mechanisms of value congruence interact with followers’ regulatory focus
and subsequently influence a wider range of positive workplace behaviors such as creativity and “in-role” performance. I might expect that PS value congruence interacts with promotion focus to be more closely related to creativity, and PS value congruence interacts with
prevention focus to be more closely related to in-role performance. Hence, in future work, it will be significant to expand the nomological work of authentic leadership by considering these potential independent and dependent variables.
Last but not least, another important next step for future research is to investigate whether my findings can be extended to other cultural contexts. Although my intention was not to conduct a cross-cultural comparison research, examining my theoretical model in the Chinese context may have influenced my results nonetheless. Given that Chinese society is characterized as high power distance (Aryee, Sun, Chen, & Debrah, 2008), employees in the Chinese context are more likely to maintain a formal relationship with their leaders that could restrict their meaningful interactions with authentic leaders (Walumbwa et al., 2010) and regard voice as a particularly risky behavior (Liang et al., 2012). As a result, authentic leadership may have less influence on employee voice. I suggest that future research should cross-validate the present findings using samples from different cultures to see whether the pattern of my findings is unique to my research context.