Some limitations or our study are based in the data available from primary research. Probably the biggest drawback of our meta-analysis is that all our analyses are based on correlation. Most available research into destructive leadership is cross-sectional in nature (with the notable exception of some of tipper’s studies). We therefore can not draw conclusion about causal effects. In addition, studies on destructive leadership and its outcomes tend to use self-report data only. This means that there is a potential for a method bias. However, the correlation difference we found between the different outcome variables will likely be similar but might differ in strength when method bias is avoided. While using self-report data only is a clear limitation, the issue lies in the nature of the research: A lot of the constructs summarized under destructive leadership explicitly refer to the perception of leader hostility as a defining attribute of destructive leadership. In addition, some of the outcomes are highly subjective.
Give that the field is relatively new, there were limits as to the possibility of analyzing moderations. We found large within studies heterogeneity which point to moderating influence. However given the limited amount of moderators included in primary research, we were not able to include moderators here other than one on the measurement of destructive leadership.
Relatedly, we summarized different outcomes into categories to achieve a larger number of studies included in our individual analyses. However, that means that we cannot say much about differentiated effects. For example, under the heading of counterproductive work behavior different behaviors such as organizational and interpersonal deviance, active as well as passive counterproductive behavior are summarized. This limits in how far differentiated relationships between destructive leadership and outcomes could be examined.
Some limitations or our study are based in the data available from primary research. Probably the biggest drawback of our meta-analysis is that all our analyses are based on correlation. Most available research into destructive leadership is cross-sectional in nature (with the notable exception of some of tipper’s studies). We therefore can not draw conclusion about causal effects. In addition, studies on destructive leadership and its outcomes tend to use self-report data only. This means that there is a potential for a method bias. However, the correlation difference we found between the different outcome variables will likely be similar but might differ in strength when method bias is avoided. While using self-report data only is a clear limitation, the issue lies in the nature of the research: A lot of the constructs summarized under destructive leadership explicitly refer to the perception of leader hostility as a defining attribute of destructive leadership. In addition, some of the outcomes are highly subjective. Give that the field is relatively new, there were limits as to the possibility of analyzing moderations. We found large within studies heterogeneity which point to moderating influence. However given the limited amount of moderators included in primary research, we were not able to include moderators here other than one on the measurement of destructive leadership. Relatedly, we summarized different outcomes into categories to achieve a larger number of studies included in our individual analyses. However, that means that we cannot say much about differentiated effects. For example, under the heading of counterproductive work behavior different behaviors such as organizational and interpersonal deviance, active as well as passive counterproductive behavior are summarized. This limits in how far differentiated relationships between destructive leadership and outcomes could be examined.
การแปล กรุณารอสักครู่..