zen and Fritz (2004), in which was stated that Bauer and Green (1996) were disappointed by their results. These authors stated their disappointment, but it is likely that other authors leave out certain results or information due to such disappointment. Another limitation of the studies discussed is the lack of longitudinal studies, as this type of study is necessary to detect variable patterns over time. Moreover, the lack of control groups of some studies is unfortunate, as control groups are essential for comparison.

On top of the limitations of the studies discussed, the literature review itself has some limitations as well. First of all, some literature discussed could be seen as rather outdated. Leadership is a very dynamic research field that is constantly changing, thus literature that is more recent is very crucial. However, this could also be considered a strength as it is also important to look at the bigger picture, for which older literature is essential. Secondly, not all relevant literature was read and discussed. Articles and studies were selected based on apparent relevance, however, it is likely that some important information and studies are missing from the review.

Some of the literature discussed focused more on the gender of the leader, other on the gender of the subordinate, and other on both. There were also papers that did not clearly distinguish leader and subordinate gender. Most studies discussed female-male, female-female, and male-male dyadic relationship without explicitly researching the role of the subjects. It is plausible that, for example, a female leader-male subordinate relationship differs significantly from a female subordinate-male leader relationship. Future research should look more into both subordinate and leader gender, and whether these influence dyadic relationships differently.

An important step that has to be made by future researchers is the improvement of study design. All the limitations discussed above are to be considered and dealt with. This means that longitudinal studies should be conducted, with larger and more generalizable samples, including control groups.

Conclusion