| Most studies examining the relationship between leader behavior and leader-member exchange (LMX) have examined only one or two leader behaviors, and many studies have employed composite leader behavior measures. The exclusive use of composite measures eliminates the examination of more distinct behaviors that may have substantially different effects on LMX. For example, composite measures of transformational leadership behavior often include a broad range of distinct leader behaviors, such as supporting, developing, leading by example, envisioning change, and encouraging innovative thinking (Bass & Avolio, 1990). The composite measures of transformational leadership do not include some important relations-oriented behaviors, such as consulting, recognizing, and delegating (Yukl, 1999).;My research included a more comprehensive set of leader behaviors than any of the previous studies published on LMX, and it examined the independent effects of specific leader behaviors on LMX when multiple other specific leader behaviors were controlled. Five relations-oriented leader behaviors (supporting, recognizing, developing, consulting, and delegating) were hypothesized to relate positively to the quality of LMX. Some task-oriented behaviors (clarifying, short-term planning, and monitoring operations) and transformational behaviors that are not primarily relations-oriented (leading by example, envisioning change, and encouraging innovative thinking) were included to determine if they added anything to the explanation of leader effects on LMX.;To examine these relationships, two survey studies were conducted, and each study used a different measure of LMX. In the analyses for meta-categories, the relations-oriented behaviors accounted for more of the explained variance in LMX than the task-oriented behaviors or transformational leadership behaviors that are not primarily relations-oriented. In both studies' analyses for specific behaviors, LMX was significantly related to supporting, delegating, and leading by example. Recognizing and consulting were related to LMX in the first study but not in the second study. Results were not significant in either study for developing, envisioning change, or the task behaviors. Encouraging innovative thinking was not significant in the regression analysis of specific behaviors in the second study. The practical applications for managers are to identify the specific types of behavior that will be most useful for improving the quality of an exchange relationship with a subordinate. |
