The Relationship Between Gender Match and Leader-Member Exchange in Supervisor-Subordinate Dyads from Information Technology

There is abundant evidence that leader-member exchange (LMX) is advantageous for subordinates (e.g., Gerstner & Day, 1997; Ilies, Nahrgang, & Morgeson, 2007). However, a basic tenet ofLMX theory is that LMX levels vary among subordinates (Dienesch & Liden, 1986; Graen & Uhl-Bien, 1995). Male-dominated professions may be a context in which women find it challenging to develop and maintain high-quality LMX. Indeed, gender mismatch within the supervisor-subordinate dyad predicts outgroup status (Duchon, Green, & Taber, 1986), whereas gender match is positively related to important outcomes, such as subordinate performance and reduced role confiict (e.g., Tsui & O'Reilly, 1989). This research aimed to examine the relationships between gender match, LMX, work interference with family (WIF), job satisfaction, and turnover intentions in supervisor-subordinate dyads from information technology (IT), a heavily male-dominated field. This was accomplished by testing two predictions suggested by similarity-attraction theory and social identity theory (SIT; D. E. Byrne, 1971; Tajfel & Turner, 1979; Tsui & O'Reilly, 1989). Structural equation modeling was used to test the proposed model using data from a sample of 430 information technology employees from 10 high-tech organizations across the U.S. Contrary to both hypotheses supported by relational demography theory, the findings suggest that supervisor-subordinate dyads in IT that include a woman, whether supervisor, subordinate, or both, report higher quality leader-member exchanges than supervisor-subordinate dyads comprised only of men. Given the unanticipated results, it is imperative for future research to reconsider if relational demography accurately describes gender (dis)similarity effects at the dyadic level, and to further examine processes underlying the development of supervisor-subordinate relationships in technology-dominated fields by employing qualitative and longitudinal research methods.