Developing a Gini Coefficient for Distributed Instructional Leadership (GDIL): How Distributed Instructional Leadership (DIL) Impacts Instructional Reform Implementation and Professional Community in Elementary Schools.

Abstract

This dissertation examines the notion of school distributed instructional leadership (DIL). Most discussions about distributed leadership focus on the “average leadership” exercised by multiple roles or individuals to conceptualize and measure the construct. However, scant attention has been paid to “dispersed leadership,” which estimates the degree to which leadership is equalized (or decentralized) across multiple roles or individuals. Nor does previous research provide a robust theory or empirical evidence regarding the effects of “dispersed leadership,” either on instructional improvement directly or conditional on certain features of school context.

To address these gaps, I developed a quantitative measure for the “dispersed leadership” of DIL — the Gini Coefficient for Distributed Instructional Leadership (GDIL). The GDIL measures the degree of “equality” to which instructional leadership functions are distributed across multiple roles or individuals. Then, using contingency theory as a guiding framework, I developed a theory about the effects of DIL on two school outcomes (i.e. “fidelity” to instructional regime and strength of professional community), contingent upon four features of school context: average instructional leadership [AIL], task, leader-leader interaction, and leader-teacher interaction. Finally, I tested the theory empirically in a series of longitudinal, multilevel models. My empirical inquiry regarding DIL was based on four-year longitudinal data on 109 elementary schools that adopted one of three Comprehensive School Reform (CSR) programs (i.e. America’s Choice [AC], Successful For All [SFA] or the Accelerated Schools Project [ASP]). The findings indicated that the influences of DIL on the two outcomes were conditional on the four school contingencies. However, the conditional effects of DIL not only varied across the four contingencies but also varied between the two outcomes.