Dağıtımcı Liderlik Envanterinin Türkçe Uyarlaması: Geçerlik ve Güvenirlik Çalışmaları

Abstract

Introduction. Leadership process is an important variable in order for accomplishing the organizational goals (Bass a Avolio, 1994; Rukmani, Ramesh a Jayakrishnan, 2010). There is also positive correlation between effective leadership behaviors of school administrators and reaching the schools' targets (Bossert, Dwyer, Rowan a Lee, 1982; Edmonds, 1979; Hallinger a Heck, 1998). For these reasons, studies on educational leadership have increased for the last thirty years (Andrews a Soder, 1987; Hallinger, 1990, 2003, 2010). Traditionally, leadership studies have mainly concentrated on the solo leaders. According to this approach, the leaders have been conceptualized as if they were supernatural hero. Similarly, the leadership theories including trait, behavioral and contingency examined the leadership processes just concentrating on the leaders and followers. Inaddition, transformative, transactional and instructional leadership studies have also tried to understand the effect of leaders on organizational outcomes (Bass, 1985; Burns, 1978; Hallinger a Murphy, 1985). However, the solo leadership approaches have been criticized by some scholars recently just because they are not suitable for the modern organizations (Elmore, 2000; Gronn, 2000, 2002; Harris a et al., 2007; Spillane, 2005). These authors have introduced an alternative leadership paradigm which they named it as distributed leadership. According to the theorists of distributed leadership, leadership processes is distributed across whole organization in modern organization. Especially division of labor, increased organizational levels and organizational complexity has played a determining role in appearance of the distributed leadership paradigm. When it is examined in educational context it is argued that schools should not be administered solely by the principals, but the other stakeholders such as assistant principals, teachers, parents and even students should also participate in the decision-making in schools. This new trend led to a new leadership study under the name of distributed leadership especially in foreign literature (Haris et al., 2007; Hulpia, Devos a Keer, 2011; Lashway, 2003; Spillane, 2006). But, there is a lack of empirical research in Turkish literature on distributed leadership. There are only few studies concentrating on the distributed leadership in Turkish literature (e.g. Baloglu, 2011; Korkmaz a Gunduz, 2011). When examined in detail, it is ederstood that those studies are not directly measure the distributed leadership levels at schools. Therefore it is assumed that there is a need for measuring the distributed leadership level in Turkish schools. However, there is not a suitable scale for this purpose in Turkish context.Purpose. The purpose of the study was to adapt a useful distributed leadership scale for Turkish schools. In line with this purpose, it was decided that the Distributed Leadership Inventory (DLI) developed by andet al. (2009) could be one of those tools. Therefore, present study aimed to investigate whether DLI was a valid and reliable data collectin tool in Turkish culture. Method. 160 voluntary teachers from eight schools located in four districts of the province of Ankara participated in the current research. Data were collected with the Turkish version of DLI which was originally developed by Hulpia and et al in 2009. The 23 items-DLI is composed of two sub-scales; one is leadership functions (13-items) and the other is leadership team characteristics (10-items). The original DLI was firstly translated into Turkish and then Turkish form was re-translated into English by experts in educational administration and English language. After translations process, the DLI wasadministered to the participants by the researcher. 160 scales were analyzed with exploratory factor analysis (EFA) to determine the factor structure of the sub-scales of the DLI. The reliability studies were calculated with Cronbach's Alpha coefficient, two half test correlation and Sperman-Brown reliability coefficient. Statistical calculations were made wih SPPS. 18 version. Findings and Discussion. Factor analysis of the DLI showed that the subscales of the DLI were made of single factor structure. Leadership function sub-scale explains 69.67 percent of the variation for the principals, 63.26 percent for the head assistant principals and 70.15 percent for the assistant principals. In addition, Cronbach alpha internal consistency coefficient values were calculated as follows: [a = .96] for principals; [a = .95] for head principal assistants and [a = .96] for assistant principals. Factor analysis of the leadership team characteristics revealed that this sub-scales is also made of single factor structure and explains 81.92 percent of the variation. Cronbach's Alpha internal consistency value of this sub-scale was calculated as .98. All these findings are parallel with the findings of the original DLE study (Hulpia et al., 2009). Conclusion. With the present study it is understood that Turkish version of DLI can be used as valid and reliable data collection tool in distributed leadership studies in Turkish school.