Understanding distributed leadership practices in and around teacher teams 5 103 5.3.5. Analysis The analyses for answering our research question consisted of two steps. Firstly, we conducted quadratic assignment procedures (QAP) within teams to study which individual level variables predict advice-seeking, the latter is the measure of distributed leadership practices. We included the significant individual variables to team level. In this way, we were able to include these in the second and main analysis step, namely, to study our unit of analysis: Teacher teams and how their context influences their degree of distributed leadership. Secondly, we analyzed links by comparing teams with higher and lower degrees of distributed leadership and their team and school level sociocultural characteristics. 5.3.5.1. First step: Individual level: Quadratic assignment procedures (QAP) Firstly, we conducted quadratic assignment procedures (QAP), multiple regression analysis, in Ucinet for each teacher team to test the link between individual characteristics and advice seeking relationships, which is our measure of distributed leadership practices. We conducted the QAP’s per teacher team since the teacher team is the unit of analysis and the boundary of the network (which means that persons could not select team members from other teams than their own team). The QAP is suited for social network data, since it can analyze observations that are interdependent, and social network data is interdependent. More conventional statistical tools are not appropriate, because network data violate the assumption of independent observations (Borgatti et al., 2013). All characteristics showed weak (.3) tomoderate (.4–.6) correlation. The individual characteristics that were significant were aggregated to the team level and included in the next analysis step, in which our main unit of analysis, the team level, i.e., teacher teams, are studied. 5.3.5.2. Second step: Team and school level: Analyzing links between distributed leadership and sociocultural characteristics To answer our research question on the link between distributed leadership practices within teacher teams and their sociocultural context, we qualitatively analyzed the links between distributed leadership of each teacher team and its sociocultural characteristics on team and school level. As stated, significant predictors of the individual level were included in the analysis on team level. Preparing two qualitative variables to be included in the analysis. For Themes of advice on team level and Reasons to implement the program on school level we first had to code the data into categories to be able to include these in the main analysis of how sociocultural characteristics linkz with distributed leadership practices.
RkJQdWJsaXNoZXIy MTk4NDMw