| The purpose of this study was to determine the extent to which collective teacher efficacy explained variance in middle and high school discipline, as measured by out-of-school suspensions, over and above that explained by school size and race. To achieve this purpose, non-experimental survey research methodology was utilized. The sample population for this study included 42 high and middle schools in one large suburban Virginia school district.;The unit of analysis for this study was the school. The school district conducted a system-wide survey to collect Collective Teacher Efficacy scores in August 2010 in all of its 196 schools. The researcher sought and received permission to access archival data for the 42 middle and high schools serving as the sample in this study. Out-of-school suspension data were retrieved from the Virginia Safe School Information Resource (2010), a publicly available database maintained online by the Virginia Department of Education. Suspension data that were a result of legally mandated requirements for out-of-school suspension were not included in the study. Demographical data were collected from the school system's website.;Hierarchical regression analysis revealed collective teaching efficacy, and its two constructs of group competence and task analysis, accounted for statistically significant variance (p < .05) in discretionary out-of-school suspensions given for serious student offenses when controlling for school size and race. An additional analysis found that collective teaching efficacy, group competence, and task analysis did not explain statistically significant variance (p < .05) for discretionary out-of-school suspensions involving lesser student offenses when controlling for school size and race. Correlation analyses found that race was strongly correlated with discretionary out-of school suspensions, a finding consistent with the literature.;A number of implications for practice and future research were discussed, including a recommendation for employing a research design using larger sample sizes to expand the external validity of the findings. |
