Is Quality/Effectiveness An Empirically Demonstrable School Attribute? Statistical Aids for Determining Appropriate Levels of Analysis

Abstract

To assist in decisions about appropriateness of data aggregation, this paper describes and demonstrates analytic techniques developed and used in organizational psychology along with contemporary multilevel analysis. Using these analytic techniques, the relations of educational outcomes (student self-reported academic performance, sense of belonging, and educational aspirations) to the school environment (possible indicators of school quality) were examined. Survey data were obtained from students (N = 11,573 or 82% of the eligible students completed the surveys) who attended 31 middle school students. Results showed that at least some indicators might be represented as school-level phenomena, although the amount of variance explained by school membership was quite small. Results implied that the appropriate level of analysis is concept- and variable-dependent. For example, students' sense of belonging may likely develop by how respect and instruction are revealed to students, such as in classrooms. Future studies of educational effectiveness might consider other groupings of students and emerging theoretical frameworks of organizational effectiveness to determine what constitutes educational quality/effectiveness and their measures at specific organizational levels.