Journal of education finance · 2007 · 24 citations · 0 references
EducationEducational MeasurementEconometricsStatistical ModelingStatisticsCorrelational ResearchData AnalysisPerformance In EducationProductivity AnalysisStatistical AnalysisStudent Achievement AnalysisCanonical Correlation AnalysisCanonical Correlation PrincipleEducation Production FunctionsInput VariablesOutput VariablesProductivityStudent Learning Outcomes
Researchers have long searched for a relationship between school resources and student achievement. The search for this relationship has relied heavily on the use of education production functions. Despite years of research on this topic, there has been little agreement concerning which resources are the most essential variables to increase education productivity. This article introduces a new method of statistical analysis to respond to this question, canonical correlation, in an attempt to bring clarity to the debate. This method allows the researcher to better analyze the complex nature of schooling by considering multiple inputs and outputs at the same time. A statistically significant relationship was found between the linear combination of input variables and the linear combination of output variables in two of the seven canonical correlations calculated from the data. The total pooled variance explained by those significant correlations was estimated to be j6% of the total variance shared between the two sets of variables. Redundancy calculations suggested that the input variables predicted 18% of student outcomes.