Addressing an old issue from a new methodological perspective: a proposition on how to deal with bias due to multilevel measurement error in the estimation of the effects of school composition

Thesis

With educational effectiveness studies, school-level aggregates of students' characteristics (e.g. achievement) are often used to assess the impact of school composition on students' outcomes – school compositional effects. Empirical findings on the magnitude and direction of school compositional effects have not been consistent. Relevant methodological studies raise the issue of under-specification at level 1 in compositional models - evident when the student-level indicator on which the ...

Expand abstract

Authors: Televantou, I

• Publication date: 2014
• DOI: 10.5287/ora-1zbv14mnr
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: regression discontinuity approach; multilevel Structural Equation Modelling; sampling error; phantom compositional effects; Big-Fish-Little-Pond-Effect; hierarchical structures; measurement error
• Subjects: Statistics (social sciences); Education; Self-concept enhancement