< Back to previous page


Determining sample requirements for modeling and reporting (stability of) school value-added estimates: sample size, inter-school student mobility and sample heterogeneity

Within the context of school accountability systems, many countries use value added (VA)-estimates as indicators of the educational quality or effectiveness of their schools. A fundamental issue in this field is the (in)stability of VA estimates across time, i.e., the extent to which school effects persist over successive cohorts of students. VA-estimates are less stable, i.e., less reliable, especially for primary schools and for schools that are located at the low end of the performance distribution. As such, for schools that are already at risk the stakes become very high, because they might be unfairly targeted for negative sanctions. Although research results start to indicate that the sample is related to the stability of VA-estimates, specific guidelines on which sample characteristics to account for when modeling and reporting (the stability of) VA-estimates are missing. Therefore, the aim of this project is to investigate the effect of school sample size, student inter-school mobility and school sample heterogeneity on the stability of VA-estimates.
Date:18 Dec 2020 →  Today
Keywords:value added, sample requirements, educational effectiveness, high stake testing, reliability stability
Disciplines:Methodology of pedagogical and educational research, Mesolevel instructional sciences, Macrolevel instructional sciences