Evergreen I
Multilevel Modeling Techniques for Standardized International Assessments (306513)
*Noelia Nirza Pacheco-Diaz, University of TennesseeLouis Marie Berry Rocconi, University of Tennessee
Keywords: Big Data, Multilevel Modeling, PISA
This poster will discuss techniques for using hierarchical linear models or multilevel models in conjunction with international standardized tests. Multilevel modeling is a statistical approach that can be used to analyze data which are clustered or nested within groups (Hox, Moerbeek, & De Schoot, 2018). Since students are typically nested within schools and schools are nested within districts, states, and countries, multilevel modeling is an ideal statistical tool to analyze these data. This poster presentation will discuss the use of such method when working with data from the Program for International Student Assessment (PISA). PISA is an international assessment that examines the literacy of high school students in reading, math, and science (OECD, 2017). This poster will provide the audience with an overview of the steps necessary to use multilevel modeling for examining international assessments and the appropriate software to conduct the analysis. Additionally, this poster will examine issues with working these types of data such as working with missing data, plausible values, and centering.