Data Files for OECD-PISA 2012 Study--Top Tier, Canada, and US
This folder contains multiply imputed data sets for the top-tiered performers in math literacy on the PISA 2012. It also includes Canada and the US data. The purpose of this research was to identify significant predictors of math literacy for top-tiered performers, Canada, and the US. It was hypothesized that top-tiered countries engage in academic behaviors that may account for, in part, the optimal performance on the PISA assessment. Comparison of the US on these significant predictors may offer some insight into math-literacy performance in the US, which is below the OECD average.
Steps to reproduce
Included in this data folder is the following: 1) R Code to reproduce the LASSO analysis. LASSO is a sparse regression, variable selection technique that identifies significant predictors in high-dimensional data sets. LASSO was used to identify the significant predictors. 2) An Excel file with the LASSO results for each country. Separate analyses were run on the student- and school-level predictor variables. 3) A zip file of the multiply imputed data sets for the 12 countries, including aggregated data sets.