The Programme for International Student Assessment (PISA) is a worldwide study of 15-year-old school pupils’ scholastic performance in mathematics, science, and reading. Every three years more than 500 000 pupils from 60+ countries are surveyed along with their parents and school representatives. The study yields in more than 1000 variables concerning performance, attitude and context of the pupils that can be cross-analyzed. A lot of data.
OECD prepared manuals and tools for SAS and SPSS that show how to use and analyze this data. What about R? Just a few days ago Journal of Statistical Software published an article ,,intsvy: An R Package for Analyzing International Large-Scale Assessment Data”. It describes the intsvy package and gives instructions on how to download, analyze and visualize data from various international assessments with R. The package was developed by Daniel Caro and me. Daniel prepared various video tutorials on how to use this package; you may find them here: http://users.ox.ac.uk/~educ0279/.
PISA is intended not only for researchers. It is a great data set also for teachers who may employ it as an infinite source of ideas for projects for students. In this post I am going to describe one such project that I have implemented in my classes in R programming.
I usually plan two or three projects every semester. The objective of my projects is to show what is possible with R. They are not set to verify knowledge nor practice a particular technique for data analysis. This year the first project for R programming class was designed to experience that ,,With R you can create an automated report that summaries various subsets of data in one-page summaries”.
PISA is a great data source for this. Students were asked to write a markdown file that generates a report in the form of one-page summary for every country. To do this well you need to master loops, knitr, dplyr and friends (we are rather focused on tidyverse). Students had a lot of freedom in trying out different things and approaches and finding out what works and how.
This project has finished just a week ago and the results are amazing.
Here you will find a beamer presentation with one-page summary, smart table of contents on every page, and archivist links that allow you to extract each ggplot2 plots and data directly from the report (click to access full report or the R code).
Here you will find one-pagers related to the link between taking extra math and students’ performance for boys and girls separately (click to access full report or the R code).
And here is a presentation with lots of radar plots (click to access full report or the R code).
Find all projects here: https://github.com/pbiecek/ProgramowanieWizualizacja2017/tree/master/Projekt_1.
And if you are willing to use PISA data for your students or if you need any help, just let me know.