Student profiles PISA 2012 data analysis.

Intro

This is a student project from the Udacity Nanodegree "Data Analyst". The goal of this project is to demonstrate the importance and value of data visualization techniques in the data analysis process in the exploratory and explanatory process.

Concepts to understand better when I am going to use data visualization:

• Exploratory data visualization: occurs during and after the data wrangling process. Is the main method to understand the patterns and relationships present in your data.
• Explanatory data visualization: after generating your findings, these visualizations help communicate the results.

Documents I received to explore the dataset:

• Dataset: PISA 2012 dataset, download here.
• Dataset dictionary, download here.
• Resources I used to know more about PISA test: pisaprodocuts and pisa_technical_report.

Dataset relevant info

PISA 2012 dataset info:

I wanted to know the student profiles that took the PISA 2012 test.

PISA 2012 data has the tests of 485,490 student representatives from 64 countries. 635 variables are evaluated. The variables I chose to analyze are the type of population (native or emigrant), countries, scores, wealth and truancy.

Main Feature of interest to investigate:

• Which continents and countries have the best and worst educational level?
• Immigrant students have a worse educational and social level than native students?
• Is absenteeism related to a worse educational and social level?
• Is wealth related to a better educational level?

Main findings I got from my exploratory data analysis:

• Scores are strong correlated: the higher a student's math score, the better the student's science or reading score.
• First Generation of immigrants is the type with the lowest scores in all the subjects while Natives are the type with the highest scores in Reading and Science.
• Immigrants observe 20% more differences between Host and Heritage Cultures than Natives.
• Asia has the top and 3 of the lowest scored countires.
• Most countries with fewer scores have less wealth.
• Natives have a longer poverty interval than immigrants.
• The greater the truancy, there is a small correlation with less wealth and a clear relationship with a lower score among immigrants.
• Scores are negative correlated with truancy.
• The smallest difference in scores between natives and immigrants occurs among those who skip classes once or twice.

My analysis:

• pisa_exploratory_data_analysis.ipynb (and its html version) exploration of dataset using Python visualization libraries. Starting from plots of single variables and building up to plots of multiple variables.
• pisa_slide_deck_ppt.html (and its ipynb version) is a short presentation that illustrates relationships and properties that I discovered in the dataset.
• pisa_slide_deck_presentation.ipynb is another kind of presentation illustrating the same as pisa_slide_deck.html.
• output-toggle.tpl file to be able to create the pisa_slide_deck_ppt.html version.