The data consists of information regarding 485490 students that took part in the PISA 2012 test, they provide answers and inputs for around 500 variabels. The data was wrangled to reach only 30 variables like country, gender, their maths score,...etc. We had to convert some data from categories to numeric to combine them. In the first notebook file (preliminary_wrangling), we tried to clean our data for the analysis part, which can be found in the "exploration_work" notebook. We then used our key findings and visualizations to prepare a presentation and a slideshow. The data can be found in this google document in the PISA 2012 section. Data:https://docs.google.com/document/d/e/2PACX-1vQmkX4iOT6Rcrin42vslquX2_wQCjIa_hbwD0xmxrERPSOJYDtpNc_3wwK_p9_KpOsfA6QVyEHdxxq7/pub?embedded=True
In the exploration, I found that there was a relationsip between the score of student and the student being Anxious The correlation was Negative, the more anxious you get the less score you acheive. Also, the family metrics(wealth,SQ levels, parent_ed) had a positive moderate correlation with the score. We also found that males tend to get higher scores than females. One more thing that caught our attention was the different effect the anxiety has on both genders. Last but not least, we found that the studets who got to use computers and internet at a younger age mostly outperform those who never used them. We also found that the education level of parents affect the score of their students as the studetns whose fathers have reached higher educational levels on the MISCED, got more scores on those whose parents got a basic education. We also found that some mothers' occupation affected the students score.
For the presentation, we want to emphasize on our original goal and it is to get the variables that affect studetns socore. We tried to explore the score distribution and then moved to make a correlation matrix to test the relationship between variabels. We found some interesting facts like that people who used computer performed those who didn't. We also found that parental education and wealth affect students score. Anxious students acheived lower scores than those who were calm. The fact that calm males are better than calm females and that anxious females are better than anxious males is one of the most interesting key findings in this project.