The data consists of features of about 500000 students. This data is based on PISA 2012 survey which consisted of a paper-based test taken by students, questionnaires taken by school principals to provide information about the students' backgrounds, schools and learning experiences and the broader school system and learning environment. All this information was gathered in the data set used in this analysis. After the data wrangling process there were 485490 students questioned with 17 selected features. Most of the variables are categorical. Some like scores and the size of the class are numeric features.
In this project, I've investigated scores in math, science and reading of about 500 000 students from more than 60 different countries. To help investigation I had considered features like personal possessions of students, personal trades, the gender of students, class size and others. I found out that:
The distributions for all scores appeared to be uni-modal. Gender distribution for each subject was similar for all scores with slightly better scores in math and science for male students and in reading for females. There was a strong and positive correlation between any pair of the three subjects for each gender. Through 3 subjects the list of countries with the highest average score was very similar. In all 3 plots the highest average was by Shanghai(China)followed by Singapore and Hong Kong (China)in math and vice versa for science and reading. The lowest averages came from Peru. It was at the bottom of the distribution plot for all of the subjects. Countries with higher averages in all subjects came mostly from regions of East Asia, Oceania, West, North and Central Europe. Countries with the lowest averages in all subjects came mostly from South America, the Middle East and Balkans. Most of the countries had even average score distribution between genders. There were some exceptions: UK, Vietnam, Mexico, Jordan, Hong Kong, Canada and Brazil had female students average score higher than male counterparts. There was a difference in the distribution of average scores between students who are never late to school and who are late, especially more than 3 times. When gender was considered for this distribution results for both male and female students were almost the same: students who never or once/twice late have a higher average score. Female and male students had the same sense of belonging to school with a number of females who agree that they belong to school slightly higher and slightly lower for strongly disagree. No matter of gender students who felt like they belong at school performs better through all the subjects. Female and male students had almost the same perseverance responses. But female students were more like to give up easily. Also both genders had higher scores if they tend not to give up easily and vice versa. Students with possessions of a private room, computer or books at home had a higher average score in all subjects. But the difference was not very significant especially for the possession of the private room. Both students and their parents agreed on the importance of math for the future. Students disagreed about it more than parents did. There was no correlation found between class size and the average score of the students Students who received teachers' feedback and who were given different tasks than other students were found to perform worse than students who didn't receive feedback or were never given a different task.
For my presentation I focused on the country and gender of students. For an introduction of data firstly I've presented all the countries which participated in PISA assessment with the number of students from the corresponding country in ascending order. Also I've provided gender distribution in the data. Afterward I've introduced scores for each subject: math, science and reading. First their distributions and then 10 countries which performed best on average for each subject. Then I've presented subjects' distributions by gender and at last pair-by-pair relationship between math, reading and science scores with gender aspect. For the presentation I have changed sizes of titles and ticks on the scales, as well as color of some plots.