Education in Brazil has a history of bad results. The most recent Programme for International Student Assessment (Pisa), released in 2018, shows that two-thirds of 15-year-old students know less than the basic in Mathematics, and the country has nearly not evolved in Reading and Science in the last 10 years. But in such a disturbing scenario, where are the good examples? And what can we learn from them? Motivated by those questions, we analyzed data about every student, school, city and state in Brazil over months to find out and tell what was behind the few cities with positive results.
Although the analysis encompassed data from Brazil as a whole, we set our focus to the state of Rio Grande do Sul, where our publication (Zero Hora / GaúchaZH) is located. We went out to visit the cities with the best results in the country’s Ministry of Education own assessments and were surprised to find out that the smaller cities appeared as both good and bad highlights. After investigating, we discovered that those with the best scores in Portuguese language (including Reading and Writing) and Mathematics during Elementary Education shared a lot of the same strategies and beliefs.
Another surprise was that the schools with some of the best results were part of the public system – highly regarded as inferior to private institutions in Brazil.
So one of our objectives was, through constructive journalism – although not at all ignoring the fact that education in Brazil suffers from bad results, but trying to point out ways to overcome this –, to allow every citizen to see what worked and what did not. We tried to do this with easy to read maps and tables, presented along stories that, despite relying heavily on data, were told in a very friendly, perhaps engaging way, with data visualization that was easy to understand and still very informative.
After it was published, the article was highly regarded as an example of where good initiatives in education are, why they work and how others can take inspiration from that to apply on their own cases. The State Education Council (CEEd) praised the effort and started referring to the articles discoveries when planning new policies. We heard from a number of schools and cities that have also taken inspiration from what we showed and started working towards better education taking it as a starting point.
Good data visualization was of uttermost importance for us from the start: we wanted to make heavy use of data, but when the time came to showcase it, everything had to be very understandable, easy to read, inviting – friendly, as we put it.
It all started with data scraping from the Ministry of Education’s database, something not publicly accessible when it comes to data about every student and school (the general results are presented every year by the government, but they focus on state and nationwide numbers – we had to go deeper). We filed a freedom of information law request to do so, which took a while to be responded, but fortunately did not involve any major inconveniences.
We initially used Excel to go through data available about Rio Grande do Sul, but went to R when it came to numbers about all of Brazil.
The intention was to use interactive maps and search engines in which readers could not only see a very visual version of the information presented (as seen on maps that show how early years are doing somewhat fine, with maps mostly painted green, while the last years of Elementary School are doing bad, with maps mostly painted red), but also quickly look for their own cities of interest. And later we realized we could also use some focus on the numbers themselves and decided to create scatter plot maps where this was all very visible.
We used technologies and tools such as JQuery, Json and Python throughout the analysis and also for the final version of the project.
What was the hardest part of this project?
The hardest part was gathering all the data: at first, because of the sheer number of entries – every student enrolled from the 1st to 9th year of Elementary School in every educational institution in Brazil, a country with a population of 209 million – and the need to figure out the scores based on classes from the same school, them from that whole year, then for all school in a state, and then of every school in the country. The project was done with a regional focus, but encompassed data from the entire country and its millions of students. That effort also means the data gathered and conveyed as this story can be used as a reference for anyone in Brazil.
In order to avoid false positives, we also analysed data from earlier years, so we could verify if the cities with the best results were actually constantly achieving the educational goals, not only having one good class during one exam, for example. The cities and schools we visited have shown good results over at least the past 6 years, as shown in three different national education assessments in this decade.
Data used was not publicly accessible, so we had to file a freedom of information law request, something that is hardly done when it comes to checking how well students are doing in the early years of education in Brazil. Finally, it was initially hard to convince all editors that the approach we decided to take was a good one: when newsrooms are usually so concerned about highlighting failures and trying to find out what is wrong, we went the other way and showed what was working, so that the article could be used as a sort of guide to those interested in achieving better results in education.
What can others learn from this project?
This project was not thought about with any audience in particular coming to mind: teachers, school principals, politicians, the government in general and any reader interested in education can find their own use of the gathered material.
But it is not only that: the extensive data analyzed, presented in a friendly manner, could also be used by reporters in our newsroom and even journalists from other outlets to work on their own stories. And time and time again we saw this happening: whenever a city faced a strike from teachers, for example, we would reach out for those maps to see how well their educational system works. Whenever projects related to education would show up at any Parliament, we would check the results of Rio Grande do Sul and Brazil in general to see where there were problems, and what was working.
We have also learned of numerous teachers and schools who referred to this project since its publication when establishing their goals for a school year, seeing the examples in there and often reaching to the cities we visited to learn more about those projects. The very State Education Council (CEEd) would encourage this behaviour.
A Word document with the entire translation of the project in English is available as a project link below in Google Drive and as a backup in OneDrive.