Did the Internet Just Make the World Smarter?

It’s easy to hate on the Internet. We used to read books. Today, we read Twitter feeds. Before the web, we used our time wisely. Today, we waste time constantly. Twenty years ago, we had thoughtful, in-person conversations, focusing for hours on a single, worthy topic. Today, we fire off nasty, anonymous YouTube comments, laced with sarcasm and typed from the lonely sanctums of our bedrooms, dorms and cubicles. We have, you might say, gotten lazier, meaner and dumber.

Or have we? We set out to compare Internet growth around the world with changes in student test scores. Specifically, we used The World Bank’s 2012 Internet penetration figures (the latest available) and test results from the Organisation for Economic Co-operation and Development (OECD) in our analysis.

Would the Internet’s myriad cat gifs, message boards, and viral videos dull students’ minds, or enhance their logical reasoning and critical thinking skills?

We’ll take a quick look at Internet penetration figures by country from 2000 to 2012. Then we’ll find out whether students’ test scores have improved (or declined) over the same period. Finally, we’ll see what kind of conclusions we can draw.

The following chart compares the percentage of Internet users per country, both in 2000 and 2012. (Note that we limited our sample to the 30 countries where we had full data for both Internet use and test scores.)

Back in 2000, Norway had the greatest proportion of Internet users, at 52%. Meanwhile, Russia had the lowest percentage, at only 2%. In 2012, Iceland led the pack with 96%, while Mexico placed last, at 38%. Note that the chart is ordered from left to right by the percentage increase in Internet penetration* from 2000 to 2012. In other words, Russia (far left) saw the greatest percentage increase in Internet users, while Canada (far right) saw the least.

*Further note: “percentage increase” is calculated by the proportional rise, not by absolute percentage points. So a rise from 25% to 50% is counted as a 100% increase in penetration for our purposes.

Now, let’s look at how each country’s test scores improved (or declined) over the same period. Specifically, we are looking at the Programme for International Student Assessment, or PISA, which measures 15-year-old students’ knowledge and skills in three key areas: reading, mathematics, and science. (Note that the countries appear in the same order as the chart above.)

We see the most significant improvements from Brazil, Latvia, Poland, Luxembourg, Portugal, Germany and Liechtenstein. Meanwhile, the United Kingdom, Iceland, Sweden, New Zealand and Australia actually show sizable declines.

More intriguingly, however, we see that nearly all the big gains occurred on the left side of the chart—that is, those countries with Internet penetration under 20% in 2000, but between 50% and 80% in 2012. Conversely, for countries that already had a sizable Internet user base in 2000, test scores remained level or declined.

Here’s what we get when we plot these two figures:

With a correlation of 0.41, we can safely say that the relationship between PISA score improvement and Internet growth is moderately strong. Big increases in Internet access tended to go hand-in-hand with better scores on the international test. Perhaps the web hasn’t been such a bad innovation after all.

Before we get too excited, however, we should be clear that correlation does not equal causation. While access to the Internet has potentially contributed to better test scores in many countries (perhaps through better access to information, more self-learning, and more ways to connect with classmates), there are several other factors at play, from changes in wealth to improved education systems to other forms of technology outside of the Internet.

And even if we stick to the hypothesis—that more Internet means better test scores—we should note that the hypothesis fails rather obviously among the world’s most plugged-in nations. Yes, the jump from 10% Internet penetration to 50% seems to lead to smarter students, but from 50% to 90%, test scores leveled off or got worse. Even with all those hilarious cat gifs, it appears the Internet might sometimes be too much of a good thing.

Here, for your reference, are the raw PISA scores (out of 600) and Internet penetration:

(Internet usage aside, test score bragging rights belong to Japan, Finland and Liechtenstein, the three highest scoring countries in 2012’s exam. Greece, Mexico and Brazil round out the list as the lowest scoring countries. Better luck in 2015.)

This article was written for TIME by Ben Taylor of FindTheBest.