Jeff's Lunchbreak: Science and Engineering Indicators 2016

NSB Logo The NSF has released their Science and Engineering Indicators report for 2016. It's a great report put out every two years documenting many aspects of Americans' relationship to science and engineering. For the past several reports (2004, 2006, 2008, 2010, 2012, & 2014), I've made it a habit to examine one specific aspect - public understanding of science. In particular, I've examined the data on how many questions people can correctly answer on a short quiz of basic scientific questions, how that has changed over the years, and how the U.S. fares against other countries on those (mostly the) same questions.

For all of the tables I'm about to publish, note that I copied the data and notes from the NSF report, but I've formatted the tables to fit onto this blog. I made the graphs myself to help visualize the data, as these particular graphs weren't in the report.

First, here's the table showing how Americans fared on a question by question basis on some basic scientific facts. The table includes data from 1988 on up to the most recent poll in 2014.

Those results aren't particularly encouraging. I point this out nearly every time I cover this report, but around 1 in 4 Americans don't know that the Earth orbits the Sun, and around half of Americans don't know that electrons are smaller than atoms! Those are simple, basic, scientific facts.

To help visualize that data, especially the trends on how it changes over time, here it is plotted on a graph by year.

Americans' knowledge has remained largely steady over the past decade and a half, though there were a few changes. Americans' knowledge on antibiotics improved the most, but has kind of plateaued since around 2006. There does appear to be a recent trend of improvement on the questions concerning the Big Bang and human evolution. Hopefully that trend is real and continues on into the future.

Next, here's the table showing how the U.S. compared to other countries.

I played around with different ways of plotting that data, but there's just so much that it's too confusing to put it all on one graph. If you're interested in seeing a graph for each individual question, you can click on the thumbnail below to embiggen* the graphs.

Click to embiggen

However, I did come up with a way to do a comparison of sorts - I took an average of the percentage of people that correctly answered questions. As an example, if it was only two questions, and 100% of people answered the first question correctly, while only 50% answered the second question correctly, the average would by 75%. I did this average three ways - overall, the physical science questions, and the biological science questions. If a country didn't pose a certain question, it wasn't included in that country's average. I admit that this is a very rough way to do a comparison, but here's how each country fared.

The U.S. actually does rather well in this comparison. It's not number 1, but it's not too far off.

I also suspected that America's over-religiosity might be affecting those questions that contradict a literal young earth creationism interpretation of the Bible, so I redid all those averages exluding the Big Bang and evolution questions.

As suspected, this did improve America's performance. This is heartening, that creationism hasn't caused huge damage to Americans' scientific understanding overall.

One lesson from this that I've pointed out before, is to keep these results in mind every time you see a poll showing people's attitudes towards anything scientific. For example, every time you see a poll showing that the majority or plurality favor teaching creationism in public schools, or a poll showing high levels of skepticism towards global warming, remember that this is the same public where a quarter of all people think the Sun orbits the Earth, and where half of all people don't realize electrons are subatomic particles. How informed can they be on scientific issues when they don't even know such simple facts?

The other major lesson is that we need to do a lot better job of teaching science. When you live in a democracy and everyone has a say in the government (at least by way of voting for representatives), you really need a well educated populace for it to work effectively. This is especially true of science in the modern age, when so many pressing issues require accurate understanding of science.

I suppose that on the plus side, as much as alarmists decry the falling quality of American education, at least in this one area, the data shows that Americans' knowledge has stayed largely the same. There's definitely room for improvement, but at least we haven't gone backwards.

*'Embiggen' is a perfectly cromulent word.