One of the best parts about being back in school is that I have full access to UofT’s extensive digital library. After I posted on the 2006 GSS, and how it showed that only 73.6% of surveyed U.S. adults knew that the Earth goes around the Sun, I figured I might like to have a closer look. So I googled the GSS, and was informed by the website that thanks to an agreement they have with UofT, I would enjoy full access to the data. Wow!

Well, turns out the full dataset is 300Mb, and in a format which is designed to be read by high end statistics packages (which I sadly do not have access to yet). So it took me a little while to write some software to begin to look at this data.

Once that was done, I thought I’d start simply — using the earth/sun question which had originally intrigued me so. In the above graph, you can see the % of correct responses to the question plotted against the highest grade level completed by the respondent (5 for 5th grade, 16 for four years of college, &c). The results are both expected and unexpected. The dominant trend of the graph shows a strong and pleasing (at least to those of us who have left comfortable jobs for difficult slogs as born-again students…and who also think it’s important to know who revolves around who) correlation between answering correctly and being educated longer.

Strangely though, the group with only 2nd grade education scored extremely high…as high in fact as those who completed 3 years of college. This group consisted of only 9 people, so it’s easy to write it off as a low-statistics anomaly. But both the 5th grade and the 7th grade groups were smaller, and fit nicely into the correlated trend. Furthermore, it’s easy to see from the error bars that despite the sample’s small size, the high marks were above the linear fit in a statistically significant sense. Sadly, the sample group did not contain any 3rd or 4th grade dropouts, so we can’t see how the data behaves in that gap.

The chart more or less speaks for itself. The good news is that education seems to correlate with heliocentrism. The bad news is that if you continued in school past the 2nd grade in the U.S., you won’t have better than a 50% chance of knowing that the earth revolves around the sun until your Sophomore year of high school.

One Response to “Heliocentrism Revisited”

This is an interesting analysis. I encourage you to pursue this further, so I hope that the following criticisms do not deter you:

1. Your error bars extend into impossible regions of probability below 0.0 and above 1.0.

2. Statistically, a single survey question provides a weak assessor of what you’re really trying to measure.

3. Ignoring philosophical issues and the physical fact the the Earth exerts a gravitational pull on the Sun, just as the Sun does on the Earth, there is the day-to-day experience of living on seemingly unmoving ground, “above” which the Sun travels across the sky. In this sense, the Sun does indeed “go around” the Earth. Given the existence which we live, I wonder whether this is a very good question to use for this purpose.

Please do not let my comments keep you from further such explorations.

Something to say?