Things that are written about:
- In the US, boys outscore girls.
- Blue dots are left-shifted more than yellow and pink dots.
- The overall triangular shape of the distribution. In fact the whole y-axis is ignored despite something kind of interesting going on there - low scoring nations tend to have wider gender-related score disparities than high scoring nations.
- Why the colors? I mean, why are all of the Americas lumped together? Why are Western and Northern Europe separated from Southern and Eastern Europe? Why isn't the Middle East its own color? And did someone forget to invite Africa to the party? The article tries to make some vague East vs West points, but then it splits the West up?
- If the purpose is to relate nationality to test score and gender disparity, then maybe the dots should be colored by something more interesting. Maybe groupings based on how much is spent per capita on education? Maybe (given the article's thrust) lump those with similar cultural norms together, as scored by another independent metric?
- This is ubernitpicky, but it would be way more awesome if when you clicked a dot two vertical box-and-whiskers plots would pop up showing the actual distribution of scores for both boys and girls in that country, not just the average. Might there be more variability, again, in the lower performing nations? Would increased variability in within-gender score range (outlier effects) possibly correlate with increased between-gender disparity?
As my editor-in-chief puts it, "Why show the nuance of all the data if you're just going to sum it up with a simplistic takeaway message about the 'anti-science stigma' felt by girls in the US?"