I’m frequently asked how science curiosity, as we measure it, relates to education and to scores on one or another scale for measuring cognitive proficiency. For answering this question, I think the information in a graphic display of overlapping probability density distributions is superior to the information in a correlation coefficient.
All these differences are “statistically signicant” (what difference wouldn’t be at N = 3000!). But are they practically significant?
I can’t confidently say. They don’t look big to me, at least on the ≥ 90th percentile side.
But at this stage in our ongoing study of science curiosity, we don’t have enough information to say that disparities of this magnitude will result in noticeable differences in how people behave; all we can say is that the higher the SCS score one group’s members are, the less vulnerable to politically motivated reasoning they will be.