Do you have a better source for this data than, well, a bunch of numbers you posted to HN? In particular, an explanation by a researcher or other expert who has

Do you have a better source for this data than, well, a bunch of numbers you posted to HN? In particular, an explanation by a researcher or other expert who has actually studied the difference and can explain it? It's very easy to draw conclusions from statistics. It's much harder to draw correct conclusions from them.