Been working on some bigger genetic projects to be published soon. In the meantime, ask me some good questions and I will try to answer them with data. Also, this year’s reader survey is here. This year it’s running on the taketest site, so it automatically links to your test results from there (if you have any, which you should), so we don’t have to keep asking people to go to some other website to take surveys/tests. It will also automatically generate a mini-report of the results. This also links people across years (if we do more surveys, which we probably will). If you don’t want to be tracked this way, just take the survey in a private tab.
Other recent posts you may have missed:
- Worldwide results for migrant outcomes, always relevant, but this time it has a comprehensive worldwide database, which we can easily expand. Post any new results you find that I don’t have so we can add them.
- Men vs. women intelligence is a perennial topic. Most data unfortunately come from children because representative samples are easier to get (via forced labor in school à la PISA). However, there are some large adult databases too. They too generally show a small male advantage, and not even a female advantage for verbal skills (depends on exact test presumably).
- People love to speculate about interactions. Maybe extroversion causes crime for men but not women; maybe openness causes leftism more in women than men etc. Statistically speaking, though, such interactions are generally very small. In other words, your Bayesian prior on the effect size of any such proposed interaction is a tight distribution around 0. With large datasets one can find any number of interactions, but with effect sizes so small as to be practically irrelevant. The crud factor also applies to higher-order statistics. Everything is related to everything else, even if only very slightly; and such relationships always differ slightly by some 3rd variable, but even less so.
