This was a fantastic collaboration with lots of people including @hilarycmartin.bsky.social @jakobgrove.bsky.social, Experts by Experience, and several others who I can't seem to find on this app.

Article: www.nature.com/articles/s41...

And a very helpful and informative commentary on the paper by @tuckerdrob.bsky.social www.nature.com/articles/d41...

Keynote speakers: @tuckerdrob.bsky.social @foswald.bsky.social Ellen Hamaker

Hands-on workshops by:
@tedmond.bsky.social & @ukuvainik.bsky.social (genomic analyses)
@dirkwulff.bsky.social (LLMs in personality research)

www.ecp22edinburgh.org/programme

Let’s not “memory hole” the Flynn Effect, whereby objective cognitive measures show large mean cohort trends that are not plausibly reflective of any sort of general aspect of intelligence, but the individual correlates are largely invariant across generations.

But what is messy? What is late? Still subject to frame of reference.

Perhaps not peer comparison but shifting frame-of-reference, concept of normative or ideal, use of language… what @michelnivard.bsky.social likens to a vowel shift. Not necessarily the case, but important to consider, rather than trusting mean trends in self-report as veridical without question.

The interpretation is that the charter schools increase conscientiousness, self control, & grit... but also make students more self-critical with respect to those skills (perhaps they are cured of a Dunning-Krueger effect), such that self reports of those skills go down while the skills improve. 5/5

One of the coolest examples of this is a natural experiment (lottery study) for oversubscribed charter schools (doi.org/10.3102/0162...). The charter schools increased student achievement but reduced self-reported noncognitive skills that are known to relate to achievement. 4/

they may be misleading when comparing across cultures, cohorts, or possibly developmental periods due to different ways that people interpret the questions and use the likert scales, reference bias, etc... 3/

Unfortunately, "objective" observation- and task-based methods for measuring personality have much weaker construct validity and criterion validity. While self-reports are a great way to measure individual differences in personality among people matched on culture and context... 2/

The twin/family studies were not clear whether it was dominance vs. epistasis vs. some other sort of nonadditivity. Dominance was just the easiest to model. We can now estimate dominance SNP h2 but can’t obtain an estimate of epistatic SNP h2. Assimilation/contrast effects may also be at play.

To close the loop on this, this is what we wrote in the review (originally posted on the other platform).

O, right I promissed JUICE! Not everyone believes in this work, one of the "haters" was very early... In fact @ent3c.bsky.social practically leant over this papers' crib and said: "well that's not going to be a looker."... Someone just as petty as me, but far more organized, kept the receipts 👇