←back to thread

256 points hirundo | 3 comments | | HN request time: 0.001s | source
1. patrulek ◴[] No.35513856[source]
N?
replies(1): >>35515224 #
2. twobitshifter ◴[] No.35515224[source]
400,000 over 12 year period
replies(1): >>35515708 #
3. runarberg ◴[] No.35515708[source]
Not quite. The data was scraped from an open source database. They used Synthetic Aperture Personality Assessment Project (SAPA) which only gives the partial test to each participant, and than interpolates for the rest of the test.

The full set of items in SAPA can be as high as 500. They supposedly found the reverse Flynn effect by only looking at results from 35 (over a 12 year period) and another set of 60 (over a 7 years period). This should be a red flag, and raise suspicions of cherry-picking and p-hacking.

This is not to mention anything about the validity of interpreting the combination of some items from SAPA as intelligence. The SAPA authors don’t do that, neither did the researchers that originally collected the data, neither did the participants, that is only done by the researches of this particular paper. And to add another red flag, this paper was published in a pretty disgraced journal Intelligence, which has been proven to publish plenty of pseudo-scientific results where they stretch the statistics with an agenda aligned with the eugenics movement.