4 ms·

This article is going to be flagged into oblivion into a few hours. We're not discussing the relationship between biological and psychological traits on HN.

• bildung 9y ago
  Alternative explanation: this article is just not good science. It: * doesn't control whether the not-first-borns are outside the target demographic (too old or too young) * doesn't look at children without siblings * more importantly: is just a cross section, so no causal relationships can be gleaned from that * also highly problematic: is self-selecting, with no discussion of the distribution of people not participating. With a poll question like that that alone invalidates every result, as obviously it invites bias towards readers who deem themselves smarter than their siblings * is not discussing, and controlling for, alternative explanations for the observed result All in all the margin of error is almost certainly a multiple of the effect size, thus rendering the results useless. Which they are anyway, because you can only answer a question like the one given in the article with longitudinal studies.
  • evanpw 9y ago
    Since birth order is immutable, what would you gain by a longitudinal study? You still only get one data point per person.
    • bildung 9y ago
      Sure, birth order is immutable, but you would gain multiple data points for "intellectual curiosity", which would be the point of the exercise.