3 ms·

I'm not a scientist, but doesn't a large enough sample-size greatly limit the chance of this, statistically?

• pekk 13y ago
  Making a sample larger cannot correct for nonrandom biases in who you sampled.
  • mjb 13y ago
    "We then randomly assigned 20 of them to take part in weekly meditation classes" It makes sense to be critical of bad experiment design, and skeptical about extraordinary results. On the other hand, claiming bad experiment design based on the amount of data presented in the article seems premature. Maybe wait for the full paper and criticize the methodology if it appears flawed.
    • jessriedel 13y ago
      > Maybe wait for the full paper and criticize the methodology if it appears flawed. Eh, if the authors and their university are going to promote this in the media--and, make no mistake, this has been promoted heavily in order for it to reach the NYTimes--then I think it's fair game. You don't get to publicize your claims on the national stage and then avoid criticism until after your paper is published and the reader has forgotten everything (except the vague impression that meditation is backed by science).
  • jessriedel 13y ago
    That's not what seren was suggesting: > it could have been random luck that people with more empathy were selected to follow the meditation training.