LORD OF THE APES
WANDERINGS THROUGH THE WORLD OF PRIMATES

Wednesday, March 3

Breast milk makes my day

I came home from the field today to find this story blinking at me from my feeds. I was quite excited.

Baby monkeys receive signals through their mother's breast milk that affect behavior and temperament

ScienceDaily (2010-03-03) -- Among rhesus macaque monkeys, mothers who weigh more and have had previous pregnancies produce more and better breast milk for their babies than mothers who weigh less and are less experienced. Scientists are using this natural variation in breast milk quality and quantity to show that a mother's milk sends a reliable signal to infants about their environment. This signal may program the infant's behavior and temperament. ... > read full article

Sweet! Now I don't need to write a Human Ethology article on why you shouldn't get breast augmentation. Though I am not sure if that is a good or bad turn of events.

4 comments:

  1. Is there a causal link between that and my comment?

    Seriously though, these authors identified four factors. A (mother size) and B (birth order) are correlated with C (nutritional content of milk). In light of temporal and other factors, their inference that A and B cause C is reasonable. C is further correlated with D (later juvenile behavior), and they infer that C causes D. This is may be right, but I didn't see any reasons not to believe it might be the another confounder (of which we already know there are at least two: A and B).

    ReplyDelete
  2. Did you read the paper? Cause I didn't. So as far as I am aware, yeah you could be right. But they can's be as bad as Satoshi Kanazawa, the winner of this week's "How to Lie with Statistics" award.

    ReplyDelete
  3. I did read the paper, and the authors aren't so stupid as to ignore the confounding variables. However, they do dismiss them as unlikely, based on either intuition or some rather half-assed statistical analyses. The regressions in the paper do not seem to be very carefully or insightfully tested, and there were several tests that could have cleared up the ambiguity that were not done. That said, however, I do think that their conclusion is most likely correct, even if it is only modestly supported by the data.

    ReplyDelete