shallowdeep -> RE: Dominant women more likely to have sons (10/14/2010 3:55:48 AM)
|
Assorted thoughts: quote:
ORIGINAL: VaguelyCurious You know that all your experiences are statistically irrelevant, right? I can't imagine why you would feel responses on this forum aren't going to be a perfectly valid statistical sample. It's not like those sorts of things should be done carefully, or anything, is it? :) quote:
ORIGINAL: Missokyst Statistical data is easy to manipulate if you choose your subjects well. To a point, but with a study likely to survive peer review you aren't free to just selectively choose subjects. As quotable as Mark Twain is, statistics are on pretty firm footing. Stochastic processes inherently have some uncertainty, but the whole point of statistics is to be able to quantify that. Such efforts are, admittedly, frequently misunderstood. quote:
ORIGINAL: VaguelyCurious She's not a nutjob, she's been published in well-respected, peer-reviewed papers, although I didn't see the research mentioned in the article (but JSTOR may well just not have a subscription to whatever it's published in and the Times sucks at citation). ANDquote:
ORIGINAL: tazzygirl I never said she wasn't respected. I asked why we weren't reading the study and results instead of a fluff piece in an online newspaper. Alright, I got curious enough to dig up the actual article from the journal Reproduction rather than researching the topic I should have been. Feel free to read the linked PDF if you want: Can mammalian mothers influence the sex of their offspring peri-conceptually? (DOI: 10.1530/REP-10-0137) (Just FYI, PubMed is usually a much better resource than JSTOR for recent medical research.) After skimming it some, I'm not particularly convinced. Grant's basic argument seems to be, "I think it would have been evolutionarily advantageous if things worked this way, so that's the hypothesis." The actual evidence in support isn't very compelling. The correlation between follicular testosterone levels and embryo gender in cows seems to have been on the fringe of statistical significance… and she dismisses a larger study that flatly contradicted her findings. That study, "Can Bovine In Vitro-Matured Oocytes Selectively Process X- or Y-Sorted Sperm Differentially?" was conducted by people with a background in biology rather than psychology and concluded, "Bovine oocytes, at least under the conditions used in this study, cannot preferentially select X- or Y-bearing spermatozoa." In fairness, she did have a faintly plausible reason for doing so. There is never even an attempt to explain how testosterone levels would possibly be able to achieve the hypothesized effect, just the somewhat suspect correlation. Based on past publications, she's apparently been pursuing this pet theory for at least two decades; it comes off like she's trying to fit data to the theory rather than vice versa. It would be surprising if there weren't some biological factors that could influence sex ratios other than chance, but it doesn't really seem like she's even isolated a likely causal mechanism, let alone proved it. In fact, the actual journal article pretty much admits that: quote:
Unless such mechanisms can be convincingly demonstrated, no matter how much behavioural and theoretical evidence accumulates, the hypotheses will remain nothing more than hypotheses. On the other hand, having the ability to modify natural sex selection processes, if they exist, would provide advantages to both livestock industries and mammalian conservation programmes, both good reasons to pursue these studies in addition to their academic interest. I think she wants more funding. :) Something I also feel worth pointing out is that, even if her hypothesis happens to be correct, I'm not sure follicular testosterone levels necessarily determine dominance in people… Edit: I just noticed the newspaper article linked in the OP is from 2008, not recently. The journal article I linked was published last month... but it makes reference to Grant's 2008 paper as well as to a couple that have been published by others since then. The other paper I linked contradicting the findings was a direct response to Grant's 2008 paper. It seems she may have backed off her original position some since then, now admitting the idea is "controversial" right in the first sentence of her 2010 paper.
|
|
|
|