Compare and Contrast
Here are some edited highlights of a BBC news story published yesterday, about some medical research that only looked at female subjects:
Women... Women... Women... women... women... The women... the study did not include men, the authors believe the findings may not apply to men. The women... men who drink might not do this... the way men and women break down alcohol in their liver may differ, which help might explain why women... you... people should be wary... women... women.
Here are some edited highlights of a BBC news story published today, about some medical research that only looked at male subjects:
you... you... your body... us... we... we... our mood... we... People... we... volunteers... people... volunteers... volunteers... patients... patients.
Women... Women... Women... women... women... The women... the study did not include men, the authors believe the findings may not apply to men. The women... men who drink might not do this... the way men and women break down alcohol in their liver may differ, which help might explain why women... you... people should be wary... women... women.
Here are some edited highlights of a BBC news story published today, about some medical research that only looked at male subjects:
you... you... your body... us... we... we... our mood... we... People... we... volunteers... people... volunteers... volunteers... patients... patients.
posted by MissPrism at 12:06 pm
18 Comments:
Wow. Very revealing, eh? Well spotted.
Classic. But then most research is done on males because our hormones mess up the results.
Makes biases pretty clear, doesn't it?
It also reminds me of blackout poetry: http://www.blackoutpoetry.com/
But then most research is done on males because our hormones mess up the results.
Mess up the results for whom? Real people, who don't have hormones?
Are you suggesting that science doesn't stand as an objective beacon free from patriarchal imperatives?
I really believed them when they said it did.
@ wriggles, in solidarity -
I didn't.
And we wonder why Viagra is covered by insurance when the morning-after pill (and even BC) is sometimes not, and why heart attacks in women are not diagnosed until too late ...
/mini-rant
Contraception including the morning after pill *is* free (well in the UK anyway)
Ah, science, you bias-free wonderland, where the messed-up version of results would be the one that applies to more than half the population.
Science may or may not be sexist and biased (well, science itself definitely isn't; scientists probably are, because the're people). The problem in this story, though, is not science. The problem is the media's reporting of science (which is very often a problem for very many reasons).
TRiG.
TRiG. Who developed the scientific method? God? People did. Thus science itself could be biased.
Sita: well, the principle of science is "predict it. then try if it works. have everyone watch you and flame wherever they can." Granted, there are lots of ways to do it wrong (see: people), and there are some things that can't be tested (in which case science will go "whatev. not my domain"), but there is no way that that basic principle can be gender-biased, unless women are magically exempt from causality or something.
Now, the scientific establishment, or even worse, science journalists, THEY are sexist and biased like whoa. Which means they constantly get it horribly wrong. But that is not the fault of the scientific method, but of doing it wrong.
I incline more to TRiG and Attie's opinion than to Sita's here (although she raises a philosophical point I'm not qualified to answer: maybe try asking Janet Stemwedel?).
However, it's important to remember that even if we grant that Science Is Pure, people and their biases are involved at every step of the process except the cold hard stats themselves. That includes deciding which questions and approaches are interesting or fundable, as well as how to present and interpret the results.
But the first study was specifically about the effects of alcohol on women. The second was a study on men that drew general conclusions for the whole population.
From the first
"Conclusion: Compared with nondrinkers, initially normal-weight women who consumed a light to moderate amount of alcohol gained less weight and had a lower risk of becoming overweight and/or obese during 12.9 years of follow-up."
And the Second
"Conclusions:
Inflammation-associated mood deterioration is reflected in changes in sACC activity and functional connectivity during evoked responses to emotional stimuli. Peripheral cytokines modulate this mood-dependent sACC connectivity, suggesting a common pathophysiological basis for major depressive disorder and sickness-associated mood change and depression"
I don't see anything wrong with the BBC article.
@Phil, the problem is that they're treating the first study, which only looked at women, as representative of the female population (fine) but treating the secondstudy, which only looked at men, as representative of the whole population, which is less likely to be valid. Just because they drew those conclusions in the paper's discussion doesn't mean the conclusions are valid, and ignoring the biochemical differences between their sample group and the majority of the human population means the conclusions are flawed at best.
@Wurzel, you're right but that's fits the proposition that the problem is with medical research conventions rather than with the reporting (in this case).
But overgeneralization from tests on highly constrained populations is a big issue - men v women is the beginning of the story, but very many studies iirc extrapolate to the entire population despite being actually limited not just to men but overwhelmingly to young, reklatuively fit male students from predominantly middle-class backgrounds and a very limited mix of racial phenotypes. And so on.
The second study also only looked at a sample of 16 people. How valid it is to extrapolate that sample to the general population is a fair question, and not one I'm qualified to answer. I'd guess there is a lot of literature discussing this out there. Experimentation is a series of compromises.
My real problem is this post is about a gender bias in the BBC which is completely unfair, and pretty misleading. The first study drew conclusions about Women so the article talked about women. The second drew conclusions about everybody so the article talked about "us".
A lot of research into new drugs and the like is done in men on the "do no harm" principle, on the basis that men are less likely to be pregnant.
Of course, for a non-interventionist study, where you just measure the effects of what people are already doing, you can study any segment of the population.
Note the difference between "16 healthy male volunteers received typhoid vaccination or saline (placebo) injection in two experimental sessions" (interventionist study: first trial in healthy male volunteers) and "Alcoholic beverage consumption was reported on a baseline questionnaire" (non-interventionist cohort study). The bias you talk about certainly exists, but the particular examples you chose to illustrate it do have other explanations.
TRiG.
"But the first study was specifically about the effects of alcohol on women. The second was a study on men that drew general conclusions for the whole population."
That you don't see the problem in these statements is precisely the problem.
Post a Comment
<< Home