Are you saying that we should ignore all non-experimental research? I'm 
concerned you're giving this message to your students! This is the 
justification the tobacco industry used to say that there is no evidence that 
smoking causes lung cancer in humans. Because we cannot produce human 
experimental results here, we have to rely on more complex correlational 
designs, theory, and animal models - all which convincingly support the idea 
that smoking causes lung cancer in humans. Since the amount of research in the 
small amounts of alcohol/breast cancer area is not even close to what we have 
regarding smoking and lung cancer, we have to adopt a wait and see and 
skeptical, but perhaps somewhat prudent attitude. This is precisely the message 
I give students. Don't ignore the research just because it is correlational; 
this is only a tad less naïve than just assuming all studies are created equal. 
As I stated previously, I don't know if this particular study attempted to 
control for other important variables. But again, ethically, what else CAN we 
do here besides correlational research?

While the BBC article certainly overstated much (as is typical with popular 
media), I think calling these types of studies "ridiculous" is very misleading, 
and in some instances, downright dangerous. Think of where we'd be with smoking 
research if everyone just disregarded it as ridiculous because of its, 
necessarily, correlational design?

Dean M. Amadio, Ph.D.
Assistant Professor
Siena College, Psychology Department
432 Roger Bacon Hall
515 Loudon Rd.
Loudonville, NY 12211
Phone: (518) 782-6768
Fax: (518) 782-6548


Amadio, Dean wrote:
> How else is one to study this issue? Much of health research is precisely the 
> same, due to obvious ethical concerns, including the research on alcohol and 
> heart disease.
Which is precisely why medical research is, in general, so lousy and its
(often overheated) conclusions keep changing from study to study,
resulting ultimately in lowered public respect for science at large.

What one should do is draw conclusions that are appropriate to the
evidence they are drawn from. If those conclusions are too bland to be
interesting, that doesn't justify falsely strengthening them to make
them more interesting. And that seems to be what has happened here.
(Note that "causes" is feature in the very first sentence of the article.)
> This is also old news (there are previous studies which find the same 
> relationship among women). I've been advising my students for several years, 
> especially females, to consider the cancer studies whenever they hear the 
> research that alcohol is heart healthy.
If those cancer studies are like this one, then I would recommend that
they ignore them, for they can tell them nothing about the effect on
their health of drinking moderate amounts of alcohol. They only tell us
that the global genetics and lifestyles of people who choose to abstain
from alcohol altogether do not result in cancer quite as much as those
of people who do not adhere to such a prohibition.

Also, they use global percentages in their presentation of risk, which
almost inevitably misleads people about the actual  increase in risk of
low base-rate conditions like cancer. (See, e.g., the recent writings of
Gerd Gigerenzer). For instance, the article says that 5,000 of the
45,000 annual cases of breast cancer are due to alcohol -- an increase
of 11% they say. The population of the UK is about 60 million. Half of
the those are female -- 30 million. About 20% of those are children --
leaving 24 million. (see
http://www.statistics.gov.uk/cci/nugget.asp?ID=6). 45,000 out of 24
million = .0019:  19 in ten thousand women are diagnosed with breast
cancer in any given year. Even if the alcohol-cancer causal link were,
in fact true, the number of cancer cases would drop to 40,000 which,
against a vulnerable population of 24 million is .0017: 17 in ten
thousand. Now ask yourself the question: Would you change you lifestyle
dramatically to reduce a risk by 2 in 10,000? And that's if the causal
link had been established, which it hasn't been.

Regards,
Chris

---
To make changes to your subscription contact:

Bill Southerly ([email protected])

Reply via email to