Worst Cherry-Picking and Math… EVER
October 19th, 2009
|
| Share |
If you happen to be a statistician, don’t read any further. This very well might kill you…
Just a month after Environmental Working Group released its online consumer guide to cell phone radiation, new studies suggest that cell phones are indeed a gadget to be concerned about — though how alarmed we should be still remains unclear.
The Los Angeles Times reports that when data from 23 different studies were pooled together, no link was seen between cell phone use and brain tumors. However, when data from the eight strongest studies from the 23 were singled out, “cell phone users were shown to have a 10 percent to 30 percent increased risk of tumors compared with people who rarely or never used the phones.”That said, seven of those eight studies were conducted by a single Swedish researcher. And the data could be affected by the fact that many people in Sweden live in more rural areas, where more radio-frequency energy is generally required for cell phone use.
What can we take away from these findings? We need more and better studies that aren’t funded by the cell phone industry, according to the researchers. In the meantime, use Environmental Working Group’s online consumer guide to cell phone radiation to find out how the radiation level of your phone compares to that other phones — and what you can do to reduce cell phone radiation exposure.
Notice a in error in the logic here? Does anyone not notice an error in the logic?
Yes, that’s right, the data looks really bad if you choose the worst data you can find. The single Swedish researcher is Lennart Hardell, who is more of an assclown than a researcher.
The data is not affected by the fact that many in Sweden live in rural areas. Plenty of other nations have a high rural distribution and there are plenty of non-assclown researchers there and in Sweden.
In a related story, it has been shown in a study that cell phones cause cancer in 100% of users. To demonstrate this, researchers looked at cell phone users and examined their medical histories to see if they had brain cancer. If they didn’t, they were omitted from the study. After removing all the non-cancer cell phone users, scientists were shocked to discover that 100% of their subjects developed brain cancer.
(Last paragraph is sarcasm)
This entry was posted on Monday, October 19th, 2009 at 4:43 am and is filed under Bad Science, Just LAME, Not Even Wrong, Obfuscation, inverse square. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.
View blog reactions
October 19th, 2009 at 5:14 am
IT BURNS! IT BURNS!
Quote Comment
October 19th, 2009 at 7:00 am
I used to call this the Sternglass method, after Ernest Sternglass, a prominent exponent of it (as applied to nuclear power) back in the 1970s or thereabouts.
Quote Comment
October 19th, 2009 at 8:32 am
Mail the article author. I did. Didn’t get much of a reply though.
Bottom post at this page:
http://www.vof.se/forum/viewtopic.php?p=306554#p306554
/Michael
Quote Comment
October 19th, 2009 at 11:29 am
The Monster from Polaris said:
Yeah, Sternglass is still around and still spouting the same bull
Quote Comment
October 19th, 2009 at 12:47 pm
This sort of abuse in lateral studies is becoming more, and more common yet for the life of me I cannot find the theoretical justification for it. One would think a technique like this would have some grounding, and indeed these studies that use it (this is not the first one I have seen) treat the maneuver as if it is a widely understood norm.
I do understand not including results positive or negative that fall far outside experimental distribution, as probably in error, but this goes far beyond that.
Quote Comment
October 19th, 2009 at 5:35 pm
How much is the rate of cancers reduced if the 6 best studies are used? Maybe we have a new therapy on our hands<:P
Quote Comment
October 20th, 2009 at 2:40 am
Hey, I just asked 23 people to flip a bunch of coins for me. When I looked at the combined results, heads came up as often as tails, but when the 8 best results from these 23 coin-flippers were singled out, it turns out that heads came up 10 percent to 30 percent more often.
Clearly, there’s something strange going on here.
Quote Comment
October 20th, 2009 at 5:10 am
Did anyone notice the bit where seven of the eight ’strongest’ studies were ‘conducted by a single Swedish researcher’ ? That tells a lot about said researcher, no ?
Quote Comment
October 20th, 2009 at 5:14 am
Willem said:
It does indeed. Lennart Hardell is infamous in this area.
/Michael
Quote Comment
October 20th, 2009 at 10:52 pm
Willem said:
Yeah, for some reason they consider his research to be the best quality despite the fact that he always seems to find connections between rf fields and (insert name of medical condition here) where nobody else seems to detect anything. He makes his whole damn living making ridiculous claims and going around to talk to idiots about the non-existent dangers.
I have made the decision to use a single word to refer to people like him: Those who claim to be legitimate scientists but who make a habbit of very dishonest and deceptive practices and who like to use their self-proclaimed authority to further bogus scaremongering. They often seem to have some kind of psycological problem that makes them need attention and to be made into a cult leader of sorts.
I call them assclowns, because that describes them quite well. They’d be funny if they weren’t such asses and they’re not worthy of any respect, like an assclown would be.
George Carlo, Andrew Wakefield and Ernist Sternglass are other well known assclowns.
Quote Comment