Slashdot Mirror
Doubts Raised About Cellphone Cancer Study (vox.com)
Vox is strongly criticizing coverage of a supposed link between cellphones and cancer suggested by a new study, calling it "a breathtaking example of irresponsible science hype." An anonymous reader writes:
A professor and research monitoring administrator at an American medical school reported that to get their results, the researchers "exposed pregnant rats to whole body CDMA- and GSM-modulated radiofrequency radiation, for 9 hours a day, 7 days a week," and the results were seen only with CDMA (but not GSM-modulated) radiofrequency. "[F]alse positives are very likely. The cancer difference was only seen in females, not males. The incidence of brain cancer in the exposed groups was well within the historical range. There's no clear dose response..."
An emeritus professor of applied statistics at the Open University in Britain also called the study "statistically underpowered..." according to Vox. "Not enough animals were used to allow the researchers to have a good chance of detecting a risk from radiofrequency radiation of the size one might plausibly expect."
An emeritus professor of applied statistics at the Open University in Britain also called the study "statistically underpowered..." according to Vox. "Not enough animals were used to allow the researchers to have a good chance of detecting a risk from radiofrequency radiation of the size one might plausibly expect."
> OK, I just realized that there are some of you who may believe that I'm overdoing my criticism of vox.com, so I'm going to post a story from their motherfucking front page today
That makes them better than Mother Jones (reporter of the original story) who has a "meme of the day" then, no?
http://www.motherjones.com/kev...
I'm curious as to how RF causes cancer only in *male* rats and why they live longer anyhow, or why the middle exposure group tended to have zero rats with cancer, rather than the low exposure groups, which had about as much cancer as the high exposure groups. Frankly, from reading the data reported in the study (you all did that... right?), all I could help think there's no clear pattern here. It really looks like the noise is larger than the effects, given all the 'anomalous' zero samples assuming the hypothesis that it really does cause cancer.
> I'm saying exposure to Vox causes cancer.
Even stopped clocks are right twice a day. I think the complaints about this study look legit here. I don't read Vox regularly and have no stake in arguing whether they're good or bad in general.
I didn't. :) But just reading the first dozen pages...
It looks like they broke the test rats into groups with 1.5, 3, and 6 W/kg exposure, CDMA and GSM, male and female. That's 3*2*2 = 12 groups. For the brain section, they looked for two types of tumors. So now they've got 24 groupings that they're searching for possible correlations.
The statistical significance of the one correlation they found (male, CDMA, 6 W/kg, malignant glioma) was p < 0.05. In other words, due to their limited sample size, just by random chance alone you'd expect such a blip to occur about 1 in 20 times even when there is no real correlation. Well they tried 24 times and got one blip.
Same thing with the heart results. 24 groupings, one blip with p < 0.05, one blip with p = 0.052. Again, almost exactly what you'd expect by pure chance alone.