Cell Phone Use Linked To Increased Cancer Risk Information Week. But the absolute risk is not very high.
“Like giving a drink to an alcoholic” New York Times. Bloomberg on the stimulus plan.
Idle parenting means happy children Telegraph
Goldman Sachs’ Friedman: 80 Cents is ‘The New Par’ Peter Lattman, Deal Journal, Wall Street Journal
American pays $50,000 to clone dead dog Financial Times. And a pit bull at that.








Note that the putative increase in risk is only 50% on a very rare disease, and that “… a Swedish salivary gland study, “Mobile Phone Use and Risk of Parotid Gland Tumor,” … found no increased risk of tumors caused by cell phone use.”
About 15 years ago I analyzed the underlying numbers in a study that claimed to show an increased risk of leukemia among electric power company employees with higher than median exposure to power line EMFs. What I found was that the number of leukemia cases among the higher-exposure group was less than you would expect from the average incidence of the disease in the general population. In fact, the claimed increase in risk was entirely due to the fraction of the “control” group that happened to fall into the lower-than-median-exposure class — but the degree to which that was so was entirely within the range one could expect to see by chance.
After a lot of spreadsheeting and Monte Carlo simulation runs, I concluded that unless the apparent increase in risk is more than 100%, the results of any such cancer study have to be treated with great skepticism. This is especially so when other studies — such as the Swedish study quoted — have not found similar effects.
Usually these studies state that the variation they have seen is “statistically significant at the 95% confidence level”. Note that all that apparently impressive claim really means is that you wouldn’t expect to see the result by pure chance more than once in 20 studies.
But those are about the same as the odds against getting two pair on the draw in 5-card draw poker.
Note that there are at least 30 types of cancer that can be covered in these studies (that was the number in the power line EMF study). So the odds of getting a result that appears “significant at the 95% confidence level” is very high.
To teach my kids about what such stats really mean, I had them take a poker deck and deal 30 5-card poker hands. The result? Two hands with two pair, and one three-of-a-kind!
Every study looking for evidence of increased cancer risk among cellphone users is like the dealing of yet another poker hand. Deal enough hands and eventually you’ll get three-of-a-kind.