Saturday, June 04, 2005
REPORT ON LEUKEMIA CLUSTERS NEAR ELECTRICITY LINES - CREATING A SCARE STORY
This morning I heard this reported as a "news" item. To be fair the BBC did actually quote the authors as saying that it was possible this was within statistical limits & could mean nothing. Then they spoilt it by getting some shrill woman who went on about how no politicians are "experts" on this, as if she was, & we should therefore "trust" anybody spreading fear.
Bull. For a full background read this on Spiked. Basically this whole case depends on "about 5" more cases out of "400-420" annually. Now on planet Earth random events don't happen at a perfectly even rate (that's why they are called random). Toss 400-420 coins (lets say 410 just to achieve an arithmetical base to make it more valid than this case) & you will be very unlikely to get exactly 205 heads (or more than 220). So to produce an excess of 5 would be meaningless even if they knew how many they should be dealing with.
This is supported by the fact that these cases have happened at up to 600m from the power lines at which point the magnetic field, the only credible cause, is non-existent. The "evidence" for the danger of passive smoking is of this order.
So the politically interesting questions are why was this non-proof published (when it is quite clear the authors are embarrassed by it) & even more, why did the media think this was actually a story. The former, I suspect, is that you have to publish something to earn your grant (& the extent to which you are under pressure to make your figures more interesting is obvious - not so much by making up figures as omitting them - had they also found a statistically equally insignificant fall in cases of flue would that have been reported - if not then if the statistically insignificant health increases had matched the statistically insignificant worse cases they would have been reporting wrongly to say they had found increased illness - this is what happened in the passive smoking case where the BMJ reported that 7 out of 40 cases had found increased cancers among passives - the corollary is clear).
And why did the media jump on this non-story. Partly, perhaps mainly, the relatively innocent story is that they have to produce news & "No Danger from Anything" doesn't cut it. However, at least within TV reporting there is clearly an agenda that everything that isn't compulsory ought to be illegal & anything involving technology, even a mature technology like this, involves incomprehensible magic. This is quite strange considering that broadcasting is a high tech enterprise but the people who run it aren't the technical staff. You never get to be promoted onscreen if you drive a camera - very much like officers & soldiers (mind the soldiers are pretty well paid & don't have to wear suits).
Maybe it would be nice to see one of them actually on air explaining something to the suits, Susan Watts on Newsnight shouldn't have to carry the whole can.
Bull. For a full background read this on Spiked. Basically this whole case depends on "about 5" more cases out of "400-420" annually. Now on planet Earth random events don't happen at a perfectly even rate (that's why they are called random). Toss 400-420 coins (lets say 410 just to achieve an arithmetical base to make it more valid than this case) & you will be very unlikely to get exactly 205 heads (or more than 220). So to produce an excess of 5 would be meaningless even if they knew how many they should be dealing with.
This is supported by the fact that these cases have happened at up to 600m from the power lines at which point the magnetic field, the only credible cause, is non-existent. The "evidence" for the danger of passive smoking is of this order.
So the politically interesting questions are why was this non-proof published (when it is quite clear the authors are embarrassed by it) & even more, why did the media think this was actually a story. The former, I suspect, is that you have to publish something to earn your grant (& the extent to which you are under pressure to make your figures more interesting is obvious - not so much by making up figures as omitting them - had they also found a statistically equally insignificant fall in cases of flue would that have been reported - if not then if the statistically insignificant health increases had matched the statistically insignificant worse cases they would have been reporting wrongly to say they had found increased illness - this is what happened in the passive smoking case where the BMJ reported that 7 out of 40 cases had found increased cancers among passives - the corollary is clear).
And why did the media jump on this non-story. Partly, perhaps mainly, the relatively innocent story is that they have to produce news & "No Danger from Anything" doesn't cut it. However, at least within TV reporting there is clearly an agenda that everything that isn't compulsory ought to be illegal & anything involving technology, even a mature technology like this, involves incomprehensible magic. This is quite strange considering that broadcasting is a high tech enterprise but the people who run it aren't the technical staff. You never get to be promoted onscreen if you drive a camera - very much like officers & soldiers (mind the soldiers are pretty well paid & don't have to wear suits).
Maybe it would be nice to see one of them actually on air explaining something to the suits, Susan Watts on Newsnight shouldn't have to carry the whole can.
