What did this study actually find? Virtually NOTHING AT ALL
-
-
Show this thread
-
Here's a link to the study https://ehp.niehs.nih.gov/EHP1837/#c15 If you scroll on down to the results section...pic.twitter.com/zHST9JdudR
Show this thread -
"What's that Gid?" "Oh, not much. Just ~40 non-corrected statistical tests of which 3 were significant" "Wow! That really doesn't sound like much, does it?"pic.twitter.com/AFviAernwN
Show this thread -
Here they are, highlighted. I should also mention that most of these were ~barely~ significantpic.twitter.com/AhFNUDyxbZ
Show this thread -
Let's zoom in on one number that is in EVERY GODDAM ARTICLE The "1.5 times higher risk of breast cancer and twofold higher risk of prostate cancer" Total. Utter. Nonsense.pic.twitter.com/qH8WOonrFz
Show this thread -
To do this, we have to quickly go over confidence intervals
Show this thread -
When we compute a confidence interval, we are basically saying that there's a decent chance that the true value lies between 2 points
Show this thread -
If the confidence interval crosses 1, it means that there's no statistical difference between the result and what we'd expect to find purely by chance If the confidence interval is very wide, it means that there is so much variance in the result that it might not be true anyway
Show this thread -
These are the tests behind that headline. Confidence intervals are in the little red squares Notice anything strange?pic.twitter.com/HU0a4dOFAE
Show this thread -
The first one is 1.00-2.17 That is ~technically~ significant, but really only just. This is likely to be a meaningless result
Show this thread -
The second is 1.38-3.03. This is a more robust result, but it's still bordering on insignificance given the sample size
Show this thread -
So the headline finding, the one that is being used to scare everyone about how terrifying light is... ...probably isn't true And it gets worse
Show this thread -
This is the conclusion. It's the closest you can get to saying "we found nothing at all" without actually coming out and saying itpic.twitter.com/sRD9NO3Pgb
Show this thread -
Note to journalists: when a scientists hedges this much in a conclusion, there's a good chance the findings were not significant at allpic.twitter.com/Uahhl1pLvS
Show this thread
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.