On the left, odds ratios and confidence intervals from a logistic model predicting autism in the children of mothers exposed to environmental hazards On the right, the headline based on the study This is a bit nutspic.twitter.com/SF93TYw6UU
You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
Can't blame journalists here. This isn't news. And yet, this from the press release Nowhere does it mention that a) most exposures were found to be fine or b) these relationships were not statistically significant anywaypic.twitter.com/ObLqvohQdy
Based on the results of the study itself, the press release should read "no relationship between exposure to industrial pollutants and autism found" but that's much less grabbing I guess ping @natalier78
Other possible headlines that this study would support: "most industrial chemicals totally unrelated to autism" "fathers' relief: study finds no relationship between men's chemical exposure and their children's mental health"
Podcast out on this now, have a listen:https://soundcloud.com/senscipod/episode-11-solvents-and-autism …
Obviously the whole thing is crazy, but this yet again highlights the silliness of p<0.05 as the one criterion used to evaluate significance. Whether you're just above 0.05 or just below is really down to chance, but even many researchers treat it as some magical theshold.
I wish I would get a dollar for every paper I read that says something along those lines, and another 50 cents for having it followed by "but with a larger sample size...".
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.