I have put the word out! DO NOT LAUGH if you see it get picked up by a TRUE BONE HEAD
"Study from a leading researcher in Massachusetts determines all e-cig users will die"
I mean, you might laugh, but that's basically exactly what this study did.
They used a selection criteria, and a classification criteria where the end-result was pre-ordained. I am hoping some high-school teacher will use this study as a test question for their basic statistics class.
Selection Criteria:
From the sample of terminal cancer patients, select those who smoke cigarettes at the start of the study.
Separate the sample into 2 groups, those who use e-cigarettes, and those who don't
Evaluation Criteria:
From the 2 sample groups, measure after 6 months, the ones who are no longer smoking.
Report but ignore the number of participants who dropped out during the study.
Do not attempt to measure whether the amount smoked changed
For those in the "smoking" group, measure whether or not they quit (without regard to whether they quit using e-cigarettes)
For those in the e-cigarette group, assume that the 60% of them that dropped out from the study all reverted to smoking. This is statistically fair, right? Because the 30% of smokers who dropped out were treated the same way.
Any high-school student taking a statistics class should be able to shoot this study full of holes. Doesn't need me.
Damn, I wish I was back in high school. Science-fair project dropped right in my lap! (Not to mention the cheerleaders.....)
