Top TipsGot a great tip? Share it with us. You know, stuff like "Prevent sneezing by pressing you index finger firmly between your nose and your upper lip."
(
rob, Wed 29 Nov 2006, 16:33)
« Go Back
Are you peddling a bogus health supplement?Everyone's heard of the placebo effect, so you can't ignore that when explaining your results, but have you heard of confounding variables or regression artifacts? Neither has Joe Q. Public!
Say you do a study into something that's supposed to cure depression, that'll attract really depressed people having a problem with their condition right? Well, they'll probably be at a particularly low point when they come to you, and they might start feeling better independently over the course of the study, because that's how moods work. That's a regression artifact and you can ignore it freely!
Heck, there's a wealth of statistical and experimental problems that can skew results, and that you don't need to report, why not not bother looking them up yourself? Happy scamming!
(, Sun 2 May 2010, 16:30, 1 reply, 15 years ago)
Or you could use the techniques outlined in Darryl Huffs book:www.amazon.co.uk/How-Lie-Statistics-Penguin-Business/dp/0140136290/ref=sr_1_1?ie=UTF8&s=books&qid=1274440660&sr=8-1
An excellent read. Everyone should have a copy.
(
lion-cake Bad lion-cake, bad., Fri 21 May 2010, 12:20,
Reply)
« Go Back