Tuesday, January 31, 2006

9 Out of 10 Doctors Hate Participating In Such Surveys


Every commercial or endorsement for health products/medicine seems to carry with it a hearty recommendation from "doctors"--specifically 9 of them.

"9 out of 10 dentists recommend flossing your teeth."

Well what does the 10th guy recommend? Flossing your teeth with a candy cane? Who is this quack? And more importantly, I should point out that this 10th doctor is actually the crux of the whole recommendation system. Without him/her the other 9 really have no bearing on the outcome of a medicinal purchase. It takes that one dissenting voice amid the chaos of the drugstore to rise above the fray and give us something to consider, if only for a moment.

"Wow, 9 out of 10 dentists recommend toothpaste! Hm, but there's one who thinks it may not be necessary..."

Of course, no product could ever claim "ALL DOCTORS RECOMMEND THIS PRODUCT" as if the decision were handed down to them by God on a stone Tagamet HB, er tablet. Who knows how far these statistical gymnastics can go! It may very well be the case that 9 out of 1200 doctors prefer Crest over Colgate, but Crest simply says "9 out of 10," a technically true statement, but a misleading one at that. I think this example is used in intro statistics classes to warn students about the dangers of numbers in general. I think numbers are generally involved in more crimes than words; at least in free societies. How did Al Capone finally get reeled in by the FBI? Tax evasion. Numbers.

Thus, I propose that we eliminate numbers from endorsements and say something else along the lines of "Given a handful of doctors, a good many of them will likely support your purchase of Crest for the purposes of cavity prevention. We're pretty sure about that. Fairly certain. kthnxbye."