We've all seen or heard commercials claiming that '4 out of 5 dentists recommend this tooth paste' or that '9 out of 10 women saw a more natural looking tan!'
Well by golly then it must be a good damn product if 80% of the dentists out there would use it! I have to have it!!
At the surface it probably seems as though its a pretty solid statistic. You know, the company did their research and is now promoting with the facts the found. Its a method used time and time again.
But here's another way to look at it:
They are lying, without actually lying. Companies promote a product in a way to make us assume that 4 out of 5 is 80%. But, they never say that. Why?
Let's pretend to take our own survey. We survey 100 people about a certain, oh lets go with, juice. Out of the 100, we'll pretend that 20 would buy that over anything else. Pretty sketchy results. A.k.a rather poor product.
But, in those 20 people who liked the juice, lets say 8 of them came in a row with the next 2 not liking it. Well, let's recap the last 10 people then. Oh my! 8 out of 10 would buy the juice over anything else!!
So, even though really 20% of the people liked it, out of a selected 10, 8 of them did. And that is not a lie, making it seem as though 80% of people like it. The more people you survey the higher chance of a probability that you get a high number of positive marks in a row and the better your product looks.
How much a year can I make figure easy stuff like this out? And to think that people need to have degrees in the business world.
Leave your feedback
Tucker
Subscribe to:
Post Comments (Atom)
3 comments:
I'll tell you before dad does. "How to Lie with Statistics". It's a book. I have it. I'll send it to you when I'm done.
Len is right. If 7 out of 10 pedestrians killed by automobiles last night while walking after dark were wearing dark clothes, then you might intuitively conclude and state that it is more dangerous to wear dark clothes when walking after dark. But you can't make that conclusion without knowing about the whole data set.
Perhaps there were a total of 100 pedestrians out walking last night and 90 of them had on dark colored clothes. In that case, if 3 of 10 or 30% of the pedestrians killed were wearing light colored clothes, you would have to conclude that it's more dangerous to wear light colored clothes after dark.
Statistics can be used to say or support anything! So keep questioning EVERYTHING!
;-) Thanks, Dad.
I'm sure Lenny remembers as well as I do that talk on a camping trip; by the 'Falling Rock' sign.
Post a Comment