Massaging the numbers
52% PC users vs 25% MAC users : ratio 2.08
1) 36% PC users are liberal, 58% MAC users are liberal
So, if we take a amount of 100 people, there will be 18.72 liberal owner PC vs 14.5 liberal MAC users ( 36% of 52% vs 58% of 25% )... so, on a sample of 100 people, there will be more PC owner who are liberal that MAC users...
2) 54% of PC owner have a college degree or higher, 67% for MAC users... same method that before, with a sample of 100 people, 28 pc users will have a college degree or higher vs 16.75 mac users...
Can continue for everything... and i will reach a fully reverse conclusion that these in the article... simply because i have not use the PC and MAC group like two independent study but use it like a single study... since we have around two time more PC owner that MAC owner in these study, the light variation in the benefice of MAC group are in fact showing that you have less Liberal or high educated people in the MAC group...
Are my number wrong? No... on 100 people ( computer user ) who have reply to the survey, there was really 28 pc users with a college degree vs only 16.75 for MAC... the link don't show the statistical analyze but only one of the numerous possible interpretation...
A last example : 42% of PC user would opt for The New York Times over USA Today, 69% of the MAC user will opt for the New York Times...
At first view, it seem that MAC users are the main audience of The New York Times... but again, the way stat are show make them lies...
on 100 people, 52% are PC users... from these 52 PC users, 42% will opt for the NY Times... it is 21.84 potential customer for the NY Times who are PC users...
on 100 people, 25% are MAC users... from these 25 MAC users, 69% will opt for the NY Times... it is 17.25 potential customer for the NY Times who are MAC users...
So, on a sample of 100 peoples, you will have more PC users who will opt for the NY Times that MAC users...
Raw numbers don't lie but be careful about interpretation of numbers...
Hehe...nice spot Thoumsin.
"Figures never lie...but liars always use figures."
First thing you learn in sales and marketing and its a perfect lesson in how politicians work all the time.
This is one of those tricks where you use big numbers to scare people but in reality they are not as big as they make them seem. Nice catch. I feel smarter now.
Well, both interpretation are correct... the article about the survey analyze it by individual group when i analyze it on the base of all computer user... if some day, MAC users become equaly numerous that PC users, conclusion from the survey will become true...
My last example, related to the newspaper will be similar to a analyze that will make the newspaper for define his target audience... yes, there is more people in the MAC group who choose for the NYT... but when related to the general computer users population, since the MAC group is a minority group, it have a low influence on the general statistic...
A good survey need to include the raw numbers, the question asked and the interpretation...
from http://www.w3schools.com/browsers/browsers_os.asp
"There are three kinds of lies: lies, damn lies, and statistics."
Benjamin Disraeli
"Then there was the man who drowned crossing a stream with an average depth of six inches."
W. I. E. Gates
Welcome Guest! Please take the time to register with us.
Sins Forums
JoeUser Forums