What is a Q-Q plot and how can it help test for common distributions?

[Tosh5457]

I've read the article in wikipedia but I can't understand it. Can someone point me to a good explanation of what a Q-Q plot shows and how to make one, or explain to me please?

 

[mathman]

I've read the article in wikipedia but I can't understand it. Can someone point me to a good explanation of what a Q-Q plot shows and how to make one, or explain to me please?

You should supply a link to the article.

 

[Tosh5457]

You should supply a link to the article.

http://en.wikipedia.org/wiki/Q–Q_plot

 

[pbuk]

Oops, answered the wrong question

 

[MarneMath]

The q-q plot is a common way to test if two sets of data come from the same distrbution. Typically, you'll plot a 45 degree line and with the two quantiles versus each other and if they come from the same distribution, then they should line up with the 45 degree line. The more the points depart from the line the more evidence you have that they are not common distributions. A q-q plot has the advantage that you can test two sample sizes of unequal size against each other, and at the same time test some distributional aspects (ie shifts, symmetry, outliners.)

For as how to make one, I suppose the first step would be to gather estimated quantiles from the two data sets.