Post

"r" here is a "correlation coefficient" and is a measure of how *strongly associated* data are with a particular model Here the author is looking at linear correlation, so how close these are to a line ie, if one goes up, does other go up? or, how close are these to forming a line?

r is always between -1 and 1 If the data match a line perfectly, r=1 If they're mostly a line but 1 or 2 points are off, maybe it's r=.95 If the data are random blobby chaos, r=0. No line! If it's a perfect line but every time one goes up the other goes DOWN, r=-1

But wait, you ask, WHAT line? Ah! Here's where it gets interesting. We don't know (yet!) what line! That comes from regression analysis, which is often the next step