The Correlation Coefficient - Explained in Three Steps 

wow they taught you that for one semester? we did this in a week and had a quiz and that's it e.e

what grade are you in by the way...

Cracked Code same

nu

What's your course? Stats or econ

meeeeeeee

i wonder why he stopped

Daniel Kirchheiner I know right!, other people would just use a mouse on a computer and confuse the shit out of u.

and my guess is, by "scatter along y-axis", you were referring to how far away are those dots from y-axis. And by "scatter around fitted line", you meant the difference between the predicted and actual y value of those dots.

D Boy, he calculated it before the video either with an equation or technology, this allowed him to show how close they actually are to 0

Ha! Errrm, a mix of felt pens that I had flying around at home... mostly Stabilo? Sorry, I'm not being helpful here!

jior6 They are Black Stabilo point 88. The tip is about a 2-3mm it looks like.

Remember r = (n(Σxy) - (Σx)(Σy)) ÷ √((n(Σx²) - (Σx)²)(n(Σy²) - (Σy)²)) If you need me to help explain this I would be happy to.

those values were just used as examples, not necessarily the true calculations for the graphs he was showing. basically the closer the correlation coefficient (r) is to the number +/-1, the stronger the correlation of the variables. if (r) is less than +/-0.3, it means there is a weak correlation. if (r) is somewhere between +/-0.3 and +/-0.5, it means there is a moderate correlation. if (r) is greater than +/-0.5, it means there is a strong correlation. so if your correlation coefficient (r) is +0.92, then you have a STRONG POSITIVE correlation. if your (r) is -0.49, then you have a MODERATE NEGATIVE correlation. i hope that helps some. i'm just learning this too.

*those values were just used as examples* Then that should've been explained with more emphasis. So basically, the only useful part of the video is explaining the slopes. I have a graph with no numbers, so this doesn't help at all.

Me too ;-;

Ich stimme sehr zu!

You can definitely work it out exactly - I just wanted to explain the principle that underlies the numbers. There are several ways to calculate the correlation coefficient, which all lead to the same result. For the most common formula, have a look here: en.wikipedia.org/wiki/Pearson_product-moment_correlation_coefficient#For_a_sample

+Christina Saunders Paused the video and came down to see someone noticing it. Thanks. :D

Christina Saunders iii