This preview shows page 1. Sign up to view the full content.
Unformatted text preview: inear correla7ons: Large/Strong: dots call close to a straight line Small/Weak: dots fall far from a straight line (harder to see a “pa`ern” 6 11/13/13 h`p://www.youtube.com/watch?v=VFjaBh12C6s Compu7ng a correla1on coeﬃcient: a sta7s7c that gives you the exact correla7on in terms of its direc2on and strength We want to ﬁgure whether highs on one variable go with highs on another… thus we need a way to compare variables But how do we do this when we have two variables that are measured on diﬀerence scales? 7 11/13/13 Z scores! Z score = number of standard devia7ons that a score is from the mean Conver7ng to Z scores is a way to “standardize” A raw score that is high (above the mean) will always have a posi7ve Z score A raw score that is low (below the mean) will always have a nega7ve Z score Z scores tell you how far above or below the mean a par7cular score was In compu7ng a correla7on, we calculate the cross
product of Z scores Cross
product = mul7plying scores on one variable by scores on another What happens in the case of a posi1ve correla1on? ▪ If you mul7ply a high Z score (+) by a high Z score (+), you get a posi7ve cross
product(+) ▪ If you mul7ple a low Z score (
) by a low Z score (
), you get a posi7ve cross
product (+) ▪ If you add up these cross
products, you get a large posi7ve number (+) 8 11/13/13 In compu7ng a correla7on, we calc...
View
Full
Document
This note was uploaded on 03/24/2014 for the course PSY 21201 taught by Professor Bernard during the Winter '13 term at SUNY Stony Brook.
 Winter '13
 bernard
 Psychology

Click to edit the document details