R square

CFA Level II
#1

On page 172 the answer to 1, why R square = correlation squared? I only understand the formula R square = RSS/SST

#2

it is a definition that you need to know. On your Regression output from Software which is the ANOVA table – Multiple-R is Correlation, R-Squared = Correlation Squared.= Multiple-R Squared. and another formula you might need to know Adjusted R-squared = 1 - (1-R^2)[(n-1)/(n-k-1)]

#3

I recall reading that correlation squared=R squared only for linear regression, not multiple linear regression.

#4

that is correct…

#5

Thank you guys. Very clear. cpk123 Wrote: ------------------------------------------------------- > it is a definition that you need to know. > > On your Regression output from Software which is the ANOVA table – Multiple-R is Correlation, R-Squared = Correlation Squared.= Multiple-R Squared. > > and another formula you might need to know > > Adjusted R-squared = 1 - (1-R^2)[(n-1)/(n-k-1)] Re: R square Posted by: ditchdigger2CFA (IP Logged) [hide posts from this user] Date: February 4, 2009 08:36AM I recall reading that correlation squared=R squared only for linear regression, not multiple linear regression.

#6

but why on page 169 the problem is a linear regression, but the answer to problem 16 on page 170 is not using the square of 1.1163, but instead RSS/SST?

#7

June - I didnt refer to the page but nonetheless, in the case of a single linear regression there are 2 ways to calculate the R^2, depending on what information is given. In an SLR, R^2 = correlation ^2 = rss/tss. The formula R^2= correlation does not work for multiple linear regression so you’d have to use the latter method for MLR.

#8

thanks for this insight