What is the difference between R^2 and regression sum of squares?

P

pmond

New member
Joined
Jun 18, 2026
Messages
0
Reaction score
0
They both measure the variation in the dependent variable explained by the independent variable. Yet are calculated differently. I feel like this should be easy but I don’t get it.
 
Regression sum of square alone (RSS), when calculated, doesn’t explain anything. It is merely a number. You must use this number RSS to divide by SST to make some sense out of it. The result would be a % of SST< which represents the independent variable’s explanatory power. I hope this helps.
 
The RSS is just the absolute amount of explained variation, the R squared is the (RSS/SST), i.e. the absolute amount of variation as a proportion of total variation.
It’s just like saying Net income is an absolute figure, whereas ROE is Net Income as a proportion of Equity.
 
So just want to confirm, I can get to R^2 two ways, by squaring the correlation coefficient or looking at an ANOVA table that thas RSS and SST?
 
pmond wrote:
So just want to confirm, I can get to R^2 two ways, by squaring the correlation coefficient or looking at an ANOVA table that thas RSS and SST?
Click to expand...
Both valid methods.
A third; (SST-SSE)/SST (where SST-SSE = RSS)
 
pmond wrote:
So just want to confirm, I can get to R^2 two ways, by squaring the correlation coefficient or looking at an ANOVA table that thas RSS and SST?
Click to expand...
However, you can only square the correlation coefficient when you have a single independent variable, while the other method always works.
 
You must log in or register to reply here.
Back
Top