AuthorTopic: degrees of freedom
DJScratch
@2009-11-02 23:36:00
I'm having trouble learning degrees of freedom. And this seems to be a part of all of the level II tests.
It is related to the number of observations?
From the number of observations, I think you subtract one plus the number of variables in the regression equation. Is this correct?

Don't miss this question on the test!!!! Past history (the last 3 years) shows it WILL be on the test.
For example:
People who pass the test = 50 + .7(hours studied) + .6(mba)+.1(luck)
Let's say the number of observations is 20
The degrees of freedom would be 20-3-1.
Is this correct? (I think so.)
ahoe
@2010-04-02 05:31:00
There are two method of calculating the degree of freedom for a regression model.

First method use n - d - 1, where the n represents number of observation and d represents number of independent variables in the model.

Second method use n - d, where the n still represents number of observation and the d, however, represents number of variable (independent variable + constant) in the right hand side of the equation.

Both method will come out with the same degree of freedom.

### CFA Discussion Topic: degrees of freedom 