 # Tcrit and degrees of freedom

I sat for L1 long ago in December 08 and I’m not very bright, please excuse all the dumb questions I will have. Hopefully they will help others in my situation:

For hypothesis testing, will there be a t critical value chart to reference on the exam? Is it possible for us to come up with t critical on our own?

Also, what are degrees of freedom and how do you know when to use a denominator of n or n-1 or n-2 etc. When I do a regression correlation problem from scratch, I calculate the means of x and y using just n as my denominator. Then when calculating variance and covariance, I must use n-1. And when I do hypothesis testing, I must use n-2. Why?

Yes, they will provide a t-table.

You start with n degrees of freedom (n = number of data points). For every statistic you calculate (e.g., mean, standard deviation, slope coefficient, intercept, whatever), you lose one degree of freedom. So, if you have 80 data points, and a multiple regression with an intercept and 3 slope coefficients, your degrees of freedom will be 80 – 4 = 76.

S2000 Magician you are the man of the hour!

I’m glad I could be of service. You’re most welcome.

You’re making me blush.

what if I have 1 slope coefficient and want to test the intercept what will be the degree of freedom?

Two.

Not sure what I was thinking there; it’s n − 2, where n is the number of data points.

1 Like