Degrees of freedom

when we look into t-table, do we look for degrees of freedom n-1 or n-k-1? (k=number of independent variables)

Or do we always look for n-2 regardless of k since we are testing for one variable at a time?


Always look for n-k-1.

For simple regression you have k=1 always…so you look for n-1-1=n-2

Hope it helped.

Yeap… Thanks!

single regression DOF = N-2

Multiple Regression DOF = N-(K+1)

For _ any _ regression: DOF = N – (K + 1)

(For a simple (single) regression, K = 1, so DOF = N – (K + 1) = N – (1 + 1) = N – 2.)

got it… thanks for confirming both of you.

My pleasure.