What is the df for the AR t test Also just wanted to make sure I wasn’t off base with the formula t = (correlation error at time t - correlation error at time t-1) ----------------------------------------------------------------------- (1/sq root (n)) thanks
i assume you are talking about to test the significance of the autocorrelation residuals (correlation among the current and lagged residuals). if this is the case; your formula is not correct. it should be: correlation_coefficient *sqrt (T) OR: correlation_coefficient / (1/sqrt (T)) where correlation_coefficient = residual correlation. df = T - 1 where T is the number of periods in the sample.