The t stat is the coefficient divided by the standard error. To compute a p-value, one needs the degrees of freedom of the error of the t distribution, dfe, which is the sample size minus the number of estimated parameters. To calculate a p-value, one can execute the following command:
pValue = 2*(tcdf(-|t|,dfe));
The t-statistic is used to test the hypothesis that the parameter being estimated is actually 0, thus it is irrelevant. Since the 95% confidence interval is about 2 standard errors on either side of the estimate, and the t statistic is just the estimate divided by the standard error, the rule of thumb is that a t stat with absolute value > 2 is significant, and rejects the hypothesis of a 0 parameter value. Hence one does not need the p-value.
Best Answer