assume 46 observations, 2 slopes, 1% significance level. H0: intercept >= -10% HA: intercept < -10% Thus, we have a one tailed test with n-k-1 = 46-3=43 DOF. The critical t value from the T table is +2.416. The t-stat is given as -0.96. Now, I am not exactly sure from this point on. What is the rule for a 1 tailed Ttest? I know what it is for a two tailed T-test… The answer was we DO NOT REJECT null because the t-stat was less than -2.416. Why are we looking at the negative critical value if it is a one tailed test?

I have not begun studying again yet, but I believe that to test the validity of the results it is based on absolute value of the t-stat so sign is not relevant in that vain, but it is relevant in the fact that +/- signs show correlation to the dependent variable. i.e. if the t is positive then the dependent will go up with an increase in the indie, and if its negative it will go down with an increase in the indie.

Because H0 dealt with a negative number to begin with. If it is negative and a one tailed test then you will need to take the critical t value as negative and go from there. Any number less than -2.416 will mean that your fail to reject the null hypothesis.