Estimating P-value for one sided hypothesis

[MelinaEm]

Hi!

I am just starting to use E-views 7. I ran a regression with two coefficients which gave me
T-statistic =18.53247 w. P-value= 0.0000
T-statistic = 2.490049 w. P-value = 0.0169

Now, as I understand it, it calculates P-value for a two-sided hypothesis. If I have to do a one sided hypothesis, both B1 <0, and B1> 0, how do I calculate the P-value for them? Someone told me that if I am calulating a hypothesis for a value on the left tail, I should just divide the P-value by two. This gives me 0.0845 however, and according to the Z-table a T-statistic of 2.49 should give me a P-value of 0.994. What am I missing?

Specifically, what are the rules of calculating correct P-values from E Views 7?

Thank you for your time!

[EViews Glenn]

You need to use the t-distribution not the standard normal. And you need to use the upper-tail (1-@ctdist) rather than just the @ctdist.

[MelinaEm]

Thank you