how to set the p-vlaues after regression to one-tailed test?

[Fred Lee]

After using regress, I want to output the p-values, I want to use two-tailed tests for controls and one-tailed tests for indepedent variables, how to output the two kinds of p-values quickly?
I know reg2docx, however, the output for p-values are always two-tailed.

[Rich Goldstein]

it is not clear where you are going with this but see the following FAQ: https://www.stata.com/support/faqs/s...nts/index.html

[Fred Lee]

it is not clear where you are going with this but see the following FAQ: https://www.stata.com/support/faqs/s...nts/index.html

Thank you, Rich!
I know we can test one-tailed tests one by one using your guidance link.
However, I want to output the pvalues of one-tailed tests rather than just doing the analyses.
More specificaly,

sysuse auto
reg price foreign length
est store m1
reg2docx m1  using "result.docx"

The code above will output the coefficient and pvalues of two-tailed tests, however, I want to out put p-values of one-tailed tests for foreign and p-values of two-tailed tests for lenth.

[Joro Kolev]

-reg2docx- is a user written command, and when you read the help file you see that it does not have an option for reporting one sided p-values.

Thank you, Rich!
I know we can test one-tailed tests one by one using your guidance link.
However, I want to output the pvalues of one-tailed tests rather than just doing the analyses.
More specificaly,

sysuse auto
reg price foreign length
est store m1
reg2docx m1 using "result.docx"

The code above will output the coefficient and pvalues of two-tailed tests, however, I want to out put p-values of one-tailed tests for foreign and p-values of two-tailed tests for lenth.

[Fred Lee]

-reg2docx- is a user written command, and when you read the help file you see that it does not have an option for reporting one sided p-values.

Thanks Joro! No, it didn't. I used "reg2docx" as an example, I want to know wheher there is a command that can report one sided p-values.

[daniel klein]

One-sided p-values are usually not appropriate. Anyway, one-sided p-values are simply obtained by dividing the two-sided p-value by 2 (and subtracting the result from 1 if the sign has the opposite direction of your one-sided hypotheses).

[Fred Lee]

One-sided p-values are usually not appropriate. Anyway, one-sided p-values are simply obtained by dividing the two-sided p-value by 2 (and subtracting the result from 1 if the sign has the opposite direction of your one-sided hypotheses).

Yes, in our research field, sometimes we can use one-sided p-values, I want to report it in a quick and convenient way.

[Joro Kolev]

To my knowledge there is no package that exports automatically one sided p-values.

Thanks Joro! No, it didn't. I used "reg2docx" as an example, I want to know wheher there is a command that can report one sided p-values.

[Fred Lee]

To my knowledge there is no package that exports automatically one sided p-values.

Thank you！