Converting p-values to z-scores - Stouffer-Method for Metaanalysis (two-tailed regressions)

Hello everyone!

I am currently working on my master's thesis where I have to conduct a meta-analysis on the issue of what determines the cost of debt capital for private firms.
For that purpose I have already collected a bunch of relevant papers which go for that research question under the use of a OLS regression.

Now the issue:
How can I convert the p-values, which go along with the estimated coefficients for the independent variables, to a z-score (which is required to apply Stouffer's-Method to combine p-values)?

What has to be considered is that all reported p-values are for two-tailed test. So, how can I translate a p-value (e.g. of 0.06 or 0.83) into a two-tailed z-score?
Will z-scores always be positive because each may have to be divided by two?

Best regards and thank you in advanced!

More informations about the method:'s_method#Relation_to_Stouffer's_Z-score_method


Ambassador to the humans
Do you only have the coefficients and pvalues or do the have the standard errors for the coefficients as well