Different results - Stata 18 versus 17

[Tolulope Oladele]

I am conducting a difference-in-differences analysis using the Gardner method on the same dataset with the same code in STATA 18, compared to STATA 17. Both versions are up-to-date. However, I get slightly different results - the same coefficient but different standard errors and p-values. What factors may be responsible for this?

[Andrew Musau]

Please show us the exact commands that you ran and the output from Stata 17 and Stata 18. Place these within CODE delimiters.

[Richard Williams]

Also indicate whether the two versions of Stata are on the same machine or on different machines.

A reproducible example, even if it used different data, could also be nice. I have 17, 18, and 19 all on the same machine so I could see if results differed too.