I am using STATA currently on a server and my own computers. I noticed during the last couple of days an inconsistency when using estadd after a regression. A simple example is the code:
Where the expected result is the following:
I do get that result on two of my computers while using STATA through a server only the first scalar after the regression is added. See below:
The first result is using STATA 13.1 on Windows 7 enterprise or Windows 10 and STATA 13.1 and the second using STATA14.1 through a server (Windows server 2012 R2 Standard). Note that I receive different results even if I use "version 13.1" on the server computer. This is a fairly new problem as I did not have a problem last week.
In both cases estout is up to date and the same do-file is used. See attachments for a snapshot pictures of the difference.
Any suggestions for why this is occurring?
Code:
sysuse auto.dta, clear
eststo xx: reg mpg price
estadd local bb "yea"
estadd local cc "yea"
estadd local dd "yea"
estadd local ee "yea"
esttab xx, sca("bb ff" "cc fff" "dd aaa" "ee dds")
Code:
----------------------------
(1)
price
----------------------------
mpg -238.9***
(-4.50)
_cons 11253.1***
(9.61)
----------------------------
N 74
ff yea
fff yea
aaa yea
dds yea
----------------------------
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001
Code:
----------------------------
(1)
price
----------------------------
mpg -238.9***
(-4.50)
_cons 11253.1***
(9.61)
----------------------------
N 74
ff yea
fff
aaa
dds
----------------------------
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001
In both cases estout is up to date and the same do-file is used. See attachments for a snapshot pictures of the difference.
Any suggestions for why this is occurring?