Time-effect among groups

[Emanuele Spina]

Hi,

i'm testing the efficacy of the same treatment on three different groups of patients.
I chose anova test (https://stats.idre.ucla.edu/stata/se...is-with-stata/) for repeated measures, and this is my command:

anova edsst gravitygroup / id | gravitygroup time gravitygroup#time, repeated(time)

in which edsst is my dependent, gravitygroup my independent variable (the three group of patients), and time represents the two time points in which i analyzed the effect of the same treatment.

I'd like to perform a post hoc analysis, comparing the difference group by group (1 vs 2, 2 vs 3 and 1 vs 3) to verify the magnitude of difference and to test if there is only one group that differs from others.

How can i obtain these results?

thanks a lot

[Carlo Lazzaro]

Emanuele:
have you taken a look at -help contrast- and related entry in Stata .pdf manual?

[Emanuele Spina]

No, can you help me?

[Carlo Lazzaro]

Emanuele:
take look at the examples reported in -contrast- entry, Stata .pdf manual.
As the Stata .pdf manula comes with your Stata copy, you can easily access it yourself.

[Emanuele Spina]

https://stats.idre.ucla.edu/stata/se...is-with-stata/

sounds good?!

[Carlo Lazzaro]

Emanuele:
yes, it does.
And it mentions -contrast-.