Overidentification tests ivreg2

[Joan Marti]

Hi,

I am trying to calculate the Sargan Hansen overidentification tests for a regression using ivreg2 but the command "estat overid" is not working. Moreover, when I convert it into a table using esttab, the Sargan test from the scalar sarganp is blank. How can I calculate the overidentification test and add it to the table?

Thank you very much,

Joan

[Andrew Musau]

estat works with official Stata commands, in this case, ivregress. ivreg2 and esttab are community-contributed from SSC, as you are asked to explain. The Sargan statistic (overidentification test of all instruments) is displayed at the foot of the ivreg2 table after estimation. If you cannot find it, update your installation of the command. Otherwise see https://www.statalist.org/forums/for...-stage-outcome on how to export specific statistics using esttab.

[Joan Marti]

Thank you very much. I managed to solve the problem by partialling out the fixed effects, and coefficients / standard errors remain the same but my R2 has decreased significantly. Is this the normal impact of partialling out the fixed effects?


Thank you very much

Joan

[Joro Kolev]

You will not get very helpful advice the way how you are asking your questions.

Show exactly what you typed at Stata, exactly what Stata returned, and phrase your questions in terms of these.

Thank you very much. I managed to solve the problem by partialling out the fixed effects, and coefficients / standard errors remain the same but my R2 has decreased significantly. Is this the normal impact of partialling out the fixed effects?


Thank you very much

Joan