Good afternoon everyone,

I have a replication task, which involves the use of outreg command. In their replication code the authors use outreg in a very unfamiliar way (paper dates back to 2013). Below the code lines, in bold problematic lines :

######

regress voted_10g treat_5-treat_3 if s_r_v == `i', robust
testparm treat_5-treat_7
local secr = `r(p)'
testparm treat_1 treat_2
local sos = `r(p)'
lincom treat_5 + treat_6 - treat_7
test treat_1=treat_2

*Test average of secrecy interventions against SOS civic duty.

lincom (treat_5 + treat_6 + treat_7)/3 - treat_4

*First table does not include the covars.

outreg treat_5-treat_3 using "logs\Table2_OLSTreatmentEffects.out", se ctitle("`collabel'") addstat("F-Test p-value on three Secrecy Treatments",`secr',"F-Test p-value on two SOS Placebos",`sos',"Covariates included?",0) addnote("Full results with covariates in Table A2") bracket adec(6) rdec(6) 3aster `repl'

*Second table includes everything in model.

outreg using "logs\TableA4_OLSTreatmentEffects.out", se ctitle("`collabel'") addstat("F-Test p-value on three Secrecy Treatments",`secr',"F-Test p-value on two SOS Placebos",`sos') bracket adec(6) rdec(6) 3aster `repl'

######

Two errors can be reported :
varlist not allowed
and the fact that addstat is not an option.

The former is because they specify as an argument of outreg not a model name, but variable names. In the current version of outreg, one would rather write keep(treat_5-treat_3) as an option.

A solution anyone? I cannot write again the code (time consuming) and I already tried to use earlier versions of Stata (from 9.0 on). Authors wrote in their code to use version 10.1. But it does not work.

Thank you very much,

Emile T.