Page 1 of 1

Presentation with -esttab-

Posted: Fri Jan 27, 2012 9:25 am
by Jamoo
Hi list,

I run a random slope model with sex (boy) and the constant, leaving cov(cons, boy) covariance term in Stata (see example below). Note the comma between cons and boy for the covariance term. When I try to input these results into an -esttab- command, it says that the 'boy option is not allowed', presumably because Stata is thinking that the comma is part of the command that I'm inputting. Any ideas about how I could get rid of the comma, so that the term would be called "consboy" or something which would work with esttab?

With best wishes,

Jamie

---example starts----

.runmlwin etc.

preamble etc

fixed effects etc

------------------------------------------------------------------------------
Random-effects Parameters | Estimate Std. Err. [95% Conf. Interval]
-----------------------------+------------------------------------------------
Level 2: |
var(cons) | .1055654 .0109918 .0840217 .127109
cov(cons,boy) | .0108248 .0109 -.0105388 .0321884
var(boy) | .0084793 .0174866 -.0257937 .0427523
-----------------------------+------------------------------------------------
Level 1: |
var(cons) | .643547 .0110398 .6219095 .6651845

---example ends---

Re: Presentation with -esttab-

Posted: Fri Jan 27, 2012 1:26 pm
by GeorgeLeckie
It would appear that you are using a very old version of runmlwin!

We now only present the comma in cov(cons,boy) in the displayed output.

The underlying parameter is reffered to a [RP2]cov(cons\boy)

We changed this half a year ago because we ran into the same problem that you did!

To install the latest version...

Code: Select all

. ssc install runmlwin, replace
And then check that you are using the latest version by typing

Code: Select all

. which runmlwin
The output should say...

*! runmlwin.ado, George Leckie and Chris Charlton, 01Jan2012

Re: Presentation with -esttab-

Posted: Thu Feb 02, 2012 10:27 am
by Jamoo
Hi George,

Lovely - sorted. Thanks.

Jamie