Mark sample in runmlwin

[RAWILDEMAN29]

Dear colleagues,

I would like to know how to activate the marksample function of Stata in runmlwin.

I know in Stata, it is marksample1 if xvar==value, and then after that if correlation, it is corr x y if sample1==1.

Thank you,
Russell

[ChrisCharlton]

The same syntax that you referred to should apply, i.e.:

Load data:

Code: Select all

use http://www.bristol.ac.uk/cmm/media/runmlwin/tutorial, clear

Run model on full dataset:

Code: Select all

. runmlwin normexam cons standlrt, level2(school: cons) level1(student: cons) nopause
 
MLwiN 2.31 multilevel model                     Number of obs      =      4059
Normal response model
Estimation algorithm: IGLS

-----------------------------------------------------------
                |   No. of       Observations per Group
 Level Variable |   Groups    Minimum    Average    Maximum
----------------+------------------------------------------
         school |       65          2       62.4        198
-----------------------------------------------------------

Run time (seconds)   =       1.08
Number of iterations =          4
Log likelihood       = -4678.6211
Deviance             =  9357.2422
------------------------------------------------------------------------------
    normexam |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        cons |   .0023908   .0400224     0.06   0.952    -.0760516    .0808332
    standlrt |   .5633712   .0124654    45.19   0.000     .5389394    .5878029
------------------------------------------------------------------------------

------------------------------------------------------------------------------
   Random-effects Parameters |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
Level 2: school              |
                   var(cons) |   .0921275   .0181475      .0565591     .127696
-----------------------------+------------------------------------------------
Level 1: student             |
                   var(cons) |    .565731   .0126585      .5409209    .5905412
------------------------------------------------------------------------------

Mark subsample to use:

Code: Select all

. mark sample1 if girl==1

Run model on subsample:

Code: Select all

. runmlwin normexam cons standlrt if sample1==1, level2(school: cons) level1(student: cons) nopause
 
MLwiN 2.31 multilevel model                     Number of obs      =      2436
Normal response model
Estimation algorithm: IGLS

-----------------------------------------------------------
                |   No. of       Observations per Group
 Level Variable |   Groups    Minimum    Average    Maximum
----------------+------------------------------------------
         school |       55          1       44.3        120
-----------------------------------------------------------

Run time (seconds)   =       1.21
Number of iterations =          3
Log likelihood       = -2761.4929
Deviance             =  5522.9858
------------------------------------------------------------------------------
    normexam |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        cons |    .059983   .0461238     1.30   0.193    -.0304179     .150384
    standlrt |   .5563728   .0164685    33.78   0.000     .5240951    .5886504
------------------------------------------------------------------------------

------------------------------------------------------------------------------
   Random-effects Parameters |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
Level 2: school              |
                   var(cons) |   .0965134   .0217593       .053866    .1391608
-----------------------------+------------------------------------------------
Level 1: student             |
                   var(cons) |   .5403317   .0156572      .5096441    .5710193
------------------------------------------------------------------------------

You can also skip the -mark- command and use an expression directly:

Code: Select all

. runmlwin normexam cons standlrt if girl==1, level2(school: cons) level1(student: cons) nopause
 
MLwiN 2.31 multilevel model                     Number of obs      =      2436
Normal response model
Estimation algorithm: IGLS

-----------------------------------------------------------
                |   No. of       Observations per Group
 Level Variable |   Groups    Minimum    Average    Maximum
----------------+------------------------------------------
         school |       55          1       44.3        120
-----------------------------------------------------------

Run time (seconds)   =       1.23
Number of iterations =          3
Log likelihood       = -2761.4929
Deviance             =  5522.9858
------------------------------------------------------------------------------
    normexam |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        cons |    .059983   .0461238     1.30   0.193    -.0304179     .150384
    standlrt |   .5563728   .0164685    33.78   0.000     .5240951    .5886504
------------------------------------------------------------------------------

------------------------------------------------------------------------------
   Random-effects Parameters |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
Level 2: school              |
                   var(cons) |   .0965134   .0217593       .053866    .1391608
-----------------------------+------------------------------------------------
Level 1: student             |
                   var(cons) |   .5403317   .0156572      .5096441    .5710193
------------------------------------------------------------------------------

[RAWILDEMAN29]

Thank you Chris.

Much appreciated,
Russell