# st: Extracting fixed effects for graphing

 From "J. Michael Oakes" To statalist@hsphsun2.harvard.edu Subject st: Extracting fixed effects for graphing Date Fri, 09 Mar 2007 14:56:49 -0600

```Greetings - I'm struggling to find an easy yet general way to extract model
coefficients (fixed effects) for use in graphing or related things. I've
examined -parmest- and similar commands but cannot seem to get what I want.

I fit an ANOVA model on some (fictitious) school-randomized trial data. What
I want is a way to extract the fixed-effect coefficients for the cond|school
effects, for when there are any number of schools (there are 20 in this
current case).

If using xtmixed, I can easily get random effect (for school) out with

. Predict re, reffects

Doing same in a fixed effect model is not obvious to me. Yes, -parmest- is
helpful but I cannot seem to get the "row" labels to work, as levels of
cond|school seem only given in the ANOVA regression table,  not in
-parmest-.

I apologize in advance is if this is easy...

===============

. anova math cond cond|school, reg

Source |       SS       df       MS              Number of obs =
311
-------------+------------------------------           F( 19,   291) =
3.39
Model |  138604.165    19  7294.95607           Prob > F      =
0.0000
Residual |  625751.327   291   2150.3482           R-squared     =
0.1813
0.1279
Total |  764355.492   310  2465.66288           Root MSE      =
46.372

----------------------------------------------------------------------------
--
math        Coef.   Std. Err.      t    P>|t|     [95% Conf.
Interval]
----------------------------------------------------------------------------
--
_cons               535.5   11.59296    46.19   0.000     512.6833
558.3167
cond
1    -19.88889     15.933    -1.25   0.213    -51.24742
11.46964
2    (dropped)
cond|school
1  2     15.90889   14.33449     1.11   0.268    -12.30354
44.12132
1 12     13.48889    18.2893     0.74   0.461    -22.50719
49.48497
1 23     12.61966   16.87825     0.75   0.455    -20.59925
45.83857
1 25    -34.96825   16.52453    -2.12   0.035      -67.491
-2.445508
1 31    -39.11111   16.52453    -2.37   0.019    -71.63386
-6.588365
1 35    -.8968254   16.52453    -0.05   0.957    -33.41957
31.62592
1 43     1.769841   14.89501     0.12   0.905    -27.54577
31.08545
1 45    -25.23611     15.933    -1.58   0.114    -56.59464
6.12242
1 70    -5.247475   17.74683    -0.30   0.768    -40.17589
29.68094
1 75    (dropped)
2  3        10.22   14.84623     0.69   0.492    -18.99961
39.43961
2 19     7.233333   16.66593     0.43   0.665    -25.56771
40.03438
2 24    -40.68182   15.23614    -2.67   0.008    -70.66881
-10.69482
2 27    -13.07143   16.97036    -0.77   0.442    -46.47163
20.32877
2 32    -38.94444    19.3216    -2.02   0.045    -76.97225
-.9166402
2 41     -11.5625   16.39492    -0.71   0.481    -43.83016
20.70516
2 44    -11.36667   16.66593    -0.68   0.496    -44.16771
21.43438
2 53         18.5    19.3216     0.96   0.339     -19.5278
56.5278
2 74     13.14286   16.97036     0.77   0.439    -20.25735
46.54306
2 86    (dropped)
----------------------------------------------------------------------------
--

===================

Thanks - Michael Oakes, UMN Epidemiology

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```