[R] Summarizing dummy coefficients in sem package

From: Gus Jespersen <jesper_at_u.washington.edu>
Date: Tue, 03 Jun 2008 14:13:51 -0700


Greetings,
I am working in the sem package on a model with 3 exogenous variables (2 are nominal-categorical), and 4 endogenous, continuous variables. To use sem with the nominal variables, I created dummy variables. Now, in my sem output I have estimates for path coefficients for the relationship between each level of the nominal variables and the endogenous variables they are associated with. However, what I am interested are the overall path coefficients from each nominal variable to the various endogenous variables, not the relationship between each level of this and the endogenous variables. How might I go about achieving this?

Here is my model ('sex' and 'trt' are the nominal variables with 2 and 4 levels, respectively) followed by the relevant output:

   Path                      Parameter     
1  sexMale->nect             sexnect       
2  sexMale->cono             sexcono       
3  sexMale->herb             sexherb       
4  sexMale->para             sexpara       
5  trtControl->nect          trtContnect   
6  trtControl->cono          trtContcono   
7  trtControl->herb          trtContherb   
8  trtControl->para          trtContpara   
9  trtNoFemale->nect         trtNoFemnect  
10 trtNoFemale->cono         trtNoFemcono  
11 trtNoFemale->herb         trtNoFemherb  
12 trtNoFemale->para         trtNoFempara  
13 trtNoMale->nect           trtNoMalenect 
14 trtNoMale->cono           trtNoMalecono 
15 trtNoMale->herb           trtNoMaleherb 
16 trtNoMale->para           trtNoMalepara 
17 area->nect                areanect      
18 area->cono                areacono      
19 area->herb                areaherb      
20 area->para                areapara      
21 nect<->para               nect<->para   
22 nect<->herb               nectherb      
23 nect<->cono               nectcono      
24 cono<->herb               conoherb      
25 cono<->para               conopara      
26 area<->area               errorarea     
27 nect<->nect               errornect     
28 cono<->cono               errorcono     
29 herb<->herb               errorherb     
30 para<->para               errorpara     
31 trtControl<->trtControl errortrtControl 32 trtNoFemale<->trtNoFemale errorNoFemale
33 trtNoMale<->trtNoMale     errortrtNoMale
34 sexMale<->sexMale         errorsexMale  

 Parameter Estimates
                Estimate    Std Error  z value    Pr(>|z|) 
sexnect         -0.01929179 0.06376330 -0.3025532 7.6223e-01
sexcono         -2.66789515 0.36084653 -7.3934344 1.4300e-13
sexherb         -0.00743378 0.00566452 -1.3123412 1.8941e-01
sexpara          0.01097552 0.00980425  1.1194652 2.6294e-01
trtContnect      0.21164052 0.08736725  2.4224240 1.5417e-02
trtContcono     -0.44791548 0.54528423 -0.8214349 4.1140e-01
trtContherb     -0.00038585 0.00756348 -0.0510144 9.5931e-01
trtContpara     -0.00015034 0.06098602 -0.0024651 9.9803e-01
trtNoFemnect     0.11658323 0.08786423  1.3268566 1.8456e-01
trtNoFemcono    -0.10709039 0.55754523 -0.1920748 8.4768e-01
trtNoFemherb    -0.00412156 0.00777418 -0.5301598 5.9600e-01
trtNoFempara     0.00420467 0.03406450  0.1234325 9.0176e-01
trtNoMalenect    0.23656338 0.08897775  2.6586801 7.8447e-03
trtNoMalecono    0.07633944 0.60103928  0.1270124 8.9893e-01
trtNoMaleherb   -0.00156381 0.00788036 -0.1984436 8.4270e-01
trtNoMalepara    0.00910566 0.03304680  0.2755383 7.8290e-01
areanect         0.01097847 0.02250164  0.4878965 6.2562e-01
areacono         0.37431397 0.12725982  2.9413367 3.2680e-03
areaherb        -0.00169721 0.00200841 -0.8450508 3.9808e-01
areapara         0.00044190 0.00365491  0.1209054 9.0377e-01
nect<->para      0.00399596 0.00425721  0.9386334 3.4792e-01
nectherb        -0.00028131 0.00245373 -0.1146438 9.0873e-01
nectcono         0.52509632 0.15739193  3.3362340 8.4922e-04
conoherb         0.01380912 0.01401081  0.9856041 3.2433e-01
conopara         0.01752004 0.02413323  0.7259716 4.6786e-01
errorarea        1.90255983 0.08980542 21.1853566 0.0000e+00
errornect        0.82560204 0.03897074 21.1851759 0.0000e+00
errorcono       26.60775814 1.25597309 21.1849746 0.0000e+00
errorherb        0.00661525 0.00031226 21.1851940 0.0000e+00
errorpara        0.01967170 0.00092856 21.1852827 0.0000e+00
errortrtControl  0.19005279 0.00897095 21.1853556 0.0000e+00
errorNoFemale    0.18331925 0.00865314 21.1852797 0.0000e+00
errortrtNoMale   0.19059782 0.00899668 21.1853547 0.0000e+00
errorsexMale     0.24736468 0.01167625 21.1852846 0.0000e+00

Thanks,
Gus



R-help_at_r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. Received on Tue 03 Jun 2008 - 22:39:00 GMT

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.2.0, at Tue 03 Jun 2008 - 23:30:42 GMT.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-help. Please read the posting guide before posting to the list.

list of date sections of archive