Ordered Probit (similar to Ordered Logit): Pseudo R2 Meaning?

#1
Ordered Probit (similar to Ordered Logit): How to graph?

Hi!

I have just run a oprobit and am curious about how to graph the results.

Code:
. oprobit  numrate NSPFrac

Iteration 0:   log likelihood = -111627.39  
Iteration 1:   log likelihood = -111415.17  
Iteration 2:   log likelihood = -111415.17  

Ordered probit regression                         Number of obs   =      69358
                                                  LR chi2(1)      =     424.44
                                                  Prob > chi2     =     0.0000
Log likelihood = -111415.17                       Pseudo R2       =     0.0019

------------------------------------------------------------------------------
     numrate |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     NSPFrac |  -.3592971   .0174446   -20.60   0.000    -.3934879   -.3251064
-------------+----------------------------------------------------------------
       /cut1 |  -1.069671   .0123295                     -1.093836   -1.045505
       /cut2 |  -.4764826   .0118506                     -.4997094   -.4532558
       /cut3 |   .0317118   .0117738                      .0086355     .054788
       /cut4 |   .6193138   .0120697                      .5956577    .6429699
From there I did this and simply graphed the results. However, I was expecting an S-shaped graph.
Code:
. predict star1 star2 star3 star4 star5
(option pr assumed; predicted probabilities)


. twoway (scatter star1 NSPFrac) (scatter star2 NSPFrac) (scatter star3 NSPFrac) (scatter star4 NSPFrac) (scat
> ter star5 NSPFrac)
Did I graph the interpretation correctly? What else should I/what should I have done instead? (picture is best I could do with sizing limits)
 
Last edited: