Probit Inverse Mills Ratio

[Gosia Aslan]

How can I compute Inverse Mills Ratio from probit?

Thanks in advance

[Oded Mcdossi]

Here is a useful link:
http://www.stata.com/support/faqs/st...s/mills-ratio/

[Emad Shehata]

Dear Gosia

the next example will explain 3 different methods to compute Inverse Mills Ratio from probit regression

clear all
input y x
0 1.9
0 2.4
0 1.1
0 0.5
0 1.2
1 3.6
1 9.1
1 8.1
1 9.8
1 3.3
0 6.6
0 6.1
end

 probit y x

* Inverse Mills Ratio from [predict probit postestimation]
 predict IMR1 , score

 predict xb , xb
 predict cdf1 , pr

 gen double cdf=normal(xb)
 gen double pdf=normalden(xb)

 gen double IMR2=pdf/cdf
 replace IMR2=pdf/(cdf-1) if y == 0

 gen double IMR3= cond(y == 1, pdf/cdf, pdf/(cdf-1))

 format y %10.0f
 format xb pdf cdf cdf1 IMR1 IMR2 IMR3 %10.4f
 list y xb pdf cdf cdf1 IMR1 IMR2 IMR3

list y xb pdf cdf cdf1 IMR1 IMR2 IMR3

     +----------------------------------------------------------------------+
     | y        xb      pdf      cdf     cdf1      IMR1      IMR2      IMR3 |
     |----------------------------------------------------------------------|
  1. | 0   -1.0656   0.2261   0.1433   0.1433   -0.2639   -0.2639   -0.2639 |
  2. | 0   -0.9097   0.2638   0.1815   0.1815   -0.3222   -0.3222   -0.3222 |
  3. | 0   -1.3151   0.1680   0.0942   0.0942   -0.1855   -0.1855   -0.1855 |
  4. | 0   -1.5022   0.1291   0.0665   0.0665   -0.1383   -0.1383   -0.1383 |
  5. | 0   -1.2839   0.1750   0.0996   0.0996   -0.1943   -0.1943   -0.1943 |
     |----------------------------------------------------------------------|
  6. | 1   -0.5355   0.3456   0.2961   0.2961    1.1672    1.1672    1.1672 |
  7. | 1    1.1796   0.1990   0.8809   0.8809    0.2259    0.2259    0.2259 |
  8. | 1    0.8677   0.2738   0.8072   0.8072    0.3392    0.3392    0.3392 |
  9. | 1    1.3978   0.1502   0.9189   0.9189    0.1634    0.1634    0.1634 |
 10. | 1   -0.6291   0.3273   0.2646   0.2646    1.2368    1.2368    1.2368 |
     |----------------------------------------------------------------------|
 11. | 0    0.4000   0.3683   0.6554   0.6554   -1.0687   -1.0687   -1.0687 |
 12. | 0    0.2441   0.3872   0.5964   0.5964   -0.9595   -0.9595   -0.9595 |
     +----------------------------------------------------------------------+

My best regards

[Gosia Aslan]

Dear Oded Mcdossi
and
Dear Emad Shehata

as usual thanks so much

Yours
Gosia Aslan

[Gosia Aslan]

Dear Oded
I tried to Compute Inverse Mills Ratio from Stata website, But I got different results, as Emad methods !!!
So, where the problem?

[Gosia Aslan]

Here Stata link

http://www.stata.com/support/faqs/st...s/mills-ratio/

Any comments?

[Emad Shehata]

Dear Gosia
First thanks for your notice
In fact the two results are correct in spite of different calculations
I will prepare an example for that in the next post.

Emad

[Emad Shehata]

check here for details

http://www.statalist.org/forums/foru...atio-a-comment

[Gosia Aslan]

Dear Emad
Thanks so much for your help

Yours
Gosia

[Xiaoguang Li]

Here Stata link

http://www.stata.com/support/faqs/st...s/mills-ratio/

Any comments?

I do not think that's correct. That seems only stata5, and phat2 is square rather than a variables. So, it actually not work now.