gologit2 autofit option error message

[simon]

Hi,

I’m using gologit2_ado in Statat 15.1 SE to run generalised ordered logit models with partial proportional odds (using autofit). When I introduced an interaction term between respondent’s occupational status (occ4) and their residential characteristic (oecd_ur) in the model, I have an error message as follows:

Step 1: Constraints for parallel lines imposed for 3.educ1 (P Value = 0.9886)
Step 2: Constraints for parallel lines imposed for age (P Value = 0.9522)
pl(0o.oecd_ur#3o.occ4 1o.oecd_ur#3o.occ4 2b.oecd_ur#3o.occ4) is not a subset of the X variables: civic_pcamat 0b.gndr 1.gndr age c.age#c.age 1.genM 2.genM 3.genM
> 4.genM 5.genM 6b.genM 0b.educ1 1.educ1 2.educ1 3.educ1 0b.religion2 1.religion2 2.religion2 3.religion2 0.ideology 1.ideology 2b.ideology 0.occ4 1.occ4 2.occ4 3b
> .occ4 0.oecd_ur 1.oecd_ur 2b.oecd_ur 0.oecd_ur#0.occ4 0.oecd_ur#1.occ4 0.oecd_ur#2.occ4 0o.oecd_ur#3b.occ4 1.oecd_ur#0.occ4 1.oecd_ur#1.occ4 1.oecd_ur#2.occ4 1o.
> oecd_ur#3b.occ4 2b.oecd_ur#0o.occ4 2b.oecd_ur#1o.occ4 2b.oecd_ur#2o.occ4 2b.oecd_ur#3b.occ4 0b.yr 1.yr
The -force- option sometimes lets you proceed anyway

For your information,

“occ4” and “oecud_ur” variables are coded as follow:

Occ4: Respondent’s occupation:

1. Highly skilled white-collar
2. Low skilled white-collar
3. Highly skilled blue-collar
4. Low skilled blue-collar

Oecd_ur: Respondent’s residential characteristic:

1. Predominantly Urban
2. Intermediate
3. Predominantly rural

What makes me perplexed the most is that when I ran the code in Stata 14.2 MP, it proceeds without any problem. Would you please help me with the problem?

Many thanks in advance,

Simon

[David Speed]

Hi Simon,

I too am getting this issue, was it ever resolved? The code worked fine on Stata 13 for me.

Cheers,

David.