Dear StataListers,

I am posting a follow up question to the one posted by Michael Evangelist on 11 June 2018 at 13:41, entitled "Error with lrtest after melogit". I have a similar problem, but with a different error message. The error occurs with "lrtest" as well as with the "lrtest, force" option recommended in the referenced post. The error can be seen in the last two lines of output below. It also occurs with or without added independent variables.

I'm using Stata/SE 15.1 updated as of today (3/14/2019) for Windows 64-bit.

Can anyone help me understand how to get the lrtest command to execute for this model or why it's not currently? Much of what I have read indicates that it should appear by default.

Code:

```
. melogit readmit30 cohort4num || new_hosp_num:, vce(robust) or
```*(iteration log omitted)*
Mixed-effects logistic regression Number of obs = 33,619
Group variable: new_hosp_num Number of groups = 2,192
Obs per group:
min = 1
avg = 15.3
max = 195
Integration method: mvaghermite Integration pts. = 7
Wald chi2(1) = 221.17
Log pseudolikelihood = -15758.173 Prob > chi2 = 0.0000
(Std. Err. adjusted for 2,192 clusters in new_hosp_num)
------------------------------------------------------------------------------------------------------------------
| Robust
readmit30 | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval]
-------------+---------------------------------------------------------------------------------------------------
cohort4num | .5597914 .0218392 -14.87 0.0000 .518583 .6042745
_cons | .2436699 .0038726 -88.84 0.0000 .2361967 .2513795
-------------+----------------------------------------------------------------------------------------------------
new_hosp_num |
var(_cons) | .0145663 .0105003 .0035461 .0598338
--------------------------------------------------------------------------------------------------------------------
Note: Estimates are transformed only in the first equation.
Note: _cons estimates baseline odds (conditional on zero random effects).
.
. lrtest
In the old syntax, the unrestricted model defaulted to a model saved under the name 0. This model was not found.
r(198);

Many thanks for your insight.

## Comment