Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Mlogit does not converge when conducting multiple imputation

    Dear reader,
    I am currently trying to implement MICE, however, it does not seem to work.
    My data has a lot of missingness, I am afraid that this might be the problem. I have currently used the following code:

    Code:
    mi set mlong
    mi register imputed overhours_new irregular_hours_new side_job_new heavy_burden_new skills_new support_new appreciation_new career_advancement_new tenure workingstudent contracttype paidwork released_finding_job seeking_after_interruption student care_household private_means flo disability voluntary_work uncertainty_new parental_leave informal_care mha mhb mhd mhc mhe mh geslacht gebjaar gross_wages_cat work_outside_office_hours work_during_evening work_at_night satisfaction_atmosphere
    mi impute chained (ologit, augment) overhours_new irregular_hours_new heavy_burden_new skills_new appreciation_new career_advancement_new uncertainty_new gross_wages_cat (mlogit, augment) work_outside_office_hours work_during_evening work_at_night side_job_new tenure workingstudent contracttype paidwork released_finding_job seeking_after_interruption student care_household private_means flo disability voluntary_work parental_leave informal_care geslacht (regress) gebjaar satisfaction_atmosphere = nomem_encr wave, add(10) force
    I added the augment option twice. I have tried adding the 'force' option, but that did not help.

    Stata issues the following error message:
    convergence not achieved
    convergence not achieved
    mlogit failed to converge on observed data
    error occurred during imputation of overhours_new irregular_hours_new heavy_burden_new
    skills_new appreciation_new career_advancement_new uncertainty_new gross_wages_cat
    work_outside_office_hours work_during_evening work_at_night side_job_new tenure
    contracttype paidwork released_finding_job seeking_after_interruption student
    care_household private_means flo disability voluntary_work parental_leave informal_care
    geslacht gebjaar satisfaction_atmosphere on m = 1

    I understand that convergence is already an issue in the first iteration. I am not sure, however, how to fix this issue. I have tried also using 'ologit' for all multinomial variables, I have read that this might make a change, however, I get the same error message.

    Thank you for taking the time to read my post.

    Best wishes,
    Ciel
Working...
X