Announcement

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

  • Weights not allowed using -meqrlogit-

    Dear all , I am using -meqrlogit-in stata/SE 14.0 in order to conduct Multilevel mixed effects logistic regression. When I regress the command:
    meqrlogit Depvar Indepvars [pweight=wvar]||level2var:,or cov(un)to my data, the stata replied to me "weights not allowed". Thus how can I deal with it?
    How can I check multicollinearity,AIC and BIC for checking Model fit with -meqrlogit- command?
    Sorry for all, I am a fresh Stata user and Statalist forum.
    Respectfully,Hassen

  • #2
    meqrlogit does not accept weights (or the svy prefix), but melogit does, so that's your only option. You'll have to designate a weight for your second level unit; see the Manual entries for melogit ("Quick Start") and meglm ("Survey Data"). For BIC and AIC, see the estat ic post-estimation command after melogit. Whether multicollinearity is an issue depends on your analysis goal and what the variables are. You can get variance inflation factors (vif) by pretending you have a least squares regression problem:
    Code:
    regress Depvar Indepvar
    estat vif
    Read the FAQ, especially FAQ 12 and follow it's instructions. In future posts, describe the analysis question and show the actual code you ran, plus the results of that code, inside code delimiters .



    Steve Samuels
    Statistical Consulting
    [email protected]

    Stata 14.2

    Comment


    • #3
      Dear Steve Samuels, Thank you very much!! Your response is extremely helpful to improve my MPH Thesis.
      I am using Huawi Mobile, so can I really apply the instruction of FAQ 12? I want to apply all rules and instructions of the Statalist Forum, but I am a fresh Statalist forum user!! How can I use them?
      Sorry for all my sweet!!
      With Best Wishes,Hassen

      Comment


      • #4
        Can you transfer files from your computer to your phone? If not, I guess you cannot show actual code and results.
        Steve Samuels
        Statistical Consulting
        [email protected]

        Stata 14.2

        Comment

        Working...
        X