Announcement

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

  • Xtlogit Random Effects - Similar funciton to FitStat to choose the model

    Hi, my name is José Guerreiro and I have a sample (with 3,190 firm year-observations) where my dependent variable is binary. My painel data is unbalanced (t= 2014 to 2022). So, I think xtlogit is the best statistic way for this sample. Am I correct?
    After tested many variables to get a model I can't choose the best result because I don't have the option to fit model as in Logit Model's or Reg Model's.
    My main question is: How I can choose the best result? Stata have some test to fit and help to choose the best result?
    I put 2 results here:

    1st Result tested
    xtlogit VariableDependentSus Indepvariable1ing Indepvariable2 Indepvariable3 Indepvariable4 Indepvariable5 Indepvariable6 Indepvariable7 Indepvariable8 Indepvariable9 Indepvariable10 Indepvariable11 Indepvariable12 Indepvariable13 Indepvariable14 Indepvariable15, re
    Random-effects logistic regression Number of obs = 2,628
    Group variable: OrbisNumberD~e Number of groups = 404
    Random effects u_i ~ Gaussian Obs per group:
    min = 1
    avg = 6.5
    max = 8
    Integration method: mvaghermite Integration pts. = 12
    Wald chi2(10) = 115.71
    Log likelihood = -267.975 Prob > chi2 = 0.0000
    VariableDependentSus Coef. Std. Err. z P>z [95% Conf. Interval]
    Indepvariable1ing 1.557365 .4080608 3.82 0.000 .7575809 2.35715
    Indepvariable2 -.2218852 .0739719 -3.00 0.003 -.3668674 -.076903
    Indepvariable3 .2681297 .0868202 3.09 0.002 .0979651 .4382942
    Indepvariable4 -.1142315 .0340206 -3.36 0.001 -.1809107 -.0475523
    Indepvariable5 .3053354 .0750156 4.07 0.000 .1583076 .4523632
    Indepvariable6 -.1192569 .0416747 -2.86 0.004 -.2009378 -.037576
    Indepvariable7 -.0979961 .0640854 -1.53 0.126 -.2236012 .0276089
    Indepvariable8 5.294751 .8101083 6.54 0.000 3.706967 6.882534
    Indepvariable9 5.156549 .8187421 6.30 0.000 3.551844 6.761254
    Indepvariable10 .0348831 .0200001 1.74 0.081 -.0043164 .0740827
    Indepvariable11 -.0074051 .0042519 -1.74 0.082 -.0157386 .0009285
    Indepvariable12 -.2477482 .1081448 -2.29 0.022 -.4597081 -.0357884
    Indepvariable13 .0676959 .0267967 2.53 0.012 .0151754 .1202164
    Indepvariable14 -.5387435 .8334802 -0.65 0.518 -2.172335 1.094848
    Indepvariable15 .6975525 2.284144 0.31 0.760 -3.779288 5.174393
    _cons -56.32592 8.704559 -6.47 0.000 -73.38654 -39.26529
    /lnsig2u 4.246848 .158375 3.936439 4.557258
    sigma_u 8.359714 .6619851 7.157921 9.763285
    rho .9550409 .0068003 .9396639 .9666381
    LR test of rho=0: chibar2(01) = 387.60 Prob >= chibar2 = 0.000
    2nd Result tested:
    xtlogit VariableDependentSus Indepvariable1ing Indepvariable16 Indepvariable5 Indepvariable9 Indepvariable10 Indepvariable12 Indepvariable13 Indepvariable17 , re
    Random-effects logistic regression Number of obs = 2,840
    Group variable: OrbisNumberD~e Number of groups = 417
    Random effects u_i ~ Gaussian Obs per group:
    min = 1
    avg = 6.8
    max = 8
    Integration method: mvaghermite Integration pts. = 12
    Wald chi2(8) = 169.54
    Log likelihood = -465.73907 Prob > chi2 = 0.0000
    VariableDependentSus Coef. Std. Err. z P>z [95% Conf. Interval]
    Indepvariable1ing 1.209362 .3202709 3.78 0.000 .5816422 1.837081
    Indepvariable16 2.120578 .8644157 2.45 0.014 .4263546 3.814802
    Indepvariable5 .2186189 .0267567 8.17 0.000 .1661768 .2710611
    Indepvariable9 9.385642 1.033457 9.08 0.000 7.360105 11.41118
    Indepvariable10 .0534956 .0177495 3.01 0.003 .0187073 .0882839
    Indepvariable12 -.4565602 .0993613 -4.59 0.000 -.6513048 -.2618157
    Indepvariable13 .0842074 .0203534 4.14 0.000 .0443154 .1240994
    Indepvariable17 8.163206 1.495749 5.46 0.000 5.231592 11.09482
    _cons -94.0886 8.011558 -11.74 0.000 -109.791 -78.38623
    /lnsig2u 4.921836 .167152 4.594224 5.249448
    sigma_u 11.71556 .9791397 9.94542 13.80076
    rho .9765919 .0038211 .9678099 .9830201
    LR test of rho=0: chibar2(01) = 1070.27 Prob >= chibar2 = 0.000

    Help is really much appreciated.

    Thank you in advance and best regards,
Working...
X