Use choice of insurance type as the outcome variable, based on their income ethnicity,
, maximum likelihood estimates require large sample size. Multiple equations also require large sample size.
multinomial logistic regression model
basecategory
mlogit ses female write
mlogit, rrr
relative risk ratio
test and constraint
. use hsb2, clear (highschool and beyond (200 cases))
. mlogit ses female write
Iteration 0: log likelihood = -210.58254 Iteration 1: log likelihood = -201.60669 Iteration 2: log likelihood = -201.41478 Iteration 3: log likelihood = -201.4146
Multinomial logistic regression Number of obs = 200 LR chi2(4) = 18.34 Prob > chi2 = 0.0011 Log likelihood = -201.4146 Pseudo R2 = 0.0435
------------------------------------------------------------------------------ ses | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- low | female | .9231032 .4010904 2.30 0.021 .1369804 1.709226 write | -.0295236 .0202396 -1.46 0.145 -.0691925 .0101453 _cons | .2599696 .9925834 0.26 0.793 -1.685458 2.205397 -------------+---------------------------------------------------------------- high | female | -.2707787 .3514253 -0.77 0.441 -.9595597 .4180023 write | .052736 .0200114 2.64 0.008 .0135143 .0919577 _cons | -3.20461 1.074845 -2.98 0.003 -5.311268 -1.097952 ------------------------------------------------------------------------------ (ses==middle is the base outcome)
. mlogit, rrr
Multinomial logistic regression Number of obs = 200 LR chi2(4) = 18.34 Prob > chi2 = 0.0011 Log likelihood = -201.4146 Pseudo R2 = 0.0435
------------------------------------------------------------------------------ ses | RRR Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- low | female | 2.517089 1.00958 2.30 0.021 1.146806 5.524684 write | .970908 .0196508 -1.46 0.145 .933147 1.010197 -------------+---------------------------------------------------------------- high | female | .7627853 .2680621 -0.77 0.441 .3830615 1.518924 write | 1.054151 .0210951 2.64 0.008 1.013606 1.096318 ------------------------------------------------------------------------------ (ses==middle is the base outcome)
. mlogtest, hausman
**** Hausman tests of IIA assumption
Ho: Odds(Outcome-J vs Outcome-K) are independent of other alternatives. You used the old syntax of hausman. Click here to learn about the new syntax.
(storing estimation results as _HAUSMAN)
Omitted | chi2 df P>chi2 evidence ---------+------------------------------------ low | -0.004 3 1.000 for Ho high | -0.307 3 1.000 for Ho ----------------------------------------------
. mlogtest, all
**** Likelihood-ratio tests for independent variables
Ho: All coefficients associated with given variable(s) are 0.
ses | chi2 df P>chi2 -------------+------------------------- female | 8.247 2 0.016 write | 13.657 2 0.001 ---------------------------------------
**** Wald tests for independent variables
Ho: All coefficients associated with given variable(s) are 0.
ses | chi2 df P>chi2 -------------+------------------------- female | 7.635 2 0.022 write | 12.326 2 0.002 ---------------------------------------
**** Hausman tests of IIA assumption
Ho: Odds(Outcome-J vs Outcome-K) are independent of other alternatives. You used the old syntax of hausman. Click here to learn about the new syntax.
(storing estimation results as _HAUSMAN)
Omitted | chi2 df P>chi2 evidence ---------+------------------------------------ low | -0.004 3 1.000 for Ho high | -0.307 3 1.000 for Ho ----------------------------------------------
**** Small-Hsiao tests of IIA assumption
Ho: Odds(Outcome-J vs Outcome-K) are independent of other alternatives.
Omitted | lnL(full) lnL(omit) chi2 df P>chi2 evidence ---------+--------------------------------------------------------- low | -45.264 -43.770 2.987 3 0.394 for Ho high | -39.714 -38.070 3.288 3 0.349 for Ho -------------------------------------------------------------------
**** Wald tests for combining outcome categories
Ho: All coefficients except intercepts associated with given pair of outcomes are 0 (i.e., categories can be collapsed).
Categories tested | chi2 df P>chi2 ------------------+------------------------ low- high | 14.478 2 0.001 low- middle | 5.799 2 0.055 high- middle | 6.955 2 0.031 -------------------------------------------
**** LR tests for combining outcome categories
Ho: All coefficients except intercepts associated with given pair of outcomes are 0 (i.e., categories can be collapsed).
Categories tested | chi2 df P>chi2 ------------------+------------------------ low- high | 16.248 2 0.000 low- middle | 6.168 2 0.046 high- middle | 7.435 2 0.024 -------------------------------------------
. log close log: d:datastatamlogit.txt log type: text closed on: 20 Jun 2006, 21:10:02 -------------------------------------------------------------------------------
example
model assumptions: function form, IIA, link function
model interpretation
sample size assumption
model fit information
alternative models
Hausman test
mlogtest
ovtest hettest
rvfplot