Re: st: Gllapred for transition probabilities

[Milda Toleikytė <mildatoleikyte@gmail.com>]

Dear statalist,

I think I should expand a bit on what I have done so far with my model

Laged employment state:
status2: unemployment t-1
status3: inactivty t-1

I estimated a gllamm model:
gllamm statust  AGE gender married education Sluoksnis unemployment
status2 status3 year2 status20082 status20083, expand(patt chosen m)
i( PB030) link(mlogit) family(binom) nrf(2) eq(a2 a3) nip(8)

I think that in order to get a transitions between employment status
in time t and t-1 I have to compute 9 marginal probabilities such as:
gllapred prob1 if status2==1, marg mu outcome(1)

and obtain confidence levels by:

ci_marg_mu lower upper if alt==1, level(95) resp(1000)

Would you comment on the routine I proposed?

Thank you in advance,

Milda Toleikyte
Master degree student at Aarhus University, Denmark

2014-02-01 20:18 GMT+02:00 Milda Toleikytė <mildatoleikyte@gmail.com>:
> I am investigating labour market transitions in my master thesis. So
> far I have fitted dynamic multinomial logit model with random effects
> using gllamm. My dependent variable is employment state- employment,
> unemployment, inactivity. Now i would like to calculate the
> population-averaged probabilities of labour market transitions. I
> believe I have to use gllapred, but cannot figure out what exact
> syntax to use. Could you advice me how should I proceed?
>
> Regards,
>
> Milda Toleikyte
> Master degree student in Aarhus University, Denmark
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/