Re: st: repeated events employment

Dear Marcel,

One approach is multilevel duration (or survival, or event history) modeling: repeated spells of employment and unemployment
are nested within individuals, thus residuals accross spells are correlated.

People who favor discrete-time event-history analysis would suggest that you use random-effects logistic regression on a person.period
dataset. If you favor continuous time survival modeling, maybe streg with the shared option is the solution, but I
There is a similar problem in demography: repeated spells of marriages, birth intervals, etc. An influential author seems to be Lee Lillard and the keyword
multilevel event history analysis oc births

On 11/05/19, marcel marcel <[email protected]> wrote:

Dear Statalisters,

I have longitudinal data on employment - unemployment histories. By construction, the data set consists of individuals that have been unemployed at least once during the period for which data are available.
Some individuals experience only one episode of unemployment, others alternate between employment and unemployment (short spells of unemployment followed by short spells of employment. The hypothesis is that some characteristics of individuals influence the probability of the number of times these individuals alternate between these two states. My question is what type of analysis to use to test this hypothesis.

I can think of duration models, poisson models or markov models, but I'am not sure which approach is best suited for this problem.

Best regards,

Marcel


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

From	"Dr. Bartus" <[email protected]>
To	[email protected]
Subject	Re: st: repeated events employment - unemployment
Date	Thu, 19 May 2011 21:07:20 +0200

Stata: Data Analysis and Statistical Software

Re: st: repeated events employment - unemployment