I have data on birth intervals of women. Birth interval refers to the time interval from one child's birth date until the next child's birth date. Here, dependent variable is the birth interval, defined as the length of time between two successive births for different parities, which are measured in single months. Here four duration of birth are used. A variety of demographic and socio-economic characteristics influence women's spacing practices., therefore the above mentioned dependent variables depend on a number of demographic and socio-economic factors. We deals with many explanatory variables like:

1.survival of the previous child,
2.mother's age at the birth of the index child,
3.mother's education,
4.mothers job,
5.sex of the index child,
6.place of residence,
7.division etc.
Are these data correlated survival data? How can we analyze this? Is this multi-level frailty model reasonable?

  • 3
    $\begingroup$ Can you give more information? The best recommendation for the appropriate analysis is always dependent on the research question which you haven't said anything about. Some more info about the data would also be helpful $\endgroup$
    – Macro
    Jun 30, 2012 at 5:03
  • 1
    $\begingroup$ I agree with Macro. @MichaelChernick 's solution might be right, but it depends on what question you are asking. Are the birth intervals the dependent variable or an independent variable? What is you research question? See my blog post How to ask a statistics question $\endgroup$
    – Peter Flom
    Jun 30, 2012 at 11:57

2 Answers 2


When events recur they are called recurrent events. There is a theory that has been developed for recurrent events and even two books. One by Richard Cook and Jerald Lawless and the other by Wayne Nelson.


Yes, these data are correlated survival data. You could use a Cox model on this data and a simple variance adjustment to deal with correlation, i.e., add "+ cluster(id)" to the coxph model, where "id" is a variable that identifies individual mothers.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.