Survival Analysis Example

[@louitje]

Ani takes the time to very clearly explain a difficult subject and to provide easy practical guidance to repeat these analyses. Thank you!

[@adrianoaxel1196]

Hi, great videos! I would like to make just a remark / suggestion. In your example, people who claim unemployment insurance stay longer without a new job and by your comments things are seen as sort of causal, or I had this impression. It should be mentioned that this analysis only show correlation and not causation. We don't know, for example (solely on this data) what are the general conditions of those who claimed the insurance. If it is due to some sicknes or some other problem, than the correlation found is expected but is not a bad thing, the insurance is there exactly to allow for this longer time without problems. The distinction between correlation and causation is very important in any data analysis and the consideration of possible external causes or biases, i.e., something not explicitly seen by the variables at hand, is always very important. This point made, I'm appreciating the videos and the iniciative. :)

[@meshackamimo1945]

Hi, This is a wonderful tutorial.so easy to understand even for a non economist , like me

[@saxboss1]

Another way to come up with the survival function value a bit quicker is to take 1- F(t) where F(t) is the total number of failures thus far, divided by the original starting sample. For the second row you’d have: (1- (374/3343)) and you get the result of .85. This is instead of multiplying out all the conditional probabilities which could be easier if you’re finding the survival rate of a later point in time where many periods have passed.

[@dataman1000]

Great lesson, thank for the do files

[@rohail260]

amazing....thank you very much

[@鄭達翼]

thank you very much! This is really useful and clear

[@kennethgottfredsen767]

Best explanation.

[@ma.kresnanavarro5745]

thank you so much!