how to analyse underlying multiple causes of death data

I have some data of tree mortality registered each 3 month. In the same date I registered sign of dryness, insect folivory or mammal predation (The plants were not always found dead with these signs). I would like now to analyse the data with SPSS and determine which of these signs are more responsible for tree death. My dataset structure is: dead 0/1, dryness 1/0, insect 1/0 etc. and I was thinking doing a survival analyse with Cox's proportional hazards.
This kind of issue always occured in human death analysis, with multiple reasons that may explain human death but I did not find any clear method.
Any advice would be helpful.


Less is more. Stay pure. Stay poor.
Welcome to the forum.

Yes, discrete time survival analysis seems appropriate here. I am not versed in ecological data, so there may be extra consideration and assumptions to take into account like the trees may not be independent. With humans and most observations, it is assumed the units are independent, but if trees are near each other, their exposures and outcomes may be correlated.
Thank you for your answer and yes I guess the units are not really independent as competition for resources may occur. It was a bit my issue here when I checked data analysis for human death that need to be independent. That's one of the reason I posted this message. Thanks anyway!!


Less is more. Stay pure. Stay poor.
I am not saying you can't use that method, you just need to investigate the indenpendency issue, in regression you would use multilevel models. I am not sure if there is a similar approach for survival or use of robust standard errors. I haven't had this issue before.