Hi Rebecca, Could it fall in the framework of survival analysis with "competing events" or "competing risks". In the medical field (not mine) it can be used to predict healing vs death, dying from disease A or B, etc. As far as I know, it assumes that the risk of competing events are independent.
I have not gone far enough to code it in R, but you can read on it in: Kleinbaum DG, Klein M (2005) Survival Analysis, A Self-Learning Text, 2nd edn. Springer, London (and many other texts I'm sure) Good luck! Mick PhD candidate Natural Resource Sciences McGill University, Canada ________________________________________ From: [email protected] [[email protected]] On Behalf Of [email protected] [[email protected]] Sent: October 24, 2009 6:00 AM To: [email protected] Subject: R-sig-ecology Digest, Vol 19, Issue 15 Send R-sig-ecology mailing list submissions to [email protected] To subscribe or unsubscribe via the World Wide Web, visit https://stat.ethz.ch/mailman/listinfo/r-sig-ecology or, via email, send a message with subject or body 'help' to [email protected] You can reach the person managing the list at [email protected] When replying, please edit your Subject line so it is more specific than "Re: Contents of R-sig-ecology digest..." Today's Topics: 1. survival analysis: flowering time (Rebecca Ross) 2. Re: survival analysis: flowering time (Chris Gast) ---------------------------------------------------------------------- Message: 1 Date: Fri, 23 Oct 2009 23:15:02 +0100 From: Rebecca Ross <[email protected]> Subject: [R-sig-eco] survival analysis: flowering time To: "[email protected]" <[email protected]> Message-ID: <756b64e07365af43be945a734a85e44528c36b0...@exmbx03.ad.oak.ox.ac.uk> Content-Type: text/plain; charset="us-ascii" Dear All, I am using survival analysis to compare flowering time between different populations in a field experiment. I have 3 possible outcomes: a) flowered during experiment b) had not flowered and were alive at harvest c) had not flowered and were dead at harvest Clearly, b) are right censored. But I am unsure what to do for c) as they were not censored (event will never happen), but equally I do not have a 'time to event' for them. To make things more complicated, dying before harvest is not independent of flowering time as being on the verge of death would make them also less able to flower, therefore recording them as being censored might be misleading. Apologies if this is a naive question, it is my first time with survival analysis! Any thoughts much appreciated! Rebecca ------------------------------ Message: 2 Date: Fri, 23 Oct 2009 23:50:12 +0000 (UTC) From: Chris Gast <[email protected]> Subject: Re: [R-sig-eco] survival analysis: flowering time To: [email protected] Message-ID: <[email protected]> Content-Type: text/plain; charset=us-ascii Hi Rebecca, I am reminded of a statistical technique from preclinical biostatistics (carcinogenicity) called the Peto prevalence-mortality test. Essentially, it is a method for assessing time to tumor development, when tumor development and mortality are related. I wonder if this, or some similar method can be adapted to your situation, wherein tumor development is akin to flowering....? Chris ------------------------------ _______________________________________________ R-sig-ecology mailing list [email protected] https://stat.ethz.ch/mailman/listinfo/r-sig-ecology End of R-sig-ecology Digest, Vol 19, Issue 15 _______________________________________________ R-sig-ecology mailing list [email protected] https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
