[R] Right censored data, abundant in zeros for regression analysis.

[Bert Gunter]

Strictly speaking, this is a statistical analysis issue, not an R
question, although I grant you that the intersection of the two is
nonempty. Nevertheless, I would suggest that you post on a statistics
list like stats.stackexchange.com . In fact, because the issue of how
to effectively deal with such data appears to be far from trivial, you
might better seek local statistical advice. Once you have decided
**what** to do, you could then come back here to inquire about R
packages and procedures to do it -- if you are unable to first find
something through internet search of course.

Cheers,
Bert




Bert Gunter

"The trouble with having an open mind is that people keep coming along
and sticking things into it."
-- Opus (aka Berkeley Breathed in his "Bloom County" comic strip )


On Thu, Dec 24, 2015 at 5:41 AM, REES T. (706713)
<t.rees.706713 at swansea.ac.uk> wrote:
> Hi there,
>
> Firstly forgive me if this seem obvious, if there is existing literature on this i can't find it.
>
> I am looking at conditioning to stimuli and there in the time taken to perform a certain task.
>
> The IV for this data is Conditioning periods ranging from 1-34 periods and the DV is the time taken for the behavioral response to occur 0-300s.
> I am aware that this could simply be looked at through a simple linear regression, however due to the nature of conditioning there is an abundance of zeros in the data.
> On top of this the response time data is right censored (i believe), in that they were given a five minute period to respond after this five minute period (300 seconds) the conditioning period was terminated, so no more data was recorded.
>
> Attached is the data (in .csv format) for time spent out, 0 indicated no time out and 300 indicated all time out during the 5 minutes.
>
> I have considered looking at zero-inflated censored regressions and others similar analysis but I cannot find an analysis that suits the data I have and actually works.
> So what is the best analysis method to deal with this data?
> Admittedly i could be completely missing the target, if that's the case please feel free to say so. Any help with the route that I should go down here would be much appreciated, even if it is blindingly obvious.
>
> Sincerely
>
> Tom Rees
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.