Re: How to test a distribution for uniformity?
From: Ross Clement (clemenr_at_wmin.ac.uk)
Date: 07/19/04
- Next message: DZ: "Re: False Discovery Rate"
- Previous message: Kevin Saff: "Re: What analysis technique do I use for this problem?"
- In reply to: Ben: "Re: How to test a distribution for uniformity?"
- Messages sorted by: [ date ] [ thread ]
Date: 19 Jul 2004 12:02:34 -0700
Ben <benjamin.kenward@zoology.ox.ac.uk> wrote in message news:<Xns952B563DABEFAbenjaminkenwardzoolo@163.1.2.7>...
>
> Thanks for your advice guys. I am using bins (those are my 45 minute
> slots), but the problem with a chi-square test is that there is no
> contiguity assumed between adjacent bins, so it wouldn't distinguish
> between say:
>
> Time slot: 07.30 08.15 09:00 09:45 10:30 11:15
> No. of observations: 2 0 2 0 2 0
>
> which isn't so bad for me, and:
>
> Time slot: 07.30 08.15 09:00 09:45 10:30 11:15
> No. of observations: 2 2 2 0 0 0
>
> which is much worse.
>
> Using the distribution of intervals between observations is an
> interesting approach which hadn't occured to me, because I was only
> thinking along the lines of bins. Sorry to not make that more clear. I
> would prefer to use bins if I could, as it will greatly simplify things.
> (This is a fairly small detail of my design and ideally I would just
> like to have a sentence or two in a methods section to briefly say how I
> tested that observations were distributed uniformly).
>
> I wonder is there such a thing as a chi-square test which is adjusted to
> assume contiguous bins, or something similar?
Well, as I said before, I'm no expert statistician. But, I can't see
why you couldn't anaylse the data with more than one size of bin. E.g.
with your example data above, the fixed 45 minute bins may miss the
clumping. But, if you also analysed the data in terms of (say) 3 hour
bins, then the clumping in your second example would be more easily
detectable. Plotting a histogram of the timings could suggest
over/under-sampled regions, and you could then use chi-square to test
the hypotheses that they are over or under-sampled compared to a
sampling from a uniform distribution.
Cheers,
Ross-c
- Next message: DZ: "Re: False Discovery Rate"
- Previous message: Kevin Saff: "Re: What analysis technique do I use for this problem?"
- In reply to: Ben: "Re: How to test a distribution for uniformity?"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
