On 18 Dec 1999 23:50:55 GMT, [EMAIL PROTECTED] (Archtopist) wrote:
[...]
> I've constructed a 2X5 contingency table for vegetation type (burn/no burn, veg
> type 1, type 2... type 5) and ran a chi square using the number of acres in
> each cell as N.
> I am concerned that this is a large and arbitrary sample - in that I could have
> used meters or inches, thus changing the rather large chi square value.
>
> Do you agree that chi square tests are not an option, and could you recommend
> an alternative (some sort of test of equality of proportions)?
Absolutely, not an option, using "number of acres" for what is burnt.
You don't give us much to work with. What are you trying to model:
the start of fires, the spread of fires, or some combination?
For a contingency table, you need the important entries to be
independent, so that "one" represents one fire -- not one acre. Thus,
you could conceivably model "where each fire started" as an entry (+1)
in the respective cell out of 5, with the "expectations" formed by the
relative acreage. That assumes that you have more than 4 or 5 fires
to be tabulated, and that the hypothesis is interesting (that is,
where it starts is not already known or easily explained).
--
Rich Ulrich, [EMAIL PROTECTED]
http://www.pitt.edu/~wpilib/index.html
