>generally speaking,if i want to do a chi2 test,the expacted count of each cell
>should more than 5.but if i have less,what statistical test should i take? is
>fisher's exact test a good idea?
There's controversy among the experts about both the rule of thumb and whether
Fisher's exact test is best (see, for example,
http://www.math.yorku.ca/Who/Faculty/Monette/S-news/0669.html ). But in most
cases, it probably doesn't matter greatly.
>my data table is as follow,and if i use exact test ,the p-valuve is 0.00.is
>it wise to use it? thank you !
>
> col
>row 1 2 3 Total
>
>1 6 12 2 20
>2 18 68 43 129
>3 9 4 1 14
>
>Total 33 84 46 163
I recommend avoiding stating p-values as equal to zero. Rather, I recommend in
this case stating it as "P<0.005" if you don't know it with any more precision.
If you're using Stata (-tabulate row col, exact-), then you can get the value
by typing -return list-, which in the case of this table is 0.0002.
Joseph Coveney
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/