Hi
On Mon, 16 Oct 2000, Kathleen Morgan wrote:
> While TIPsters are on the topic of statistics--can anyone out there give
> me the names and "recipes" for post-hoc tests to use with Chi-Square?
I don't know if there are named post-hoc procedures for
Chi-square, but it is certainly possible to partition a
chi-square with >2 levels in ways analogous to planned contrasts
in ANOVA. My distant memory is that it is essentially reducing
the table in appropriate fashion and re-calculating the
Chi-square, although there might have been some other or possibly
shortcut approach. For example, given a 3(1,2,3)x2 table, one
might do the overall Chi-square and then partition it into
2(1,2)x2 and 2(1+2,3)x2 tables. If truly planned, then perhaps
little or no correction of ps required. If "fishing" in post-hoc
fashion, then some adjustment helpful. The tricky thing with
higher levels (e.g., 5x4, 3x4x2) would probably be working out
the orthogonal components.
Best wishes
Jim
============================================================================
James M. Clark (204) 786-9757
Department of Psychology (204) 774-4134 Fax
University of Winnipeg 4L05D
Winnipeg, Manitoba R3B 2E9 [EMAIL PROTECTED]
CANADA http://www.uwinnipeg.ca/~clark
============================================================================
