🚀 go-pugleaf

A couple of questions for those who use STATISTICA:

In the "process analysis" module of STATISTICA v5.0 there is the
capability to do a power analysis for a test of two samples (or null and
alternate hypothesis).  But what if you want to produce power curves for
a one-way ANOVA where there are more than two groups?  Related
question:how to find sample size giving a specified alpha, beta, and
deviation between null and alternate (for an ANOVA with more than two
groups).

Second question: Sokal and Rohlf give an iterative algorithm for finding
required sample size (Box 9.13 in 2nd edition), but the solution to this
algorithm (for an ANOVA with
only two groups) differs substantially from the answer provided by
STATISTICA.  Is Sokal and Rohlf's iterative algorithm overly conservative,
or am I missing something, or is STATISTICA missing something?

Thanks for the help, and if you need more details, I'd be happy to supply.


____________________________________________________________________
Mark S. Conrad                                 4msc@qlink.queensu.ca
Department of Biology, Queen's University, Kingston, Ontario, Canada