Re: st: pk analysis suite: elementary questions about -pkequiv-, -pkcross-, and -pkshape-

["Joseph Coveney" <jcoveney@bigplanet.com>]

David C Airey wrote:

I thought carryover effects are salient whenever there are repeated measures in
a designed experiment. You should worry that giving a person treatment#1 will
change the person such that giving treatment#2 to the same person will not be
the same as giving treatment 2 to the same person, previously untreated. So the
order in which you give multiple treatments to the same person should be
designed at least, and also modeled if possible. Any text on general
experimental design should address this problem; there are a couple in the Stata
book store. George Cassella's new text, Statistical Design (2008) might be a
good alternative to Winer et al. (http://www.stata.com/bookstore/sped.html, out
of print!)

--------------------------------------------------------------------------------

I should note that what I've been casually calling "carryover" includes what a
biopharmaceutics analyst would actually distinguish in a model:  period,
period-by-treatment interaction, and (first- and higher-order) carryover.  All
of these relate to the influence of a participant's in-study history on the
outcome, and so I've been informally lumping them together under the one term.

Joseph Coveney


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/