Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: Re: vanelteren (was st: ologit - conditional?)

From   Joseph Coveney <[email protected]>
To   Statalist <[email protected]>
Subject   Re: Re: vanelteren (was st: ologit - conditional?)
Date   Wed, 27 Jul 2005 07:06:45 +0900

Ricardo Ovaldia wrote:

I am a bit concerned about using -vanelteren- on these data. My understanding 
is that this test performs a rank-sum test for each dam and then constructs a 
weighted linear combination of these Wilcoxon statistics. My problem is that 
the rank-sum
test assumes that you have two independent samples, however in my case, the 
embryos allocated to the two medias come from the same dam and thus are not 
independent within strata (dam).

Is the test still valid?


You can think of van Elteren's test as analogous to an analysis of variance 
with a blocking factor.  In the normal case, you would have -anova embryo_score 
culture_medium dam-.  Dam is a blocking factor within levels of which you 
anticipate clustering of scores.  Conditional on dam, however, the scores would 
be independent.

One common use of the van Elteren test is with clinic as the stratification 
variable in a multicenter clinical study.  van Elteren's test is used in lieu 
of randomized blocks ANOVA with clinics as blocks.  Instead of -anova response 
treatment clinic- (or perhaps -xtmixed response treatment || clinic:-), you'd 
have -vanelteren response, by(treatment) strata(clinic)-.

Joseph Coveney

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index