Re: Nested ANOVA
From: jwalla (john_wallace_at_affymetrix.com)
Date: 11/19/04
- Previous message: Bruce Weaver: "Re: Nested ANOVA"
- In reply to: Bruce Weaver: "Re: Nested ANOVA"
- Next in thread: Bruce Weaver: "Re: Nested ANOVA"
- Reply: Bruce Weaver: "Re: Nested ANOVA"
- Messages sorted by: [ date ] [ thread ]
Date: 19 Nov 2004 12:12:32 -0800
Thanks for your reply, Bruce. Maybe I can clarify what I think you
mean with an example:
In a nested study, the researcher wants to identify the effect of a
teaching technique on standardized test scores for students. She picks
one school with two classrooms for the control technique and another
school with two classrooms for the test technique (with identical
numbers of idealized, homogenous students, naturally). Since students
in this example cannot simultaneously be members of more than one
classroom, students are nested within classroom for the purpose of
analyzing the data.
In a repeated measures version of the test, you'd only need one school,
and your populations would consist of the same students before and
after exposure to the teaching technique. In that case, you wouldn't
need to nest the students in classrooms, since the independent variable
(teaching technique) has nothing to do with how the students are
grouped?
Naturally there's all kinds of confounding that would need to be
controlled for in a real study like that, but for the purpose of
discussing grouping of replicates and nesting terms is my example
accurate?
- Previous message: Bruce Weaver: "Re: Nested ANOVA"
- In reply to: Bruce Weaver: "Re: Nested ANOVA"
- Next in thread: Bruce Weaver: "Re: Nested ANOVA"
- Reply: Bruce Weaver: "Re: Nested ANOVA"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
|