Hi:
A student of mine has a rather unusual design and I can't quite figure
out how to code it for an ANOVA. I should've caught it earlier but
unfortunately
I didn't and the data collection is now over. Any thoughts or
suggestions would be
appreciated sincerely.
Here's the scoop. She looked at the influence of either vinyl or
carpet
in a hospital room on a variety of objective (acoustic, air quality,
etc.) and self-report measures (satisfaction, mood, etc.). At one
level the design is very simple; that is, pariticpants were essentially
randomly assigned to rooms that were either carpeted or not. There
were
three rooms of each type and anywhere from 5 to 7 patients were at one
time in each
of the rooms. If we just look at the "objective" data, I believe we
have a
straightforward nested design with Room nested within Floor Type and
Patients nested within Room. If we collapse across Room the design
would reduce to a simple two group comparision but the rooms differed
enough that I think we have to deal with the loss of power and bite
the quasi-F bullet.
The problem is when we look at the self-report data. Here, we have
ratings from both staff and patients on the same questions. The rub is
that a given patient only answered questions about the room they were
in,
whereas the staff answered questions about both the vinyl and carpeted
rooms generally. As a result, what I initially though was going to be
a simple
2X2 design (Participants (Patients or Staff) by Floor covering (Carpet
or Vinyl) is actually not so simple because the Floor Covering is
crossed for
Staff and nested for Patients. Any suggestions?
Thanks.
lou tassinary
--
****************************************************************
“Colamus humanitatem” [Seneca]
To violate a person's ability to distinguish fact from fantasy
is the
epistemological equivalent of rape [Noretta Koertge]
Louis G. Tassinary, Ph.D. Texas A&M
University
Environmental Psychophysiology Lab College Station, TX
77843-3137
College of Architecture
[log in to unmask]
http://archone.tamu.edu/architecture/faculty/tassinary/mainframe.html
http://www.earlyamerica.com/review/summer97/secular.html
****************************************************************
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|