|
|
Thanks to all who replied and helped me to find the individal and group
means for my three scaled survey.
No, on to the meat of the investigation. I need to do a cluster analysis of
the sixty individual cases (participants). I think it's a hierarchical
cluster analysis procedure due to the smaller number of cases. Each person
has his or her own profile for the three scales (mean for each scale can
range from 1-5), i.e. : Case #1âscale 1 score =3.9, scale 2 score= 4.0 and
scale 3 score=2.5.
Scale score 1, 2 and 3 (individual means) are all variables in the data
editor.ID #'s are the cases.
I need to sift the sixty cases into subgroups/typologies that maximize
differences between groups/types and minimize differences within
groups/types. I suspect it will be a 4-5 group solution. For example, there
probably will be a group of cases/participants where all three scale scores
are "high", a group of cases where all three scale scores are "low," a group
of cases where one scale score is low and the other two scale scores are
"high," etc.
When others have done similar research they have used the Ward's method.I
don't know if that would work here or not. I really have no idea how to go
about doing this and specific help (steps) would be welcome.
Thanks!!
JC
|
Locked
