cluster vs principal components analysis

Sharon Mickan (
Fri, 21 May 1999 14:41:34 +1000

I am trying to develop an hierarchical model of effective teams, after
eliciting 480+ bipolar constructs from healthcare managers. Initially I have
completed frequency counts using qualitative analysis software ( I
now want to compare this preliminary hierarchy with a mathematical one -
does anyone have any comments about whether cluster or principal components
analysis would do this better. My reading suggests that I choose which one
suits the data best - but are there any other philosophical or practical
reasons to consider one over the other?

Any thoughts will be appreciated...

Sharon Mickan
PhD Candidate
NHMRC Public Health Postgraduate Research Scholarship

Department of Occupational Therapy
The University of Queensland
St Lucia 4072

Ph: 3365 3456