The first part of the course
is an overview of those aspects of experimental design, data analysis,
and analysis interpretation that are relevant to Multidimensional Scaling.
The course then continues with a survey of Multidimensional Scaling models,
with emphasis on the geometrical (visual) aspects of these models and their
use for exploring dissimilarity data to generate hypotheses. Hypothesis
testing is not covered. Following a review of the available MDS algorithms
and programs, a Lisp program implementing a multidimensional scaling algorithm
will be written as the major class project.
The course is designed for quantitative
psychology students who have a mathematical understanding of Multiple Regression
and a working knowledge of LispStat.
Note that the most links in this section are Adobe Acrobat PDF files that
require the free Adobe
Theory and Experimental Design
Kruskal, and de Leeuw
Squares Monotonic Transformation
Nonmetric Multidimensional Scaling Algorithm
and ALSCAL WMDS Algorithms
of MDS Results.
There will be a individual projects invovling the analysis of personally
provided data. The analysis will be written up as a publishable quality
paper. There will be no exams.
Readings are from links provided above and below, and from the following
F.W. & Hamer, R.M. (1987) Multidimensional Scaling: History,
Theory and Applications Erlbaum, New York.
There is very complete set of
related information prepared by William G. Jacoby for his Scaling and Dimensional
Analysis course at the ICPSR (Inter-university Consortium for Political
and Social Research). If you are interested, there is an abstract of next
summer's ICPSR course. On-Line Data Gathering for MDS! Check out this movie