STA254
Department of Statistics, Stanford University, Fall, 2008
Correspondence Analysis and Related Methods
Universitat Pompeu
Fabra, Barcelona
Slides
Week 1 (colour) Week
1 (black&white)
Week 2 (colour) Week
2 (black&white)
Week 3 (colour) Week
3 (black&white)
Week 4 (colour) Week
4 (black&white)
Week 5 (colour) Week
5 (black&white)
Week 6 (colour) Week
6 (black&white)
Week 7 (colour) Week
7 (black&white)
Week 8 (colour) Week
8 (black&white)
Week 9 (colour) Week
9 (black&white)
Week10(colour) Week10(black&white)
R
scripts
Week 1 chidist braycurtis
jaccard
Week 6 See also Appendix
B of Correspondence Analysis in Practice
R
packages (zipped)
Correspondence analysis ca
Three-dimensional graphics rgl
Classification & regression trees tree
Multivariate analysis in ecology vegan
Data
sets
BioEnvGeo EU Salud Salud2 WomenWork Environmental
questionnaire data
Homeworks
Week 5: read chapter 10
of Correspondence Analysis in Practice
Data
sets from ISSP 2002 survey: get documentation
in issp02.pdf
Working woman can establish just as warm a relationship with her
child (v4)
Men’s job is to work; women’s job is to look after household
(v11)
Couple can live together without getting married (v22)
Working mother should get paid maternity leave (v27)
How satisfied with life in general (v52)
How satisfied with family life (v54)
Week9: supplementary
reading chapter 20 of Correspondence Analysis in
Practice
Articles
Popular article on
correspondence analysis
Publishing
quantitative results
Appendix A (Theory of
Correspondence Analysis) of Correspondence Analysis in Practice
Additional
reading
Chapter on non-Euclidean
distance
Chapter on distance and
correlation between variables
Chapter on hierarchical
clustering
CARME
network
Basic information
The accent in this
course is on learning about tools, mostly for visualizing multivariate
data, that can be applied to practical
problems. Correspondence analysis (CA) plays a central role in this area because it applies to count data, the
most basic form of statistical measurement.
Many other data types (raw categorical data, preferences, ratings,
continuous measurements, distances) can be recoded in a form suitable for being
visulalized using correspondence analysis, hence CA is a versatile framework
for data visualization. CA is routinely
used by ecologists, who count the occurrences of plants and animals, and social
scientists, who count the responses of people; also by market researchers, linguists,
psychologists, biomedical researchers and archeologists, to name a few...
Classes: Tuesdays & Thursdays
12h50-14h05, Sequoia 200.
Some reading and/or
homework every week, some of which is graded.
Homework mainly consists of
applying R functions to data sets and interpreting the results-- Naras’s course
on Computational
Tools for Statistics is highly
recommended.
Homework counts 20% towards final grade.
Final project counts
70% towards final grade.
Class attendance
& participation counts 10% towards final grade.
My office hours for
seeing students are Mondays 15.00-18.00 and Thursdays 16.00-18.00 –
my office is room Sequoia
128.
You can arrange a
meeting at other times by contacting me by email.
You can use my
regular email addresses for communication but I actually prefer to channel all
emails about teaching through this address:
mmr.upf (at) gmail.com
so please try to use
that address if possible – this will be more efficient.