The
Fourth Course in the ECAS Programme:
Analysis of
Categorical Data |
LEIDEN, THE NETHERLANDS |
September
20-24, 1993 |
Scientific
Programme Committee
J.C.
Gower (Leiden), Chair
D.
Edwards (Maeloev)
A.
Giovagnoli (Perugia)
W.J.
Heiser (Leiden)
W.
Krazanowski (Exeter)
W. Lebart
(Paris)
W.
Molenaar (Groninger)
Organising
Committee
W.J.
Heiser (Leiden), Chair
I.A. Van
Der Lans (Leiden), Secretary
J.J.
Droesbeke (Bruxelles)
S. Heiler
(Konstanz)
J.J
Meulman (Leiden)
D.
Peña (Madrid)
G. Saporta
(Paris)
S. Verdel
(Leiden)
Scope
of the course
Categorical
data arise in many different situations, and in a great variety of
scientific fields. They may appear as counts in contingency tables, or
in the form of a matrix of cases by attributes. The course covers
recently developed methods, for both exploratory and confirmatory modes
of analysis. Exploration is often guided by dimension reduction and
visual display of relationships among variables and their categories,
by grouping of observation units, and data transformation. Methods for
confirmation can be considering models in which the concept of
conditional independence -that applies to manifest and latent variables
alike- plays a crucial role.
Topics
• correspondence analysis
• homogeneity analysis and
optimal scaling
• generalized canonical
correlation analysis
• cluster analysis and measures
of association
• logistic regression and
discrimination
• independence graphs and
graphical modelling
• latent class and latent
structure models
• discrete space / discrete
time Markov modelling
Lecturers
Erling B. Andersen
David Hand
Willem J. Heiser
Rolf Langeheine
Jaqueline J. Meulman
Gilbert Saporta
Joe Whittaker