We are pleased to announce two forthcoming short courses which will be
presented by Professor Dallas E. Johnson, Kansas State University, USA.
Analysis of Messy Data: Experimental Designs used in Medical Research 15-17
February 1999
Design and Analysis of Crossover Experiments using SAS GLM and SAS
MIXED procedures 18-19 February 1999
For further information and registration details please contact Kellie
Watkins at The University of Reading ([log in to unmask]), providing
an address and/or fax number. Please note also that the deadline for
registration is Friday 22 January 1999.
____________________________________________________________________________
___________________
*** Analysis of Messy Data:Experimental Designs used in Medical Research ***
This course is specifically designed for those who are involved with
analysing certain kinds of messy experiments that involve ordinal and/or
continuous data. In particular, emphasis is placed on analysing those
experiments in which restrictions have been used when assigning
experimental units (often subjects) to treatment groups. These kinds of
experiments include those done in many locations and those that involve
taking repeated measurements on each subject. A correct analysis of such
experiments usually involves using models that contain more than one error
term. This course will show how to use the SAS procedures GLM and MIXED to
obtain correct analyses of these kinds of messy experiments.
The objectives of this course are to discuss and compare different
alternatives for obtaining a correct analysis of repeated measures and the
other kinds of mixed model experiments, how to write the appropriate SAS
commands, how to choose an appropriate error structure, how to interpret
the resulting outputs, and how to make correct inferences. The advantages
and disadvantages of using SAS-MIXED and SAS-GLM will also be discussed.
*** The Design and Analysis of Crossover Experiments using SAS-GLM and
SAS-MIXED Procedures ***
This course is specifically designed for those who are involved with
analysing certain kinds of messy experiments that involve crossover designs
with ordinal and/or continuous data. Crossover designs are a special type
of a repeated measures experiment where the experimental units are given
different treatments prior to taking the repeated measures. Crossover
designs offer some real advantages over traditional designs when one is
comparing different treatments, as crossover designs allow each
experimental unit to serve as its own control. However there are certain
pitfalls that must be avoided if crossover designs are going to be used
effectively. A correct analysis of such experiments usually involves using
mixed models that contain more than one error term. This course will
introduce the important issues and discuss appropriate statistical analyses
of crossover designs using SAS-GLM and SAS-MIXED.
Professor Johnson is currently Head of the Department of Statistics at
Kansas State University. For more than 20 years he held a half time
appointment as a Statistical Consultant in the Kansas Agricultural
Experiment Station where he was often faced with finding a correct analysis
for messy experiments. Professor Johnson is a Fellow of the American
Statistical Association, a member of the International Biometric Society, a
member of the Institute of Mathematical Statistics, and the current and
founding editor of the Journal of Agricultural, Biological and
Environmental Statistics. He is the co-author of two books with George A.
Milliken entitled Analysis of Messy Data, Vol 1 - Designed Experiments and
Analysis of Messy Data, Vol 2 - Nonreplicated Experiments.
Kellie Watkins
Statistical Services Centre
The University of Reading
Harry Pitt Building
PO Box 240
Whiteknights Road
Reading RG6 6FN
UK
Tel: +44 (0)118 931 8689
Fax: +44 (0)118 975 3169
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|