We are announcing the following course, taking place at the Statistical
Services Centre, The University of Reading.

For further information and registration details please contact Kellie
Watkins ([log in to unmask]) providing an address and/or fax number. 


Course: "Using SAS PROC GLM"	Dates:  14-15 October 1999

Content:

o	Review of the general linear model.  Comparison of regression models.
o	Analysis of variance (balanced and unbalanced).
o	The MEANS and LSMEANS statements.
o	Analysis of covariance.
o	The four types of sums of squares.  The pros and cons of Type III sums of
squares.
o	Estimability and estimating functions.  The ESTIMATE and CONTRAST
statements.
o	Cross-over designs.
o	Repeated measurements analysis.  The REPEATED statement.

The GLM procedure in SAS uses the method of least squares to fit general
linear models, including those used in simple and multiple regression,
analysis of variance for balanced and unbalanced data, analysis of
covariance and cross-over studies.

PROC GLM allows specification of the model using a very flexible model
formula, calculates parameter estimates and analyses of variance, estimates
combinations of parameters, handles multiple error terms and multivariate
responses.  The generality of the procedure means it is often less than
straightforward to specify an analysis and interpret the results.  The
output includes jargon which does not appear in elementary or intermediate
level text books on regression and analysis of variance, such as "TYPE I-IV
sums of squares", "estimable functions" or "least squares means".  In this
course the facilities of PROC GLM are described, the nature of the four
types of sums of squares and the use of estimating functions will be
covered in detail and the output interpreted.  The course will also include
a critical assessment of the facilities offered by PROC GLM.

----------------------------------

It will be assumed that participants are SAS users, who understand the main
components of the system, in particular the DATA step.  It is also assumed
that participants are familiar with multiple regression and analysis of
variance, and have had previous exposure to the analysis of variance for
unbalanced data.

---------------------------------


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


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%