help ctems
--------------------------------------------------------------------------------------------------------------------------------------------
Title
ctems -- Cornfield-Tukey expected mean squares
Syntax
ctems , index(list) levels(list) random(list) [ comment(string) ]
options description
--------------------------------------------------------------------------------------------------------------------------------------
Main
index() list of all the individual subscripts
levels() list of levels for each subscript
random() list of random effects (1) and fixed effects (0)
comment() optional comment
--------------------------------------------------------------------------------------------------------------------------------------
Remarks
ctems applies the Cornfield-Tukey alorithm to a special dataset to compute the expected mean squares for an anova model. The dataset
has two string variables: effect and subscript. There are as many rows as there are terms in the anova linear model. In each row,
effect is an ascii representation of each effect and subscript are the letters (and parentheses) of the index subscript.
ctems is not a particularly easy program to use. The program knows nothing about ANOVA or your design other than what is entered in
the dataset and the command options. It cannot detect inconsistencies or deficiencies in your model. Please use ctems carefully.
Examples
. use crfdesign, clear /* two factor design */
. clist
effect subscript
1. A j
2. B k
3. A*B jk
4. e i(jk)
. ctems, index(i j k) levels(8 4 2) random(1 0 0)
. use crf422design, clear /* three factor design */
. ctems, i(i j k l) l(8 4 2 2) random(1 0 1 0)
Reference
Kirk, Roger E. (1998) Experimental Design: Procedures for the Behavioral Sciences, Third Edition. Monterey, California:
Brooks/ColePublishing. ISBN 0-534-25092-0
Author
Phil Ender
Statistical Consulting Group
UCLA Academic Technology Services
ender@ucla.edu
Statistical Methods and Data Analytics
