Readme file
SERIES C
 Applied Statistics
The effect of drop-out on the efficiency of longitudinal experiments
G. Verbeke and E. Lesaffre
Journal of the Royal Statistical Society, Series
C, Applied Statistics, Appl. Statist. Volume
48 (1999), 363 - 375
The file 'rats.dat' contains the rat data described in Section 2 of the paper.
The data set has 4 columns:
1 : defines the treatment groups
('con' = control, 'hig' = high dose, 'low' = low dose)
2 : identification number of the rat
3 : age of the rat (in days) at which a specific measurement was taken
4 : the response of interest (in pixels)
The results reported in Table 1 of the paper were obtained from fitting the linear
mixed model (1) in Section 2 of the paper, in SAS (version 6.12). This can be done using
the following program :
/* Input of the data */
data rats;
infile 'a:\rats.dat'; /* specify path ! */
input treat $ rat age y;
run;
/* Logarithmic transformation of the time-axis */
data rats;
set rats;
t = log(1+(age-45)/10);
run;
/* Fitting the linear mixed model */
proc mixed data = test covtest ;
class treat rat;
model y = treat*t / solution;
random intercept / subject=rat;
contrast 'treatment effect' treat*t 1 -1 0, treat*t 1 0 -1;
run;
Contact address:
Geert Verbeke
Biostatistical Centre
School of Public Health
University of Leuven
Kapucijnenvoer 35
B-3000 Leuven, Belgium
Tel : 32 / 16 / 33 68 91
Fax : 32 / 16 / 33 69 00
E-mail : geert.verbeke@med.kuleuven.ac.be
Website : http://www.med.kuleuven.ac.be/biostat/
|
