Biostatistics 537: Longitudinal Data Analysis - Fall 2008
Problem Set 1 - Due: September 11, 2008
There are two data sets for this problem set: vt.dat and box.dat. Both of these can be
downloaded at the class website http://www.uic.edu/classes/bstt/bstt513/ (towards the bottom
of the page).
1. The data (in vt.dat) consist of a sample of 22 subjects from a study of affective facial
expressions (Vasey and Thayer, 1987). In this study several pieces of music were played to
each subject in an attempt to elicit selected affected states. Trial 1 was a baseline, relaxing
music condition. Trial 2 was designed to produce positive effects, and trial 3 was designed
to produce agitation. Each trial lasted 90 seconds and the response variable at each trial
was the mean electromyographic (EMG) amplitude (µV) from the left brow region. The
variable order in the datafile is subject number followed by the three trial EMG
measurements. Consider the question of whether each of the two non-baseline trials differed
from the baseline trial in EMG. For this, choose an appropriate set of contrasts that relate
to this question. Posit a reasonable model and interpret the results. In doing this, you
should choose between the univariate and multivariate repeated measures analysis of
variance approaches, and give reasons for the approach you choose. If you choose the
univariate repeated measures model, estimate the intraclass correlation and comment on its
magnitude. Also, to better normalize the outcome variable use a log transformation. Here’s
some SAS code that will help with this problem:
FILENAME in1 ’C:\vt.dat’;
TITLE1 ’Vasey & Thayer Data - Left Brow EMG amplitudes in 22 subjects’;
/* input the data in multivariate format */
DATA MultData; infile in1;
INPUT subject 2. (trial1 trial2 trial3) (4.) ;
ltrial1 = log(trial1);
ltrial2 = log(trial2);
ltrial3 = log(trial3);
PROC MEANS;
RUN;
/* set up the data in univariate format */
DATA UNIDATA;
SET MultData;
ARRAY T(3) ltrial1-ltrial3;
DO TRIAL = 1 TO 3;
logEMG = t(TRIAL);
OUTPUT;
END;
DROP ltrial1-ltrial3;
RUN;
PROC MEANS;
RUN:
