Dear allstat,
In a pair comparison experiment, two different groups of people did the
test with the same factors and methods. Test involves with a specific image
in different conditions and variables. Variables are either ordinal or
continues. People in group A have seen image A and people in group B have
seen image B- which is a totally different image. Each person in the group
has done the test twice with two different methods (the general method of
pair comparison is constant but the sequences are different). I’d like to
do a significant testing between the results from different levels of
variables and also correlation tests on this data and report the overall
trend regardless of methods and images they have seen. Now I am wondering
if it is possible to do the tests by combining the whole data and do the
tests, if not, do you have any suggestion to get an overall results from
the data, because the data collected from each group is very limited (15
subjects in each group)for a reliable statistic test.
Many many thanks for your time and advice in advance,
Navaz
-------
Navaz Davoudian
PhD Candidate
School of Architecture, University of Sheffield
Western Bank,
Sheffield
Tel: 0114-2220365
Email: [log in to unmask]
|