Hello!
My name is Daniel Arvidsson and i perform methodological studies in physical activity. I have a question concerning how to create a specific dataset in SPSS for mixed model analyses with repeated measures. The aim is to validate an activity monitor (method-test)
against a reference (method-ref) for 10 different activities (activity). In addition, the test method has two different algorithms (algorithm 1 and 2) versions for the output. Hence, the 3-way mixed model is:
Method x Algorithm x Activity
To assess the fixed effect of each of these factors I create a SPSS dataset in long form, denominating three columns as Activity (1-10), Method (1 and 2) and Algorithm (1 and 2), respectively. Two other columns are subject ID and Output (energy expenditure,
EE, which is the dependent variable to be compared). This setup results in, that for each activity, each subject will have three rows with outputs, as method-ref provides one row and method-test provides two rows is it has two algorithms to be evaluated against
the method-ref. The columns for activity 1 look like for subject 1:
Activity Method
Algorithm ID
Output
1 1 ?
1 4.33
1 2
1 1
3.56
1 2
2 1
2.39
My question is, what figure do I put for method 1 in the algorithm column, as it is the reference method to compare algorithm 1 and 2 against?
Best regards,
Daniel Arvidsson
Lund University
Sweden
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
