Hi. You need variables that are paired. It may be possible to go from what you have to what you need by using the CASESTOVARS command (as long as the variable you want to "pair" has a unique case identifier, like the "id" variable in this example):
dataset close all.
data list free /group score id (3F2).
begin data.
1 15 1
1 13 2
1 16 3
1 17 4
1 13 5
2 12 1
2 13 2
2 12 3
2 10 4
2 10 5
end data.
SORT CASES BY id group.
CASESTOVARS /ID=id /GROUPBY=VARIABLE.
LIST.
------------------------------
Rick Marcantonio
Quality Assurance
IBM
------------------------------
Original Message:
Sent: Mon May 01, 2023 03:28 PM
From: Megan Smith
Subject: Marginal Homogeneity test
Hi, is there a way to use a grouping variable (pre=1. post=2) here to pair the data or do I need a separate pre variable and post variable for each pair?
Thanks,