Hi there,
I have run a linear mixed model (using MIXED) with 4 categorical factors (gender, time, condition,
trial: each with two levels) and 1 continuous factor (brain thickness). There is a significant 3-way
interaction between time, condition and brain thickness. I am struggling to figure out the TEST
subcommand in order to tease apart this interaction. I have plotted the results, using EMMEANS to
obtain estimated means for condition X time for low and high values of the continuous factor. I have
interpreted the F tests from the COMPARE function as essentially test of simple slopes when specified
to a given value of the continuous variable (Is this correct?). All my tests of simple slopes are
significant (indicating an effect of time across both levels of condition and high/low brain thickness),
but I am more interested in conducting difference in slopes to fully explain the interaction...
Specifically, is the simple slope of "condition 1 X time" different for low and high levels of brain
thickness. Is this possible to do using the TEST subcommand? I would greatly appreciate any help
with this!
Thank you,
Nandi
=====================
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
