GLM Simple effects post hoc

[Kam T]

My question concerns pair-wise comparisons in SPSS output. I am running a full
factorial 4-by-2 repeated measures GLM and for my comparisons I'd like to see
e.g. 3-way post hoc comparisons for two of my factors. When I enter
/EMMEANS=TABLES(fac1*fac2*fac3)COMPARE(fac3)ADF(bonferroni)
COMPARE(fac2)ADF(bonferroni)
SPSS outputs only comparisons for fac3.
When I enter
/EMMEANS=TABLES(fac1*fac2*fac3)COMPARE(fac3)ADF(bonferroni)
/EMMEANS=TABLES(fac1*fac2*fac3)COMPARE(fac2)ADF(bonferroni)
SPSS complains about redundant command and skips everything after first
/EMMEANS statement.
Would someone please explain to me why after 18 versions, arriving at various
combinations of post-hoc comparisons for simple effects is so elusive? Is this
a statistical violation that I have not understood? And if not, how does a
simpleton do this simply? Thanks in advance.

=====================
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

[Bruce Weaver]

A 4 x 2 model has only two factors, one with 4 levels, and another with 2 levels.  So I'm a but confused about how there can be a 3-way interaction, such as you show in your EMMEANS lines.  It might help if you post the complete syntax for your GLM command.

K T-4 wrote

My question concerns pair-wise comparisons in SPSS output. I am running a full
factorial 4-by-2 repeated measures GLM and for my comparisons I'd like to see
e.g. 3-way post hoc comparisons for two of my factors. When I enter
/EMMEANS=TABLES(fac1*fac2*fac3)COMPARE(fac3)ADF(bonferroni)
COMPARE(fac2)ADF(bonferroni)
SPSS outputs only comparisons for fac3.
When I enter
/EMMEANS=TABLES(fac1*fac2*fac3)COMPARE(fac3)ADF(bonferroni)
/EMMEANS=TABLES(fac1*fac2*fac3)COMPARE(fac2)ADF(bonferroni)
SPSS complains about redundant command and skips everything after first
/EMMEANS statement.
Would someone please explain to me why after 18 versions, arriving at various
combinations of post-hoc comparisons for simple effects is so elusive? Is this
a statistical violation that I have not understood? And if not, how does a
simpleton do this simply? Thanks in advance.

=====================
To manage your subscription to SPSSX-L, send a message to
[e-mail] (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