Post hoc analyses for a 3-way interaction in an ANCOVA


I am doing a repeated measures with a covariant. a 3-way interaction (gender-group-memory) emerged as significant. I understand that when doing a repeated measure with a covariant the "post hoc" option for SPSS is not applicable. I selected "compare means" in the option window, but I find it difficult to interpret the results. Is it looking at the confidence interval in the estimated marginal means section and seeing if the confidence interval don't overlap? I'm adding an output for example:

1. New group * Gender * memory
Measure: MEASURE_1
New group Gender memory Mean Std. Error 95% Confidence Interval
Lower Bound Upper Bound
Low female 1 4.247a .273 3.703 4.791
2 3.817a .283 3.253 4.381
male 1 5.054a .320 4.417 5.692
2 3.570a .332 2.909 4.231
High female 1 4.790a .310 4.173 5.407
2 4.001a .321 3.360 4.641
male 1 3.861a .366 3.132 4.590
2 3.689a .380 2.933 4.446
a Covariates appearing in the model are evaluated at the following values: dep = 6.1585.

Also: any help on how to build the 3-way interaction at the same graph using SPSS will be highly appreciated.

Thank you very much!