Hello Bieke Helwegen. I am struggling to follow your explanation of the design, but I think you are saying that it is not a full factorial design. Is that correct? If so, you may be much better off using the MIXED command with a LONG file format rather than the more conventional GLM REPEATED MEASURES approach. This allows you to have complete control over what terms are in (or not in) the model, whereas GLM forces certain terms to be in the model whether you want them or not. MIXED with a LONG file format also allows for missing values on repeated measures factors that GLM does not.

See the following UCLA pages for examples of how to use MIXED for repeated measures. HTH.

• https://stats.oarc.ucla.edu/spss/seminars/repeated-measures/
• https://stats.oarc.ucla.edu/spss/library/comp_repeat_spss/

Yes, you can perform a three-way mixed ANOVA with repeated measures in SPSS. SPSS provides the necessary procedures to conduct this analysis. Here's how you can do it:

SPSS will generate the output, which includes information about the main effects and interaction effects of your factors, as well as tests of significance (p-values), effect sizes, and other relevant statistics. The output will also provide information about the assumption of sphericity and corrections if applicable.

Please note that for a three-way mixed ANOVA with repeated measures, it is important to consider the assumptions of the analysis, such as sphericity, normality, and homogeneity of variance. SPSS provides options to assess these assumptions within the output and allows for corrections or adjustments if needed.

It is recommended to consult a statistical expert or refer to additional resources for a comprehensive understanding and interpretation of the results from a three-way mixed ANOVA in SPSS

Thanks everyone, this helps a lot!