I currently have a setup, for which I want do do an ANOVA in spss. In my model I need to include four factors (I'll call them factor1 to factor4) each with two possible states (true and false). However not all combinations are possible. To be more precise, factor4 can only be true if one of factor1 or factor2 or both are true.

Hence in total I only have 14 possible combinations instead of the total 16 which would result from a full combination of all four factors. Is it possible to us a repeated measures ANOVA with such data, or do I have to use something else, since factor4 cannot be varied independently?

I tried to construct a repeated-measures ANOVA for this in the SPSS GUI, but I cannot click on OK, unless I give all 16 combinations, which is impossible given the setup I have. Is this somehow possible to program in SPSS, maybe by entering the syntax directly?

Or maybe is there a way I can add the missing data, for example by adding the means of some other variable for the missing combinations, without changing the final outcome of the model?


I just also tried to add the missing combinations with empty variables and then run the model on this completed set of variables. However this also does not work, since in that case SPSS just complains, that there are no valid cases in the data (since all are missing those two combinations).


It appears you're trying to run a factorial ANOVA but your predictors are not all crossed with each other, this is impossible to do. You would need to provide much more information on the design of your study to give you the best advice but you typically have one of two options here.

You can analyze all of the factors that are crossed, perhaps factor 1-3, and then factor 3-4 in two separate ANOVAs. You might be able to reach good conclusions there. It's possible you don't even really want to say anything about the interaction between 1 and 4 (for example). Or, you could just make one variable that allows you to analyze the data in a one way ANOVA. Without more details I wouldn't unequivocally recommend this. You're going to have to go through and see if there's something you need to extract from the results that cannot be addressed through the former method and whether such a design makes sense. It would mean you have one factor with levels like TTTT, FTTT, TFTT... FFFF.

  • $\begingroup$ Thanks, most likely it will not be necessary to analyze all interactions, so the analysis will focus on just a couple of the possible interactions. Mostly the third factor and it's interactions with the other factors are interesting for the study. So I will just create new variables by averaging all the cases with the appropriate settings of the other factors and then analyze this interaction in isolation. $\endgroup$ – LiKao May 7 '12 at 9:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.