Question about contrasts and covariate in cross-design paradigm

#1
Hello everyone,

I need some expert advise for the following stats question.
I am evaluating data for a psychological learning experiment with a cross-design. Two groups of participants (A and B) are exposed to either stimulus set A or B and subsequently tested with stimuli from the joint set A and B (single binary response for each stimulus). The hypothesis is that participants are more likely to endorse stimuli that they were exposed to rather than the other ones. Further, the nature of the stimulus materials entails a distinction of 5 different stimulus types ("features") which are grouped into 2 kinds (feature 1-3 and 4-5). The goal of the analysis is to test whether (a) there is a general learning effect at all (participants more likely to endorse stimuli they were exposed to), (b) which of the 5 features are learned or not, (c) whether there is a general difference between the learning performance for the two kinds of features (1-3) and (4-5). In addition there is a between-subjects difference in prior experience.

Now, my current approach is to use an ANOVA with group * experience * feature type (composed of the 5 features) to assess question (a) by a group * feature type interaction. Further (c) and (b) should be tested through self-defined post-hoc contrasts.
In this context I have two questions:

(1) Is this the best way to carry out the analysis?
(2) Based on feedback by a colleague's, there may be a potential impact through a prior response bias. This is represented by two 5-value vectors matching the 5 features for each A and B stimuli and represents a potential bias that is identical for all participants of group A or B respectively. How can this be incorporated in the analysis? I am puzzled here since from my understanding it seems that the standard covariate variable approach does not work in this case. Consulting my textbooks didn't help either.
It would be great if you could explain to me how to approach the covariate in this case.
It would be even greater if you could help me how to do the contrast/covariate in SPSS.

Thank you very much!

All best wishes,

M.