2x2x2 (?) experimental design analysis


I conducted an experimental study to see the effects of race (factor 1) and communication source (factor 2) on message viewers' purchase intention, and the below are 5 types of stimulus conditions. The images were added to textual message, which was identical throughout all five conditions.

1. image of white expert
2. image of white peer
3. image of black expert
4. image of black peer
5. control (no addition of an image)

I also want to see if the message viewers' race (i.e., white & black, factor 3) affects how they process these 5 types of messages. In this case, is the design of the experiment 2 (race: white vs. black) x 2 (source: expert vs. peers) x 2 (viewers' race: white vs. black) as race and source within-subject factor an viewer's race as a between subject factor? (then what about the control condition? - I also would like to know whether addition of images (regardless of types) are more effective than no addition of an image.

To analyze my data, I think I can run two ANOVAs: (1) a one-way ANOVA comparing the control group with the overall treatment average and (2) a 2x2x2 ANOVA comparing the two factors (2 within, and 1 between) and their interactions. But the results will not allow me to simultaneously compare the three factors with the control group.

Would you let me know 1) a proper analysis procedure, and 2) whether the aforementioned design is inaccurate/flawed?
Thank you for your help in advance, and please let me know if you need any additional information for this question.
Last edited: