Comparing similar variables for two groups who answered the variables based on similar pictures with small changes

#1
Hello,
I want to measure differences between two groups who answered similar questions based on similar pictures of food.
Each group were shown 8 pictures of food in its packaging, but one of the groups were exposed to the pictures of the 8 food products where the protein content of the product had been photoshopped onto the packaging as a label. A radomization variable was used to ensure, that only half of the participants saw the 8 food pictures with protein labels and the other group saw the same pictures without protein labels on.

Under all 8 pictures, for both groups, the respondents were asked about (based on 7 point bipolar scales):
1) Attitude (how much they like the product)
2) Intention to buy the product
3) How healthy they perceived the product to be.

Each picture and the 3 questions were displayed on an individuel page in the survey, thus 8 pages were shown with a picture of a product and the questions, thus both groups answered the same 3 questions for each product individually.

Now, what I want to test is whether the participants who saw the 8 pictures with protein labels on and then answered the questions about attitude, intention to buy and health perception significantly different compared to how the other group, who saw the 8 pictures without protein content displayed on the packaging answered.

I want to test/compare both how each group answered regarding attitude, intention to buy and health perception for each product individually, and how each group answered in terms of overall attitude, intention to buy and health perception regarding all the products combined and whether those who saw products with protein labels answered different from those who saw the pictures without protein labels.

Overall i want to test whether adding protein content as a label to food packaging increases the consumers attitude, intention to buy and health perception in general and for specific products.

How can this be tested? And is it possible to e.g. add another variable; e.g. if being more biased/less biased towards protein affects how each group answered the questions?

I hope you can help me out, because I am very inexperienced with statistics.
OBS: im using SPSS to test the data.

Kind regards,
Nicklas.