Principal component analysis to group items for two different countries

Hi guys,

We are two master thesis students doing our thesis about craft beer. We are performing a principal component analysis to group items from a survey we carried out in relation to a theoretical concept "Country of Origin" and how that affects attitudes towards craft beer from two countries. Our survey asked the respondents questions about their opinions towards Italian people, products, and beer and American people, products and beer. We then performed a PCA for the attitude towards Italian people, products and beer and made this into 9 components, but when we do the same for attitude towards American people, products and beer, the items group into components in a slightly different way.

As we want to compare people's responses for both countries, how would you suggest we proceed from here?
1) Should we have components made up of different items for each country?
2) Should we drop items until the components are made up of the same items
3) Any other ideas?

Thanks so much :)