MA thesis

Hello,

I'm Shama from Tunisia. I'm very interested in code switching. In fact, i'm working on my MA thesis and it's about code switching and borrowing on social networking sites (specifically on facebook).
I'm still trying to improve the methodology part though and I was wondering whether you guys can give me any feedback.
My study will focus on analyzing switching code between 3 lges (French, Tunisian arabic and English). But the problem is that not everyone in Tunisia speaks english so if i analyzed the participants statuses or msg, their lge choice will depend on whether the person their talking to also speaks/understands English or not. and that would jeopardize the sudy's results.
so i thought about creating a page on facebook where i can add all of my participants who all speak the 3 lges in question and each time propose a topic that they will discuss and then i analyze their comments (instead of analyzing statuses and posts).
My question is do you think that such an approach can allow me to get data that appropriately represent lge use on facebook????

i would really appreciate any ideas/feedback/criticism.

thank you