Statistical approach to analyzing style
I'm doing a research comparing news texts of two English-language press agencies, Reuters and STA, from a stylistical point of view.
I have a corpus of both two agencies, and would like to perform a sort of statistics analysis to give me a clue about the style used.
Could you give me some hints what to look for and how to do it?
I'm focusing especially on the difference between L1 agency (Reuters) and L2 (STA) and how the style differs due to authors's L1 background.
I don't want to look into the contents of the texts, but rather in style only.
Would appreciate any help or hints you can give me.