Authorial ignorance. It doesn't test the scope of knowledge of the author
Authorial context. Similar texts produced in the similar time.
General source context.
Survival rate. Did only controversial idiocy survive in the libraries. Did this survive monastery burning because it was being used to insulate a bamboo wall?
Reception context at time of authoring. Did everyone know that "relocated to the East" meant burnt in an execution camp.
Internal purposes. Documents written for internal consumption (beer barrel receipts) are often more trustworthy about what they lie about than documents written for external consumption (all our monks are sober, the broadsheet bill)
Transmission of the text. Forgery, reinvention, copying, selection for survival
Language of text. SMSes from protestors in London are not going to sound like, "I have found a great and not very well protected big screen television store here at grid ref follow."
Internal consistency, is this actually a single text? Is it a text of texts, is it a chapbook, Fred's favourite quotes, a hypertext of biblical references?
Modern presentation of the text, is this presented in an undergraduate or highschool textbook as illustrative? Why? Which sources were excluded? Is it actually representative, or atypical?