ABSTRACT
This chapter discusses how the findings related to the research questions set, noting that as a by-product of their analysis, some of the authors actually answered a question which was not given to them. It notes overlaps in findings between the corpus-driven approaches. For example, both the lexical bundle and the multi-dimensional analyses noted the highly interactive aspect of the Q+forums, with the Family & Relationships forums being most involved. Also, the keyword and key semantic categories approaches found that the UK had more focus around politeness and mentions of US politics, as well as India using more acronyms. These two approaches also had similar findings relating to religion. Tony McEnery described keyword analysis as a useful method for comparing corpora in order to discover what makes them distinctive or similar. The usefulness of corpus-driven methods, such as keyword analysis, depends on contextualization and interpretation from the researcher.
