« bibliographic & ironical | Main | this time »
October 19, 2006
how T.H. finds out what writing does and how it does it
what is it good for?
bazerman, from Introduction (distracting/irritating overuse of boldfacing removed): "For researchers, discourse analysis provides a means of examining communicative practice so as to uncover signs of social identities, institutions, and norms as well as the means by which these social formations are established, negotiated, enacted, and changed through communicative practice. For teachers of writing in colleges and schools, discourse analysis provides ways of going beyond the simple and perhaps confusing terminology of our everyday language for texts and writing" (3).
following directions
T.H. says "content analysis...involves a synergistic blending of quantitative data gathering and qualitative analysis....Some [studies] may start with a proposition and use data-gathering in a deductive manner to confirm or disconfirm the proposition, whereas others may be more exploratory, using qualitative analysis in an inductive, flexible manner" (16), but in general they should all follow his 6-step (recursive, cyclical, rarely-this-linear) process:
1. Pose a Research Question
2. Define the Appropriate Construct(s)
3. Select an Appropriate Text or Body of Texts as the Study Corpus
4. Determine Appropriate Units of Analysis (Text Features), Using Multiple Raters if Possible
5. Gather Data
6. Interpret the Findings
elaborations
on "construct(s)": "...a researcher who hypothesizes that more mature students tend to be more rhetorically sensitive than less mature students is implicitly working with two constructs, 'student maturity' and 'rhetorical sensitivity.' these constructs will have to be defined, however, in a way that can be converted into measurable units. for example, 'student maturity' might be defined in terms of age, years living on one's own, or some other variable; 'rhetorical sensitivity' might be defined in terms of attention to purpose, attention to audience, and/or other parameters" (17). (so i guess i shouldn't ask how, exactly, "attention to purpose" is "measurable"?)
on coding: make it up as you (start to) go along, using your first few samples of material to find/make categories that can be applied to later samples.
limitation #1 as reason #1 for arguing that the project shouldn't stop here: "coding in quantitative analysis is too surface-based and thereby lacks validity...[as it] is said to valorize the textual artifact in unreflective fashion, ignoring the reader's or writer's engagement with the text (anderson 1973). as van dikj (1997b) put it, content analysis is 'a method which in fact has less to do with meaning than with the more observable aspects--mostly words--of discourse' (p. 9)" (26). <--why we have to also ask the writers.
Posted by ttobryan at October 19, 2006 12:28 PM