no title--surveys and pluralism

The Lauer and Asher text didn't say much that an intro. to statistics course doesn't say, but it was a nice concise way to organize the statistical side of surveying. I would definitely refer back to this chapter before going into any survey research, but I'd hit up a stats textbook and a knowledgeable colleague too. I had one question about table 4-1 on page 58. For the Texas researchers' example on page 59, wouldn't we be able to plug in 42% and 48% as p and q to find the confidence interval? Should we always use the "worst case" scenario of a 50% p value? If we would plug in the actual p and q values, does this eliminate the need to use the corrections in table 4-3? I am feeling sick right now, and have been taking day/nyquil for two days, but this chapter wasn't the easiest to figure out. When L&A say that "for the Texas research, the use of the correction Table 4-2 was unnecessary, because the sample size did not approach the population," I wonder how big the population of writing directors actually is compared to the 127 who completed the survey. I'm guess there are over 2,540 college writing program directors in the country, so they are under the 5% required for a correction, but couldn't L&A comment on this for clarity?

I was disappointed to not read much about creating unbiased questions. L&A basically referred us to the references and gave us a list of problems that can befall writers of questionaires. I should interject here that I personally hate questionaires. I always try to cheat at least a little bit on surveys to skew the researchers' findings. At West Perry, where I teach, we have to survey our students each semester to comply with a technology grant we received as a school. The kids ask me, "What can I put down so we get better computers?" Of course, I don't know exactly what criteria the state uses, but my point is that almost everyone is savvy and skeptical about surveying.

Kirsch's text is dense! I wish I were feeling more physically up to the challenge today **cough cough, puppy dog face. I'll just dig into a couple of the things that caught my eye.
Kirsch questions Berieiter and Scardamalia's statement "that 'a holistic ideology ... poses an actual threat to writing research [because of its] opposition to any research (or instruction) that deals with less than the full act of writing carried out under natural conditions'" (250). I have to agree with Kirsch here. Experimental research is limited in the types of questions it can answer and could potentially produce a type of writing founded in and limited by experimental research. Holistic approaches to composition are not necessarily less valid just because they are difficult to experiment with.

After Kirsch discusses the arguments for and against a single methodology in composition studies, she states, "The question, then, is whether scholars are willing to break from a relatively rigid adherence to their disciplinary orientation in order to entertain alternative methodologies" (256). This quote made me think of the common secondary-ed imperative that "all teachers teach writing." The parallel is far from perfect, but an important point is revealed in the comparison. The writing demands of each discipline are different, and all disciplines can contribute to composition studies. As far as whether we should have a pluralist methodology for composition studies, Kirsch's arguments for cultural equality (251) and Newkirk's reminder of the reality of research processes used in the field today make a pretty strong case for pluralism. To me the "question is" whether researchers should or must take a pluralistic route and become jacks of all trades, or whether researchers can share the effort--playing nice--as a community.

I really liked the Kirsch chapter's change in direction to an example of a forward thinking piece of research that involved the input of subjects as participants with clearly articulated roles in the research. The purpose of composition studies is to produce better writers, right? How can anyone oppose including the subjects? What Kirsch's chapter highlights most for me is that methodological pluralism is the only way to seek a body of knowledge about composition that is fair to diverse subjects (I mean "people," but subjects as disciplines applies as well).

I'm pooped. Ttfn.

Survey Says...

Of the two readings this week, I was more drawn to the Kirsch article. I'll admit, after reading the Lauer and Asher text detail the various definitions, methodologies and strategies surrounding surveying, reading the phonebook probably would have been a nice break, but I liked several aspects of the article. The idea of pluralism within research methods is something that we've talked about fairly often throughout the semester as a means of bumping up our research to give it more clout, especially when looking at in the context of case studies and ethnographies. Usually, the sort of thing that they're suggesting is something that will produce a quantitative result, something that can be carried by numbers without much interpretation. In this weeks readings we got both ends of the spectrum: we got a big pile of numbers and then a big pile of data that begged to be interpreted.

What struck me about the Lauer and Asher text detailing the various sampling and survey information was that in order to make these numbers not only convincing but important, inferences and research had to be done painstakingly beforehand. I naturally compared the work between crunching a set of numbers and writing and rewriting survey questions with that of actually interacting with people and I began to wonder which could tell you more about a set. There are positive and negatives to both methods: qualitative could have some discrepancy with the observations and quantitative could have discrepancies with the representative results of a set. Constructing survey's to back up your qualitative studies though was one thing that I wasn't sure was totally addressed in the chapter. It s made very clear that survey's don't allow for very many inferences at all because they are solid numbers to represent a larger population, but what would you do with survey's received from your control group from your observation experiments? What kind of inferences could you possibly make about that information that would pertain to your numbers.

I guess what ultimately bothers me about the various ideas of numerics used throughout this chapter is that by taking these various random samples, you are supposed to some how take into account the methods of a range of different people. I wonder if I'm alone in being frightened by the thought that some random "integer" next to me could be selected to represent where I fall within this spectrum without me ever being consulted. It helps me see some of the gray area in the numerics as well, and depending on how charts and numbers are depicted, they can tell a story that may not be any better than a case study or ethnography.