Monday, February 23, 2009

"And researchers need to...

avoid the temptation to manipulate and interpret the data in order to find what they were looking for" (Kirsch and Sullivan, 238). I'd wager that's the toughest part of research. After you go through all the Chi Square analysis...or the F (or T) test...or Spearman's rho (whatever that is), the bottom line is that you need to properly represent your results. It must be totally depressing to go through all that work and then find out that what you'd hoped you'd find is not there at all.

I really had to force myself through the Lauer and Asher chapter this week. Much of it was over my head - or it felt like it anyway. Parts were very familiar, some vaguely familiar, and much of it made my brain feel like exploding. I can talk all day about standard scores and standard deviations. Even percentiles. I really liked the normal curve on p. 91 that even included letter grades on the normal curve. But I wonder how many classroom grades really do fall on the normal curve like that- I don't map mine out that way. In fact I don't pay any attention to how many students earn A's or B's. Students earn what they earn when it comes to grades (which we all know are far from a perfect science!)

For a big picture learner like me, the illustrations on p. 100 and 101 really helped me grasp the point about interaction and no interaction. And I wish I'd read Beach's explanation of data analysis on p. 225 before I tackled L/A's chart on page 90. I probably wouldn't have had to write nearly as many margin notes!

One powerful point was the importance of a thorough literature review before the study begins. Bottom line - it can help you find unanswered questions that might be the exact ones you're looking to research. Your might find contacts and additional leads. Sometimes students I work with in our writing center question why they have to do preliminary research before they formulate a thesis. Now I've got some new ammunition to torture them with.

All told, quantitative descriptive studies might be the best of both worlds. They allow the researcher to discuss ideas and thoughts in words yet also back it up with hard numbers and data that speaks on its own. I just hope that, when I attempt this kind of study, I have a really powerful computer next to me, a knowledgeable mentor, and a big bottle of wine.

No comments: