I now have thirty dollars in library fines. And I'm buying more books than usual too. I'm starting to think that climbing this particular mountain may wind up being expensive -- despite the lack of preparation entailed or the need for special protective gear (unlike, for example, climbing the real Everest). Luckily, I have some great sherpas in the form of all the people on amazon's top reviewers discussion list.
Also, I recently made up a series of assignments for a stats class I'm teaching which require them to:
1. randomly sample a selection of Amazon top 1000 reviewers
2. come up with hypotheses about the variables entailed in moving up the chart
3. figure out how they are going to operationalize and measure the variables (i.e. willingness to like any book)
4. perform frequency distributions
5. make cross-tabs
6. make up two samples(one for top reviewers under the old system and one under the new) and then compare the median, mode and mean for various qualities (i.e. average number of stars given, average grade level on which the review is written) -- and then perform statistical tests to describe the differences between the two groups
7. run bivariate correlations
8. make a regression model
and so forth. Hopefully by December 2009, I will have more data than any of you would ever want on what it REALLY takes to get to the top of this list. I look at this as a win-win situation -- my grad students learn stats, and I get some potentially useful advice. Ah, harnessing the power of the student. One of the real joys of being a professor.
Subscribe to:
Post Comments (Atom)
I would certainly be interested in learning the results, especially if I turn out to be oneo f the random reviewers sampled. :)
ReplyDeleteThat is an idea of genius.
ReplyDelete