Does Nothing Matter?

No, this isn't a nihilistic posting or philosophical musing. I was looking at a lot of the quarter stats on student performance, and there's a lot of good news - I can see strong growth in grammar and vocabulary, and the writing seems stronger but I haven't created any good technical standards for tabulating that kind of information yet. It will be done, but not yet. It's pretty straightforward to do this using ACT/ISAT style rubric based grading, but I'm just not quite there yet.

With all this good news, though, the problem is the "nothing" that I refer to in the title above. Statistically, there seems to be "no reliable relationship" between some of the class tools that we're using to improve performance and the actual changes in performance. So while everybody is improving, increased utilization of resources doesn't seem to correlate with greater improvement.

Here are the stats:

So, 1st q means first quarter, total is total grade, imp is improvement. The first number column is a correlation and the second is covariance. It looks like there is very little relibility between any of these relationships, except that improvement from pretest to post-test is more closely tied to quia practice quizzes than anything else, but at .252, it's not very compelling. I don't really understand the covariance, but my guess is the relationship between hits on sharepoint and overall performance and improvement is pretty small. Is it possible that it takes 171 hits to account for one point of improvement?

There are, or course, contingencies to account for this. For example, I force AR students to practice in ways that are optional for Regular English students, so many of the students with generally weaker scores or slower improvement are using the resources more often than average. I also have small samples and a couple of wild outliers. But perhaps what is most important is that I feel compelled to try some new interventions to compare against them.

Well, anyway. I've gotta go. I've got work to do. Also, starting now I've got to log time that I work with people. That's all for now.