--Originally published at FLN Hub – nodes
I gave a test a week or so ago. When I scored the test the first time, I didn't think I made my key correctly. So, I double checked it and re-scored. Same result.
I immediately implicated myself in the low scores. My test may have been too hard. Perhaps I included things that weren't taught. I went back and went over the entire test, item by item, and matched each to a set of notes, a lab, and the state standard I marked on my plans.
I taught everything on the test in at least three different modes. I did my job.
So, I turned to statistics to help me out. I ran item analysis on each question, both using Biserial values and Chronbach's alpha test. It got nerdy fast.
With the exception of a few questions with low Biserial values (which I'd already identified as problematic, this just confirms) the test is reliable. All of the questions were within the accepted range of variance.
This confirms that the test was, in fact, valid and provided reliable results.
I can't force students to engage. I can provide opportunities to engage and I can prod them that way, but at 15-18 years old, they need to make their own decision. Some classes have made that decision.
Two classes in particular are beginning to thrive. Two don't know what they want, and two are continuing to really, really frustrating years.
I think it all comes down to the fact that students need to realize that everything they do as an individual affects every other person. That's what a community is. We're all affected, including me.
How do you teach that?