Nashville education blogger TC Weber talks about testing (and a lot of other things) in his latest post.
Specifically, he talks about the release of data on TNReady tests and the comparisons being made to previous TCAP tests.
Keep in mind: We didn’t have a complete administration of TNReady in 2016. Which means the 2017 test was the first year for TNReady. It also means the comparisons being made are based on different tests taken two years ago. So, you have analysis of 5th grade results and “growth” on TNReady being made in comparison to 3rd grade results on TCAP.
It’s apples and oranges.
Here’s what TC has to say:
Everybody at the TNDOE is talking about how much better this new test is and how it is more rigorous than our previous tests. But when I ask, how do we know that? I get no clear answer. It seems to me that the reasoning is, it’s a better and harder test because it’s closer aligned with standards and more kids failed.
Let’s approach this in a different manner though. Say I annually run a 5k race and each year my timing goes up a little bit, so I’m feeling like I want something different. After year 5 I change to a 10k race. My time for that race is substantially lower. What conclusions can I draw from that difference in time? Am I really not that good a 5k runner? Is the course really that much harder than the 5k I was running? Is my training off? Am I not that good a runner?
I’d say there are very few conclusions, based on comparing the results between my 5k and my 10k time, that can be drawn. It could be that the length of the course was a bigger adjustment than anticipated. It could be that conditions were worse on the day I ran the 10k vs the 5k. It could be that one course was flatter and one was hillier. A kid could be good at bubble in questions but not write ins. How do we know that improvement isn’t contingent just on familiarity with the course? Or the test?I know people will argue that we should all be training to run hills instead of a flat races. But does running hills well really indicate that I am a better runner? Terrain is just another variable. My liberal arts education always explained to me that in order to get the most accurate measurement possible you need to remove as many of the variables as possible.One year of data is not a real indication of anything other than, kid’s are not very good at taking this test. In order to draw any meaningful conclusions, you would have to have a set of data that you could analyze for trends. Simply taking a 10k race and comparing it’s results to a 5k race’s results, just because both are races, is not a valid means to draw conclusions about a runners abilities. The same holds true for students and testing.