Next month I'm giving a virtual PD to a group of Indiana teachers on data collection, so I've been thinking about what it takes to start the school year on the right foot.
It comes down to one question: how do I determine which of my students are at risk?
If your school does universal screening, the results are supposed to answer that question.
But there are two tools that claim to do universal screening, and one is much better than the other.
The first (better) kind of screening
One kind of universal screening is given one-on-one. Every student takes it, one at a time, sitting with you. Each measure takes about a minute, and the whole assessment takes 10 minutes or fewer per child. You read the directions exactly as they are written, and you time it.
Every student gets the same items, in the same order, with the same directions. That sameness is what lets you compare one child's score to another child's, and to the benchmark goal for that grade.
This kind of screening helps answer the question, "Is this child at risk, or are they on track to do well at reading?"
Examples of this kind of screener include Acadience, Dibels, aimsweb, FastBridge, and easyCBM.
The other (not as good) kind of screening
The other kind of screener is a computer-adaptive test. STAR, i-Ready, and MAP are ones you may have seen. Students take the test on a computer, and the test adjusts as it goes, giving harder questions after correct answers and easier ones after wrong ones.
There are a few reasons why computer-adaptive tests may not be the best choice for screening.
PROBLEM #1 ---> Lack of standardization. Students are not all answering the same questions, so their scores are not standardized, and standardization is what makes one child's score comparable to another's.
Two students can come back with the same number having taken two different tests, which means the score cannot tell you they have the same gaps.
PROBLEM #2 ---> On a computer-adaptive test, a student picks from options on a screen instead of actually producing the answer.
PROBLEM #3---> Computer-adaptive results are hard to interpret.
PROBLEM #4 ---> Worst of all, they tend to underestimate risk.
So if you are required to give a computer-adaptive test, ask what score a student needs, at each grade and at each time of year, to be predicted to do well on the state test at the end of the year. If the test publisher cannot give you a reliable predictive score, the test cannot tell you who is at risk.
In my newest video, I go through what a good screener actually measures, and what to do once the scores come back.
P.S. This month I'm doing something I almost never do: hosting a free, live webinar. I'll walk you through the four steps to reach your struggling readers, so you can catch things early this school year. Come with your questions; I'll be there live to answer them! Sign up here.