Showing posts with label STAAR. Show all posts
Showing posts with label STAAR. Show all posts

Sunday, September 27, 2015

The Big Data Derailment


Let me start this article off on the right foot:

I do like data.

I really do.

Not as much as a Triple Chocolate Andy's Frozen Custard, but I do like it. I sit around at home with cat and play with data for fun. (Frozen custard on the side would make it even better)

Now that I've admitted that I'm going to say something bold. Something that might make you curl your toes a little bit because it (along with some of my other views on literacy) go against the current of what's popular in testing, er, teaching these days.

Analyzing STAAR data is derailing literacy instruction in Texas.


Data Driven to Death


Data is good, but it's only good if you know how to read it, and the longer I watch trends of analysis the more I see data being completely misinterpreted all over. 

We're missing the forest for the trees, folks.

Like I mentioned in a previous article, Reading isn't like the other tests. Reading tests have two components: passages and questions.

If you're only analyzing the questions and using that to draw conclusions about quality of instruction, student ability, and overall literacy and direction of you or your school's direction, then you could be sending your train out the front window.

Why are so many schools struggling so much to catch up to current Reading standards and failing. There's a lot of reasons, because an entire state of people and schools is so vast and diverse, but I'd like to focus on just one aspect today: miscalculations.

Teachers and administrators are reading their SE hotspots correctly and even putting them into perspective by using blueprints and testing frequency charts. However, with this advent of testing culture and the need for everyone to be 'data driven' all the time, it has taught districts, schools, and teachers to think about literacy like this:


We study and study and study. We discuss and discuss and discuss. We plan and plan and plan. We do it all for the sake of being data driven! We want to fix those particular SE's! We must! We're doing it for our kids' sakes, our schools' sakes, our own sakes!

But it's not working.

If it were the real answer, for all the work we're putting in over four years worth of STAAR now, we ought to be seeing more results. Instead, more schools than ever are being classified by the state as "Improvement Required."

Why, oh, why?

Because, like I mentioned, we're missing the forest for the trees. Because of popular trends in educational jargon, we've focused on the tiles, the small pieces of what we think the problem is based on what a testing company paid by the state (and now fired by the state) told us our problem was.

But here's the thing.

The tiles aren't literacy. They're good for analysis to an extent, but they're not enough.

Why? Because literacy isn't tiles. It isn't isolated concepts or isolated skills. All those little things: character roles, conflict, main idea, inferencing, drawing conclusions, context clues--they all link up into what literacy actually is. If we want to improve, we have to zoom out to the big picture.

Literacy is a mosaic of cognitive skills. 



Those skills that the test doesn't open examine or quantify are what break our students if we don't address them in the classroom, even if we don't have numerical data from a test to quantify or justify their instruction. 

If you have a population of students who are not passing STAAR, don't only look at your SE's.

Also consider: 
  • Metacognition (Rigor): Can your students think about thinking?
  • Lexile/Guided Reading Levels: Can your students understand enough of the passage to have a chance of passing?
  • Do your students believe in themselves? Do they think they can be good readers?
  • Fluency: Do your students read fast enough to remember what they read?
  • Decoding: Hand in hand with fluency--Do your students need serious intervention?
  • Vocabulary: Do your students know what to do when they don't understand?
  • Background Knowledge: What are you doing to broaden the knowledge base of students whose families can't afford to take them places to gain new experiences?
  • Language Transition: What SIOP strategies are you using to help students find success in their second language?
  • Stamina: Are you giving your students time every day to read self-selected texts? Reading is a skill, not a content. You don't get better at doing it by learning about it.




Sunday, August 23, 2015

STAAR and Lexile Levels

Higher Order Thinking vs. Basic Comprehension


There's been a trend since STAAR rolled out that we have to push our students hard with higher order thinking questions if we want them to be successful. The general message has also been that the SE's that have weak percentiles meant weak teaching.

I've begun to seriously doubt all of that.


Horse of a Different Color


Let's face it. The STAAR Reading test (and the subject matter itself) is not like Math, Science, or Writing at all. 

Not at all.

Why is that so? 

Because the Reading test isn't just questions. It's composed of questions and passages. Therefore, when reviewing data about how your students did on any given Reading test, you're only looking at part of the picture. 

But if you really want to accelerate your students' growth, you have to look at both pieces.


Passages


So how do we talk about the passages? How do we analyze them? I think many people in the education training field have neglected this because it wouldn't seem that you could turn the passages into any kind of data or anything concrete or predictable. Yes, TEA gives us some idea of how many genres might be represented per grade level and how many passages there will be, but other than that, what else is there to say? You can't predict the topic. Unlike Math or Science standards where you can guess pretty close what the test will look like, Reading has remained rather nebulous. In an attempt to predict STAAR and do something to foster growth, districts and professional development people have gone question-crazy.

But what Genie used to tell me on Saturday mornings is true:




Great minds don't think alike. Great minds think for themselves.






Thinking for Ourselves


What if I told you that Reading passages from STAAR do follow a pattern, a pattern concrete enough that you can accurately predict your students' progress and likely performance based on a single set of numbers that is easy to track and easy to explain to small children?

It took many, many hours of research, investigation, and questioning, but here's what I've discovered: if students' Lexile level matches that of the average of the test, the student will pass.

Not only that, but I've calculated the average Lexile trends of the last three years and you can see what they are for free by downloading them here:






Using this data, you can set concrete, measurable, frequent goals and formative assessments as long as you have access to a program that provides Lexile level reports like Istation does. 


Back to Godzilla & King Kong


So what is more important? What's more powerful? Higher order thinking questions or basic comprehension?

Neither.

Because it's not a fight.

Harder questions are not the goal. Harder questions are the conduit through which the assessment discovers whether or not the student was able to comprehend the passage fully.

So what's the goal? The goal should be comprehension. The tools to test that comprehension should be higher order questions because they most accurately weed out who got it and who didn't. 

Where are test strategies in all this?

Test strategies are how teachers and students try to compensate for the difference between a student's reading level and the level of the passage he/she has to read and answer about. This can help them score a few points, and for bubble kids, it could help push them over into passing, but it's not the ultimate answer.

The ultimate answer is higher comprehension.

Set yourself and your students up for success.

Set comprehension as your goal and watch your students take control and grow, grow, grow.