Tag Archives: Educational Data

Vol. #82: Read Theory

I set up ReadTheory for my students Friday and I am wondering where this FREE tool has been all my life! It definitely belongs on my list of free reading RTI strategies.

ReadTheory is a literacy tool which tailors itself to the student’s individual performance in reading. It selects a passage and questions for the student at random from the pool of available quizzes at the student’s level:

Students “choose a level to start.” My students were not aware that it meant “grade level” and some assumed they would start on “level 1”. After completing a passage that was entirely too easy, it quickly adjusted for them.

The video references how it adapts to a student’s performance as they go. Here’s how:

Level up: If a student performs outstandingly on the quiz (score 90% or more), then the quiz is never shown again and the level increases by one.
Level unchanged: If the student passes this quiz (score between 70% and 89%), it is never shown again and the student remains at the same current grade level of reading.
Level down: If the student performs poorly on the quiz (score 69% or less), then the quiz is replaced into the pool of available quizzes and the level decreases by one.

The teacher receives data charts and progress reports which are interactive and intuitive. The class average, student start level, current level, average level, and number of tests completed are all shown.

I especially like that students get immediate feedback on the questions they get right and wrong, and that for incorrect passages, they can click to get the “explanation behind the answer”.

There are several ways ReadTheory could be improved:

  • Being able to upload a csv file would have been really nice, although entering  students’ names one at a time didn’t take too long.
  • I would have really appreciated an easy pdf download by class that has the website, default password, and each students’ username in rows to cut apart for easier distribution.
  • I have some parents who would love a parent log in, similar to what Edmodo, Class Dojo, and Class Charts have, so that they could see their child’s ongoing progress.
  • Some of my students have noted it’s a lot like Study Island (which many of them had in Elementary school) but without the fun gaming/reward part.
    • Study Island costs money and I appreciate that Read Theory is free. It does keep “points” in some fashion, but I’m unclear how these are obtained and what they represent. They do not appear to be attached to badges or any type of reward within the actual web App outside of the statement of “You now have X many points.”  As I learn more, I may look to how I can reward them “outside the screen” in my classroom.
  • There’s no “stop”. Our students are trained to look for the stop sign when testing. These passages keep going on until a student chooses or a teacher tells them to stop. Choosing how many passages to do (or a time limit) and having a “stop” pop up would be a nice option.

These suggested improvements aside, I really like ReadTheory so for its ease of use, intuitive data, and personalization for students.

Vol.#78: Never The Destination

accountableI read Karl Fisch’s great post over at The Fischbowl about the word “accountability” and how too many in education erroneously equate it with using standardized testing to justify educational actions and decisions.

It got me to thinking how this current phenomenon often has educators, sometimes myself included, pinned in the corner of “all standardized testing is bad.” This is an understandable reaction to the ridiculous, high-stakes, over-emphasized testing of today. When one feels they are under attack, they take a defensive stance. Testing gives a snapshot of a narrow facet of skills, and while it shouldn’t be the focus nor the be-all-end-all… it isn’t completely useless.

After writing recently about my frustrations of the frequent pre-screening before the pretesting before the big test, it must sound like I’m completely backtracking. However, it’s the way the data is used that is important to examine.

Testing should be small, incremental, low-stakes, and personalized. If  I have a student who is struggling, as a language arts teacher I should be able to request testing to indicate issues of fluency vs. comprehension to know how best to help him/her.  It should be targeted and prescriptive, but this would require trusting educational decisions of professional educations, which is not what’s happening in the political scope of education right now.

Even the larger tests that level students in achievement ranges could be helpful if it were early in the year so teachers could use it to help inform their instruction for the year. However, it’s used at the end of the as a  summary of what the student and teacher have “done right”. This, again, is a misuse of the data. It’s an autopsy when only a biopsy can help a teacher help a student. Also, inferences are being drawn from the data which does not measure what it’s being assumed to measure. (ie: “teacher effectiveness.”)

Therefore, high-stakes testing becomes the “goal”. Schools can’t test to see what they need to teach, they are too busy scrambling to teach what’s on the test that contains what someone else decided was important and another said it would carry serious consequences for the student, teacher, and school if some bubbles aren’t colored as well as last year.  And consider what that these tests could never measure for just a moment…

Your doctor does not decide your heath on a BMI score or triglyceride reading alone.  However, that small piece of data can inform a medical professional if its part of a larger picture. The problem is when non-educators in charge of education (which is a problem in and of itself) decide to measure the doctor’s competence by his/her patients’ BMI average (teacher’s test scores). This is a misuse of the data, and a ridiculous way to measure the doctor. 

TL; DR:

Education Haiku

Vol.#72: Data Happens (And What To Do Next)

I have  data about literacy – my students’ and own children’s – coming at me on a regular intervals; tidalwaves on the beach of what is otherwise a peaceful school experience.

For my own son, he camScreen Shot 2014-09-21 at 5.08.52 PMe home with an mClass report with all little running men at the top of their little green bars – save one – and a lexile level that corresponds with a 3.6 grade level early in his third grade year.  However, another letter says he’s been flagged as a “failing reader” based on the preliminary standardized test given in the beginning of third grade. This would have perplexed me if I didn’t already know how ludicrous it is to assess literacy of children with these frustrating bubble tests.

For my sixth-grade students, I have access to their standardized test data from the end of fifth grade – the ones with passages that are way too long assessing way too many standards and simply expecting way too much of the poor ten-year-old test takers.

We also give our middle schoolers quarterly timed tests on basic skills in reading and math. Based on these results, students are sorted into green, yellow, and red, with intervention plans written for those in the “danger zones”. Also, there are standardized benchmark tests at the end of each quarter to see if they are on track to attain a passing achievement level for the standardized state test at the end of the year.

demaNdingIf anyone counted, that’s seven tests during the year for students, including the “real” test. But not including any tests given by the teacher. (And that’s just for reading, don’t forget to then add in math. And science. And social studies… But I digress.)

I am not naive enough to think I am going to change the path we are going down right now, but I feel strongly that if we are going to make students do all this, I’d better find a way to make all the resulting data helpful to my instruction.

And therein lies another layer of my molten lava white-hot fury. What has been sorely missing from the dialogue in all these data-sessions is the next steps. Ok, Sally Sue is “red”.  What does she need now?  Or, even more frustrating, she passed one test, but is “red” on the other. So…now what? What do I DO for her? (You know, that I wasn’t going to do anyway? Like…teach her?)

Perhaps this oversight is because those who pushed this agenda only wanted to sell us all the screening tests so they don’t actually know what to do next? Or, maybe their answer is they want us to buy their scripted program to “fix it”, but we are all out of money?

At any rate, here’s where I am with this new normal.  I need pragmatic (*ahem* free) ways to address all this conflicting data. What follows is a list of  strategies I have to that end:

  • Offer the same article in several different lexile levels using Newsela. Some articles have leveled questions as well. (Newsela has a free version and a “pro” version.)
  • ReadWorks “The Solution to Reading Comprehension” offers both nonfiction and literary passages, questions, and units for free. It includes lexile leveling information.
  • You can also check the reading level of any text or website at  read-able.com for free.
  • Offer clear instructions for how you want students to complete a close reading of a text. Here’s mine. Sorry for the shameless plug. 🙂
  • Mr. Nussbaum’s webpage has reading comprehension passages and Maze passages that score themselves for free! It only goes up through grade 6, so it would only help students up through about a 960 lexile.
  • ReadTheory is free, and allows you to create classes and track reading comprehension progress.
  • There are several reading leveler apps you can pay for and they are probably fancier, but I’ve found this one handy, both as a mom and as a teacher. For example, I used to have long conversations with my students who kept picking up books during DEAR time, not an occasional graphic novel, but always a graphic novel, cartoon books, picture book …you know the type? Anyway, scanning their bar code and simply telling them it has a 2.4 grade level has been more effective than the long conversation. 🙂
  • One on my horizon to try: curriculet.com  It’s free and I’ve heard good things!
  • I have also found the following conversion chart handy, because of course the data does not always come in the same format:

4879716These have helped me in more than one “What are you doing for my child?” conference and to complete the required intervention plans based on all the data. I don’t know if they have revolutionized me as a literacy teacher, but I suppose time scores will tell.

Have a strategy, tool, or resource for helping your students as readers? Please share in the comments!

Vol.#63: Simplifying a Teacher’s Life: Free Technology Tools for Assessment

Last week, I posted my presentation   “Every Teacher a Literacy Teacher Using Technology Tools” from what I shared with the 2015 Kenan Fellows at the North Carolina Center for the Advancement of Teaching (NCCAT) in June.  As promised, though a little late, I am adding the other presentation: “Simplifying a Teacher’s Life:  Free Technology Tools for Assessment” this week.

The video is long (30 minutes), but as with any flipped lesson, it provides the benefit of being able to pause, skip, or come back to it as needed. Plus, the focus is free technology tools to collect student data so you spend less time grading, so in the end you will get your 30 minutes back, I promise! 🙂

cc-by-nc-sa

  • Care to share your experience or planned use for any of these tools?
  • Have another tool to add?

Please share in the comments!

Vol.#47: Education Infographic Inferences

I learned last week that communicating data via an infographic is a power not to be wielded lightly. The ongoing feedback in the comments had me repeatedly changing and re-editing numbers and phrasing. For example, at the bottom teachers work “more than others” became “more than full-time” because so many commenters seemed to be taking it as a personal judgement on how much they did or did not work in their own jobs.

Back in July, I’d gotten quite a lot of feedback (almost 200 comments) about the infographic on the data of the change in average teaching salaries over a decade, so I wondered: what about an infographic on just the average teachers’ salaries per state? What would that information look like?

The most recent data I found was from 2012, and as I created the high-to low list, I thought I also saw a pattern on how states voted during that same year’s presidential election.

I color-coded the infographic according to those results:


easel.ly

I thought about how NCAE almost always supports the democratic candidate, and I found those who “stood out” like Alaska and New Mexico very interesting. Also, I wondered how much of the “average salary” was higher from retention of experienced teachers (particularly abysmal in my own state of North Carolina) or other factors outside of education specifically, such as the general cost-of-living. For example, even though Hawaii is in the top half of the states for a teacher’s average salary, according to at least one source the “comfort index” on that salary is actually the lowest in the nation due to how expensive it is to live there.

What do you infer from this data?

Volume #44: Literacy Data, Part Deux

In my last post, I argued against the use of the current practices for gathering data for measuring growth and proficiency in literacy.

I suggested that for math, formative standardized test data is a biopsy. For literacy, it’s more like an autopsy.

And while the data indicates strong versus sickly readers, this information is usually no surprise to the professional educator, and more importantly it offers no treatment plan: advice on which medicine to administer.

With the release of my state’s scores re-renormed to the Common Core, there’s lots of focus on all the new data. What it all means. Why the scores are lower. How it will be improved.

And while the politics rage on, I have to explain to parents that their child simply went from twelve centimeters to five inches, and yes the number may actually be smaller, but I believe it to show growth in his/her reading ability.

And I need to take this new information and figure out how it should inform my instruction. I need the data to indicate a treatment plan for the literacy health of my students.

During my participation in VoiceThread titled “Formative Assessment and Grading” in October 2011, Dylan Wiliam said something that has always really stuck with me:

“One of the problems we have with formative assessment is a paradigm that is often called, “data-driven decision making”. This leads to a focus on the data, rather than on the decisions. So, people collect data, hoping it might come in useful, and then figure out sometime later what kinds of decisions they might use the data to inform.  I’m thinking that we ought to perhaps reverse the ideas in data-driven decision-making and instead focus on decision-driven data collection. Let’s first figure out the decisions we need to make, and then figure out the data that would help us make that decision in a smarter way.”

~Dylan Wiliam   “Formative Assessment and Grading”,  Slide 5   [My emphasis]

I’ve pondered this at great length. If my goal is decision-driven data collection, what would I want out of a standardized literacy assessment? What do I want the data to tell me?

What else? What other information (as a teacher or as a parent) do you believe the data should provide about students’ literacy abilities?

Vol.#43: Literacy Data

countsSeveral years ago, an ELA colleague and I were presenting writing strategies to another middle school’s PLTs. The IRT’s office was in the PLT meeting room, and during a break between our sessions she remarked how she always had math teachers coming in to scan the results of their County required standardized test benchmarks immediately. However, she always had to chase down the language arts teachers to “make” them scan the bubble cards for the data. They’d given the test as required, just not scanned the cards for the results. She asked us what to do about it, and we sheepishly admitted we were often the same. Amazed, she asked… “Why?”

“Well, that data doesn’t really tell us anything we don’t already know.

Standardized data from the math benchmark practice tests tells our math teammates if students are struggling with decimals, or fractions, or two-step equations. In short, if students need more help…and if so, with which with specific skills.

The truth is…the data on these reading benchmarks tells us that since our AIG students score higher gifted readers must be better readers and our ESL students who are learning English don’t score as well on a test for…reading English.”

Image Credit: Pixabay User Websi
Image Credit: Pixabay User Websi

None of that is new information to any literacy teacher, and even if it were it doesn’t speak to how to shape his or her instruction. We are Data Rich, Information Poor. (D.R.I.P.) Analysis of that data does not help us see the path forward clearly for our students. Perhaps worse, it doesn’t necessarily even reflect the quality of instruction they’ve been given.

And while greater educational titans like Alfie Kohn have already explained  the many problems of relying on standardized data for, well, anything, it is my contention that using it to measure English Language Arts, both for measuring teachers and students, is an exceptionally erroneous practice.

Standardized testing by definition is supposed to be an “objective” assessment. However, subjective factors such as beliefs and values aren’t shouldn’t be separable from measuring literacy. While math is cut and dry (there is a right answer) interpretation of a literary work it not black and white. The students who can argue support for more than one of the four cookie-cutter answers – and do so in their heads during the test thereby often choosing the “wrong” one – are likely in reality the best readers. Disagreement on what an author meant by effective figurative language use or dissention in supporting different possible intended themes are not to be transcended in analysis and assessment of literature but embraced.

Am I missing some insight in interpreting formative standardized benchmark data? Is there some value here that I am overlooking? Please let me know in the comments!