Danger Ahead?

As Tennessee prepares to test more students than ever via an online platform, there are some signs of potential trouble.

Chalkbeat reports:

State officials said Thursday they are confident the new digital platform will work under heavy traffic, even as their new testing vendor, Questar, had headaches administering computer-based tests in New York on Wednesday. Some students there struggled to log on and submit their exam responses — issues that Questar leaders blamed on a separate company providing the computer infrastructure that hosts the tests.

 

Tennessee officials say they are working with Questar to ensure similar problems don’t occur in Tennessee. They also point out our online testing infrastructure is different and that Questar will have troubleshooting staff in the state during the test administration.

Here’s the problem: Across multiple testing vendors and dating back to TCAP, Tennessee has had problems with testing. This includes the now perennial issue of not being able to deliver scores back to districts in a timely manner. In fact, in December, districts were told scores might not be back in a timely fashion this year, either.

It’s possible the state and Questar have all the issues worked out and this year’s test administration will be nearly flawless. However, the record over the past few years is not encouraging.

Then, there’s the issue of what happens with the results. If they are available for factoring into student scores, it is up to districts to choose the method. I’ve written before about why that’s problematic. Here’s a quick summary:

The cube root method yielded on average a quick score, the score that goes for a grade, of 4.46 points higher. In other words, a studentscoring basic with a raw score of 30 or higher would, on average, receive an extra 4.46% on their final quick score grade, which goes on their report card. A student who scored a 70 last year could expect to receive a 74 under the new quick score calculation.

The additional points do drop as one goes up the raw score scale, however. For the average basic student grades 3-8 with a raw score between 30 and 47, they would receive an extra 5.41 extra points under the new method.

The average proficient student grades 3-8 with a raw score between 48 and 60 would get 4.32 extra points under the new method.

The average advanced student grades 3-8 with a raw score of between 61 and 67 would receive an extra 1.97 extra points under the new method.

The difference varies much more widely for below basic students, but the difference can be as much as 25 points in some cases.

So, for those districts using quick scores in report cards, there could be a wide variance across districts depending on the method chosen. It seems to me, districts should have already communicated to families how they will calculate quick scores with some justification for that choice. Alternatively, the state could have (should have?) mandated a method so that there is score consistency across the state.

Then, of course, there’s the issue of using these scores in teacher evaluation. Let’s say testing goes well this year. This would be the first year of a test without problems. If that happens, this should serve as the baseline for any test-based teacher evaluation. Yes, I think using value-added scores is a misguided approach, but if Tennessee is going to go this route, the state ought to take steps to ensure the data is as accurate as possible. That would require at least three years of successful test administration. So far, we have zero.

If TNReady is a great test that has the potential to offer us useful insight into student learning, it’s worth taking the time to get it right. So far, it seems Tennessee has yet to learn the lesson of the NAEP outlier — we don’t need rapid acceleration, we need to be patient, take our time, and focus.

 

For more on education politics and policy in Tennessee, follow @TNEdReport


 

Not Exactly Helpful

In a story yesterday about TNReady scores not being ready in time to be counted in student final grades, I noted a statement published in the Clarksville Leaf-Chronicle and attributed to Tennessee Department of Education spokesperson Sara Gast. Here’s that statement again:

But Sara Gast, spokeswoman for the state Department of Education, said school districts would receive their scores based on how quickly they returned their materials.

This was the first week school districts could receive data back, and districts across the state will get their scores on a rolling basis over the next couple of week through the week of June 5, she said.

She said some districts will not get their scores in time to be counted in final grades “because they did not meet the deadlines.”

Since then, Clarksville-Montgomery County Schools has posted an update on their Facebook page:

The state department of Education has clarified that CMCSS did NOT miss any deadlines. According to Sara Gast from the Tennessee Department of Education, “We provide three different timelines for a reason, and all are equally fine and acceptable for districts to be on. We are neutral on which deadlines districts meet, and it is reasonable that larger districts would need additional time to ship materials back and may use the entire window to do so. We have always fully expected that we will have districts on all three tracks based on their local decisions.” Assistant Education Commissioner Nakia Towns confirmed that with this comment: “We emphasized that there was no “miss” of deadlines. We just provided three timelines.”

What’s not clear from this statement is whether it was anticipated that scores would not be ready by the end of school depending on the track chosen by districts.

It’s also interesting how the DOE’s explanation has shifted from blaming districts for missing deadlines to now saying that having more than 75% of districts not getting scores back before the end of the year was the plan all along.

I offered a solution yesterday. It’s simple, really.

Stop using this single test as the primary indicator of student performance, teacher effectiveness, and school accountability.

For more on education politics and policy in Tennessee, follow @TNEdReport


 

 

The NeverEnding Story

Another day, more stories of districts reporting to families that TNReady scores won’t be back in time to be factored into student grades. I first reported that Williamson County sent word that scores would not be back according to the original timeline. Next, it was MNPS telling parents that TNReady scores won’t be back until June, meaning they won’t be factored into report cards.

Now, two more middle Tennessee districts have sent notices about TNReady results not being ready in time.

Here’s the notice from Clarksville-Montgomery County:

The TNReady materials from CMCSS have been returned to Questar. Tennessee has noted that they will be unable to provide the district with the test results until after the end of May. Based on CMCSS Administrative Policy INS-A023, effective April 17, 2015 in alignment with HB 36 SB 285 Amendment (005744), Clarksville Montgomery County School System will not include students’ state assessment scores in their final spring semester grades if the state assessment scores are not received by the district at least five instructional days before the end of the academic year. As we will not be receiving the scores until the end of May the scores will not be included in students’ grades for this year. The second semester average for elementary and middle will be 50% 3rd 9 weeks and 50% 4th 9 weeks. The second semester grades for high school will be 40% 3rd 9 weeks, 40% 4th 9 weeks, and 20% final exam.

And one from Wilson County Schools:

Good Afternoon!

The end of a school year always brings about a flurry of activity and excitement, but I wanted to take a moment to update you on report cards for the spring semester.

A couple of weeks ago, we announced that report cards would be available, VIA Skyward, on Tuesday May 30th. Unfortunately, it’s unclear whether the district will be able to meet that date, due to a shipping delay that was beyond our control. While our district met all of the required deadlines to ensure that our raw scores would be returned by May 22nd, the state vendor responsible for picking up the completed materials arrived several days later than scheduled. This has affected a number of large districts across the state, including Wilson County.

The TN Department of Education is aware of these delays. They’ve assured us that they’re working with the vendor to “find a remedy” for the school districts impacted. Our hope is that a solution WILL be found, and our raw scores will be returned on time. Having said that, we thought it was important to make you aware of what’s happened, in the event that report cards have to be delayed for a week.

You may remember, TNReady scores came back later than expected for the fall semester, causing report cards to be delayed. While school districts have the authority to exclude TNReady scores that are returned more than 5 days late, it is the position of Wilson County Schools that the scores be included for this semester, as they were in the fall. This is not a decision that was taken lightly. Many conversations have taken place with teachers and principals about this issue, and the overwhelming consensus is that we include the scores on report cards. Students have worked incredibly hard all year to show of their skills, and we’re eager to see just how well they did!

Thank you for remaining patient, as we work through the process. We’ll keep you updated, as we receive additional information from the state. If you have any questions, feel free to submit those to “Let’s Talk” at the following link: http://www.k12insight.com/Lets-Talk/embed.aspx?k=WK9F4DLT. You can also reach out to me directly, using the information below.

Sincerely,
Jennifer Johnson

Here’s what the Department of Education has to say about the importance of state assessments:

Our state tests serve multiple objectives:

  • They provide feedback about students’ academic progress and how it aligns with grade-level expectations, providing parents and teachers a big-picture perspective about how a student is progressing compared to peers across the district and state, including a student’s strengths and growth opportunities.
  • This builds confidence and transparency about students’ readiness for college and the workforce among Tennessee universities and employers and holds us accountable to serving all students fairly.
  • Assessments help educators strengthen instruction and reflect on their practice, and allow us to highlight schools where students are excelling, so we can learn from those who are doing well.
  • State assessments also help inform decisions at the state level and help state and district leaders determine how to allocate resources, better invest in schools, and identify where we may need to offer additional support.

All of this sounds pretty important. But, not important enough to get it right. Last year, TNReady was a complete disaster. For the past four years, there have been problems with scores being either not available or not clearly communicated.

This year, the state is not providing quick scores to districts — those are the scores used to factor into a student’s final grade. Instead, the districts were to receive the raw data and choose a method of tabulating quick scores. An analysis of the various methods indicates a significant difference in scores depending on the calculation used:

The cube root method yielded on average a quick score, the score that goes for a grade, of 4.46 points higher. In other words, a student scoring basic with a raw score of 30 or higher would, on average, receive an extra 4.46% on their final quick score grade, which goes on their report card. A student who scored a 70 last year could expect to receive a 74 under the new quick score calculation.

The additional points do drop as one goes up the raw score scale, however. For the average basic student grades 3-8 with a raw score between 30 and 47, they would receive an extra 5.41 extra points under the new method.

The average proficient student grades 3-8 with a raw score between 48 and 60 would get 4.32 extra points under the new method.

The average advanced student grades 3-8 with a raw score of between 61 and 67 would receive an extra 1.97 extra points under the new method.

The difference varies much more widely for below basic students, but the difference can be as much as 25 points in some cases.

So, for those districts using quick scores in report cards, there could be a wide variance across districts depending on the method chosen. It seems to me, districts should have already communicated to families how they will calculate quick scores with some justification for that choice. Alternatively, the state could have (should have?) mandated a method so that there is score consistency across the state.

Of course, since a number of districts now won’t have data back in a timely fashion, there may not be many districts using quick scores at all this year.

Here’s the key point: Last year’s TNReady was a debacle. That means this year is really the first year we’ve done TNReady. Instead of jerking districts (and their students) around, the state should have waived use of TNReady scores to evaluate teachers and grade students this year. Doing so would have provided insight into the time it takes to get scores back to districts and allowed for possible changes in administration for next year. Instead, the plan was rushed with a new vendor. Now, we’re where we’ve been year after year: The school year is ending, and there’s a problem with test data.

One more thing: Despite this being the first year of a successful administration of a new test and despite the gap in test results — TCAP in 2015, no results in 2016, TNReady in 2017 — the scores from TNReady will still factor into teacher evaluation.

A word of caution to districts during the 2017-18 testing cycle: The state’s track record with deadlines and score results is not so great. Maybe when they promise you scores will be ready according to a certain timeline, you should be making plans for that timeline not being met.

For more on education politics and policy in Tennessee, follow @TNEdReport


 

Back to the Future

After yesterday’s opening day fiasco with the new TNReady test, Commissioner Candice McQueen announced that all TNReady tests for this year will now shift to paper and pencil tests and a new testing window will be created. No testing will happen before February 22nd.

Here’s the letter Commissioner McQueen sent to Directors of Schools about the shift:

Thank you for your patience as we faced technical challenges with the MIST platform this morning. At 8:25 a.m. CST the state’s vendor for TNReady, Measurement Incorporated, experienced a severe network outage, causing significant problems with the MIST platform. Like you, we are incredibly disappointed that the MIST platform was not accessible to schools across the state as the Part I testing window opened.

 

Shortly after learning about the issue, we advised that schools experiencing problems with the test discontinue testing, and return to their normal classes.

 

Throughout the 2015-16 school year, the department has continuously worked with Measurement Incorporated to strengthen the online testing platform. As a result of district feedback and through our efforts to collaborate, we have mitigated and eliminated many technical issues. The online platform has undergone many capacity tests, yielding actionable information to drive improvements. Following Break MIST Day last October, we’ve made significant investments in server capacity. As a follow up to our Jan. 12 capacity test, the department’s technology team also spent multiple weeks in the field visiting select districts around the state to reproduce system errors in a real-world, real-time situation to gather better diagnostic information. As a result of this continued analysis, we offered districts the option to move to paper testing as we saw continuing issues with how the platform interacted with districts’ infrastructure.

 

Unfortunately, issues have continued to arise with the online platform. The new nature of the issue this morning has highlighted the uncertainly around the stability of Measurement Inc.’s testing platform, MIST. Despite the many improvements the department has helped to make to the system in recent months and based on the events of this morning, we are not confident in the system’s ability to perform consistently. In the best interest of our students and to protect instructional time, we cannot continue with Measurement Incoporated’s online testing platform in its current state. Moving forward, during the 2015-16 school year TNReady will be administered via paper and pencil (both Part I and Part II).

 

We thank districts, schools, and teachers for their commitment and perseverance to move our students to a 21st century learning environment. We know this is what the real world requires. We understand and appreciate the investment of time, money, and effort it has taken to attain readiness.

 

As a result of a statewide shift to paper and pencil, we will delay and extend the Part I testing window. Measurement Incorporated is currently scheduling the printing and shipping of the paper tests, and the department will share the revised testing window with districts by Thursday of this week. We understand that the shift to paper and pencil testing has many scheduling implications for your schools, teachers, and students. We thank you for your patience and cooperation as we transition to a test medium that we are confident will allow all students to show what they know.

 

TNReady is designed to assess true student understanding and problem-solving abilities, not just basic memorization skills. Regardless of the medium of assessment, this new and improved test will provide schools, teachers, and parents with valuable information about our students college and career readiness.

 

Warning Signs

Prior to Monday’s scheduled test administration, some educators across the state were raising concerns about the testing system and its ability to handle the load of student all across the state.

Amanda Haggard reports:

In a letter sent to the Tennessee Department of Education Commissioner Candice McQueen on Jan. 31, RePublic CEO Ravi Gupta outlines exactly what happened when the school made its attempt at the test. From the letter:

Our experience on January 28, however, raised substantial concerns about the technical capacity of MIST [Measurement Incorporated Secure Testing platform] to support state-wide testing. RePublic has only 1,200 kids — a tiny fraction of the State’s 500,000. On January 28, we attempted to administer the Math practice test on MIST as a step toward preparing kids for the first round of state exams. More than half of our kids were unable to log on, were kicked off the platform after logging on, or could not submit a completed test. The critical issue, confirmed by MIST representatives, was an error or series of errors on MIST’s own servers.

Haggard goes on to detail other concerns raised ahead of Monday’s test administration.

A Call for a Pause

In response to the challenges presented by the TNReady test administration, some legislators are now calling for a pause on test-based accountability for students, teachers, and schools. The tests would still be administered, and results reported, but they would not impact student grades, teacher evaluations, or the state’s priority schools list.

What Happens Now?

The state now has asked its vendor, Measurement, Inc. to provide paper and pencil tests. These will not start before February 22nd. Districts and schools will have to reschedule testing based on the availability of tests and guidance from the Department of Education.

Now that the tests have shifted to pencil and paper, some are asking how they will be graded. Human graders were always a part of the equation due to the constructed-response nature of the tests, but they will now be assessing handwritten responses.

A Pattern?

This is the third consecutive year the state has had problems with its testing regimen. In 2014, quick scores were not ready in time to be factored into student grades. Last year, there was a change in quick score calculation that was not clearly communicated to districts and which resulted in confusion when results were posted.

 

For more on education politics and policy in Tennessee, follow @TNEdReport