Should We Alter the Reading Benchmarks Because of the Pandemic?

Teacher questions: Over the past few weeks, I’e fielded many questions about testing — from policymakers and teachers. Here are a couple of examples:

1. Should schools lower grade level benchmark reading expectations due to lost instructional time during the pandemic? Please advise ASAP.
2. A question that keeps coming up is, if a student cannot read grade level text on their own, can they then listen to the text and answer the questions on an assessment in order to be considered “meeting” reading standards 1-9 in grades 2-5 since there is a specified Lexile band for those grades through standard 10?

Shanahan’s response:

Over my career I’ve worked on many tests (e.g., PARRC, ACT, NAEP, SAT, and various state tests and commercial tests, too). I have also done research on classroom testing, including informal reading inventories and cloze tests. Despite that, I think we overdo it with testing. I’m not just mouthing the usual complaints about intrusive accountability tests, but I think we do more classroom assessment than necessary as well.

Nell Duke and I were on a podcast for the National Association of School Boards recently and a question about state tests came up. Both Nell and I were unified in our opinion that given the terrible disruptions to education this year, annual accountability testing should be suspended this time around.

There are several reasons for not bothering with that kind of testing right now. The most persuasive is that instructional time is at a premium. Too much instruction is being lost. Devoting any instructional time to accountability assessment at this time would be profligate.

We should forego next spring’s accountability tests. But we also should be prepared for an early round of testing at the beginning of the next in-person school year (Fall, 2021?) to provide schools with worthwhile information early on. Summative data this spring won’t help, but formative data early the next school year would.

But it isn’t just policy makers who are concerned about evaluation. Teachers and principals seem to be at sixes and sevens over classroom testing. How do we test over Zoom? Should we test at all? How do we interpret the tests? Should we lower our standards?

To tell the truth, though I’ve tested hundreds — maybe thousands — of students, I’ve never done so from outside a classroom. I think, with sufficiently high quality equipment, I could test students’ oral reading at distance — and, yet, teachers I respect are divided over the matter. Those who’ve tried tell me they don’t trust the results, while others are more confident. These tests have not been validated under current circumstances. We should be able to figure out that one — how to evaluate performance at a distance — but we may take too long.

I’m less sanguine about other kinds of “at a distance” testing, since kids can figure out work arounds to avoid having to know what the tests are evaluating.

Of greater concern, is that many schools have lost sight of the purposes of these classroom tests… either to offer predictions about students’ ultimate success (so we know whom to give extra resources to) or to provide benchmarks that show where students are right now so we’ll know what should come next. The introductory questions above, I suspect, are driven by concerns over fairness. In other words, the teachers are treating these screening and monitoring tests as if they were high-stakes assessments like the ACT or SAT.

Imagine if physicians decided that their diagnostic tests had to be interpreted according to such notions of fairness. “Mrs. Jones the results of your mammogram would usually be concerning. But since we are in a pandemic, I’m not going to recommend a biopsy; the disease might not have been this advanced without all the stresses that you have been under. I think it would be fairest if we treated this as a less advanced tumor and just not worry about it right now.”

We have a word for such notions of “fairness” — malpractice.

Schools often evaluate students’ early reading performance using DIBELS-style measures. The benchmarks of such tests are intended to be diagnostic … “Oh, Johnny still is having trouble with phonemic awareness, I’d better continue PA instruction for him, but the rest of the class can move on” or “The test indicates that Janie is struggling with decoding, especially with the vowels … I’ll teach those next.” Adjusting those benchmarks may make everyone feel good: “Johnny and Janie might not be doing as well as in the past, but doggone it, it’s not their fault that we haven’t had as much teaching as in the past.”

Those kinds of benchmark adjustments can only disguise the fact that Johnny and Janie need additional tuition with particular reading skills. The kids and teachers might — for now — feel great about the results of tests with lower benchmarks. Unfortunately, that good feeling will be temporary, a sugar high if you will. The test is telling you there is a problem. Accepting a lower score as being sufficient, just will hide the problem and keep it from being addressed.

Those reading text level goals were not established with the idea that they’d be easy for everyone to reach in a certain time period. No, they were aimed at setting a long term continuum that, if students advanced along successfully, would result in adequate proficiency. Adequate here means that students could enlist in the military, enroll in higher education, or get a job — and thrive.

I doubt very much that colleges, employers, or the military intend to lower their standards down the road because of today’s pandemic. Accordingly, we must do everything possible to get students to the levels of achievement that will allow them full access to our society’s economic, civic, and social benefits.

The idea of concluding that students can read well enough if they have strong listening skills is more of the same and merits the same response. The reading comprehension standards require that the student be able to the read texts independently. Our job is to teach them how to make sense of text at those levels of difficulty indicated in the standards.

I have no problem with reading accommodations for learning disabled or non-English speaking students. If I want to find out what they know about science, their inability to read English could lead me to incorrect conclusions. But, if the point is to teach students to read, then teaching them to listen instead is a rip off.

This pandemic is an educational disaster for many of our boys and girls. Lowering our standards and our efforts to accomplish them will not make it better for the kids; it will just reduce the likelihood that we’ll do what is necessary for their success. Please don’t lower those benchmarks.

See comments here › (opens in a new window)