Tuesday, January 3, 2017

Welcome 2017: Let’s Teach, Not Test

             I believe in being upfront with my readers, so let me start with a confession:  I don’t hate testing.

            I know it is a horrible thing for a so-called “educator” to admit. It’s sort of a social disease.

            Perhaps someone has a 12-step program that could help me… Assessment Anonymous. Perhaps.

            When I was a practicing teacher working on my Master’s degree, I loved collecting tests in a big notebook. Sight word lists, multiple-choice phonics quizzes, informal reading inventories, motivation questionnaires. 3-holes punched in their left margins. Organized by purpose. I loved them all.

            In one of my jobs I even did school entry testing, putting prospective kindergartners through their paces.

            Then, later, as my habit worsened, I started working on tests… the ACT, the SAT, the National Assessment, eventually even co-authoring a state test in Illinois.

            You probably know how this story comes out… everyone hooked on testing eventually hits bottom, the dark night of the soul when you know you have to change or it will be all over. I reached my nadir when I found myself writing a positive review of DIBELS for the Buros Mental Measurements Yearbook.

            Okay, now that I have that off my chest, let’s get real.

            Over the past 15 years or so, we have so overdone the testing thing. Not just the tests that educators usually don’t like—the high stakes accountability tests—but even the instructionally relevant ones that we believe can be beneficial… the running records, informal reading inventories, DIBELS-style screeners and monitors, and a slew of acronym-titled diagnostic measures. All of them. Too much. Too damn much.

Accountability Testing
            Accountability testing was not a bad idea… it just hasn’t worked the way its proponents thought it would. Nothing wrong with that: You have an idea; you try it out on millions of kids without any empirical evidence that it will work; then after a couple of decades of doing that with few victories… you keep doing it?

            The basic idea was this: schools should be run more like businesses. Business figured out how to improve quality by measuring quality. By carefully monitoring their products and services—by testing them, they could ensure higher quality. It’s why your car starts in morning, every morning.  

            By analogy, the idea was that if we tested kids, we’d see which districts, schools, and teachers weren’t getting the job done, then resources and efforts would be focused and kids’ learning would improve. That movement started back in the 1970s, but really got going full-bore in the 1990s… more than 20 years ago. Needless to say, we are still waiting with baited-breath for the uptick.

            I still like the idea of the public knowing how well schools are doing, even if that has no direct impact on kids’ learning. However, we don’t need to test as much as we do to find out how schools are doing. Such tests need to be as brief as possible, and they only should be administered to samples of children, not all children (the National Assessment does a very good job of this on a national basis, testing fewer than 100,000 kids every two years).

            But whether or not we adopt an accountability-testing plan that makes sense, there is NO excuse for teachers to spend inordinate amounts of time getting kids ready for these exams. So-called “test prep” should be banned if it goes more than a couple of hours a year; like having kids take a practice test the week before testing. Almost all of the time currently devoted to prepping kids for the PARCC, SBAC, STAAR, Aspire, and the other state tests should be devoted to…wait for it… teaching! That time could be profitably spent teaching reading, writing, math, science, social studies, and the rest of the curriculum.

            Why would I recommend such a crazy thing? Because the surest way to raise reading achievement is not through test prep, but through teaching kids to read.

Instructional Testing
            But the testing glut is not just due to the politicians and their accountability schemes. A good deal of the over-testing we have brought on ourselves. Again, the theory has seemed reasonable.

            If we know which kids are lagging in which skills, then we can be sure to teach those skills to the right kids, and voila, higher reading achievement. This idea is especially prevalent among those responsible for kids with learning problems; often it is proposed that those children be tested weekly! The claim is that such testing represents a more rigorous effort on behalf of the strugglers.

            But that claim has no basis in research at least as far as reading achievement goes. I’m not arguing against occasionally testing certain skills to see what kind of progress is being made, and if anyone is falling through the cracks… but that can be accomplished well by testing 2-3 times per school year. I’m also not talking about the teachers who observe kids’ performance within daily instruction and who look carefully at kids’ written work (in fact, they’re my heroes).

            But interrupting instruction frequently to have kids take tests—even tests aimed at focusing instruction—is a big time waster. There is no evidence that such testing regimens actually improve learning, but there is plenty of evidence supporting the teaching of reading. Our New Year’s resolution should be, “Let’s teach, not test!” Let’s devote our instructional time to teaching kids to read—not to preparing them for tests, not for administering tests. 


Elisha Carr said...

I agree with you that we have become a test-driven society. However, in many school districts teachers are not given the option to not test--They are mandated by superintendents, principals, etc to give the tests. I work in one of Florida's urban districts and we spent three weeks in December testing students in every subject (Reading, Math, Sci, Writing, Achieve 3000 and IReady Math). I appreciate and loved your post, but what advice do you have for educators that are stuck with testing mandates beyond their control? Even now in the new year it will be coming down the pipe line to asses kids more based on which standards they performed the lowest? I do not agree with this philosophy at all because no research supports these methods but how do we deal with this?

Anonymous said...

Yes! From your blog to administrations' ears! Ever since RtI began, my district has been so heavily "data" driven that the only way to get kids extra help has been to overtest them. Careful teacher observations carry no weight anymore. To be fair, I think this may be have been in response to some teachers coming with too general statements about kids. The push for common language has also become a push for common testing. I agree that formative testing is needed for instruction, but testing for endless data has to go.

Elise said...

I would venture to say few people have ever shared about a particular score on an Iowa Assessment as an adult, but I can assure you they will remember the day they proved a mixture can be separated. They tested out their hypothesis, separated the mixture, sorted it into its components, and graphed the data, and then ate it all. Real learning.

Tim Shanahan said...

I agree that often the problem is not the teacher's fault (when it comes to selecting tests or setting up a test policy--such as how many times per year the tests have to be given). That is more likely a bad decision of a curriculum director, special education director, or principal. Teachers, however, are the ones who often decide to do a lot of test prep--including test prep for the monitoring and screening tests (teaching phonics using the nonsense words from a particular DIBELS form is a good example of this bad practice). Usually when a problem is this big there is a lot of blame to spread around.

What can teachers do? (1) Limit test prep in their classrooms to no more than an hour or two about one week prior to the testing (unless mandated otherwise).; (2) I would not encourage teachers (or principals) to disobey orders, but I would strongly encourage them to ask a lot of questions and challenge the policies. Teachers could do worse than to print out copies of this blog and some earlier ones relevant to this issue and share them with the other professionals in their district--even sharing them with school board members. Teachers should ask for evidence--research evidence--supporting this amount of testing in reading. Such evidence does not exist. Professionals sometimes have to do things that they do not like. But that does not mean that they cannot professionally add information to the system and ask questions of the system.

Keep your chin up and good luck.

Tim Shanahan said...


Some testing for RtI is fine. We are doing too much and research is suggesting that RtI isn't paying off as well as we thought it would --perhaps even because of the over-testing. Giving kids increasingly intensive instructional support based on their demonstrated needs is wise and at least some of this decision-making could be based on analyses of work samples and teacher observation. However, tests every few months might help as well. More testing than that in reading; there is no evidence that it improves anything for the kids.

Bill Morelan said...

Love your blog and quote you often. Two points on this topic ...

First, Arkansas is very fortunate in that our State Director of Student Assessment (Hope Allan) strongly supports minimal testing. She often says, "Student growth comes from TEACHING, not TESTING!" And our district approach is "Don't test unless you use the results to monitor and adjust instruction." This is a HUGE change from just a few years ago when we literally devoted weeks to state-mandated, high stakes testing ... the results of which were ambiguous at best since the state used three different tests over a three year span (from Benchmark, to PARRC, to ACT Aspire). The new system isn't perfect by any means, but at least we're headed in the right direction.

Second, we've found that assessment can dovetail seamlessly with RtI ... but only if everyone, including students, clearly understands its purpose. And yes, I absolutely agree that over-testing can quickly invalidate RtI. Too often it's introduced with a top-down, compliance mentality ... and when it doesn't work, administrators tend double-down on testing, or throw the entire program out without ever exploring the root causes for the problem. In a world of quick fixes, sound bites, and instant gratification, that's not surprising ... but it is very sad.

Donna said...

Dear Tim,
What a refreshing article! I too have been concerned about the overuse of standardized testing, and the misuse of progress monitoring for years. However, I’d like to take it a step further. High stakes testing is politically motivated by the assumption that student performance is linked to teacher incompetence. It is perpetuated by pundits who are more concerned with their political agenda and constituents than the well-being and success of children. Mandated reforms such as NCLB did nothing to ensure equity in education. In fact, the reformist movement which has driven our national obsession with the corporate model of data and testing, has undermined public education and unraveled the public trust. The data from what has become the corporate educational complex has been skewed to support school choice and the conservative effort to dismantle public schools. The high stakes result of standardized tests have become a divisive tool which pits teachers against their students and administrators against their teachers as they desperately scramble to cover themselves in the discriminatory eye of the public.
In the end, the national obsession with testing is the undoing of decades of pedagogical research and everything we have learned about brained-based learning. The political emphasis on “test scores” has pushed education backwards into a results based (rote) mentality. It is far from the “21st century” ideals of project-based and differentiated education, and certainly not in sync with a 21st century brain that has been weaned on technology bytes and fast-paced brain-pop. It is an unnecessary distraction from the real work of teaching and it bogs down the immense resources of a well-trained work force with the crunching of data. Meanwhile, politicians polish their brass buttons with our blood, sweat and tears. Our time would be far better spent on designing standards based and engaging lessons with an emphasis on formative assessment that would teach our students how to read, respond and think across the curriculum.
Furthermore, the politicization of test scores has not served children in any way, and in my opinion, has only widened the achievement gap. It is nothing more than an insidious, racist and classist camouflage that perpetuates real estate red-lining in an effort to appease the far right, who are misguided in their belief that teacher unions are the source of the problem. It has perpetuated the myth that the failure of public schools is because of all those “bad teachers” and the unions that protect them. But worse, the publication of district test scores in newspapers has become a secret code abused by the middle class in their efforts to flee the challenges that public education presents. Public schools have suffered from brain drain and middle class flight in poorly funded districts, while those who are are able, to take advantage of charter and magnet schools and abandon ship. Rather than staying to fix the leak or at least offer life boats to the rest of us, families have turned away from public schools in the biggest marketing scam of the century – the move toward privatization. Our neighborhood and “struggling schools” have eroded into dystopian relics, or been closed for good. However, we don’t have a choice, but to welcome and accept the poor, the tired, the disenfranchised in the new colossus poem of public schooling.

Tim Shanahan said...


I disagree on a couple of points: (1) I don't think the accountability testing scheme is based on the idea that teachers are "incompetent," but instead on the notion that workers who are protected by unions tend to be lazy. Their concern isn't competency, but effort. I suspect that is the reason that the accountability testing scheme has basically failed. The problem is that while you can increase motivation with tests--people will try harder--it can only work if principals and teachers have a clear idea of what they need to do to raise reading achievement. I have worked with thousands of educators over the years and I think it is fair to say that most have no idea what steps they would need to take to improve reading achievement (except for getting different kids or remaking the society). That's why we see so much test prep--and we have long before these tests were actually high stakes (I was writing about this almost 30 years ago).

(2) These schemes are not the province of the far right--or, more accurately, of the far right alone. Unless Barak Obama, Bill Clinton, Hillary Clinton, John Kerry, Al Gore, and several Democratic governors of the last generation are part of the far right. George W. Bush put in a testing policy that allowed the federal government to intervene in districts by requiring extra tutoring for struggling students (paid for with federal funds) and educators screamed. Barak Obama reshaped those policies, requiring states to use the tests not to determine whether kids be tutored or allowed to change schools, but to fire teachers and close schools (and educators were pretty quiet about that). We have to stop being anti-stuff when one group of politicians embraces it, but for it or quiet about it when a different group of politicians signs on. These are good ideas or bad ideas--no matter which party is their champion at the moment. (For example, Republicans in 1990 supported test driven policies, but were against them in 1993-2001. Then they were for them again, etc.--The Democrats have taken the opposite journey. I guess that is just how politicians should operate--but researchers, teachers, teacher educators should not get sucked into that on issues that they have expertise on).