astronomical cost, business model of education, ETS, grading buildings, higher profits, intellectual development, learning opportunity, Mandatory high stakes testing. data points, meaningless results, nuclear testing security, prepackaged curricula, reliable measurement
In the first article of this series–The Tyranny Of Testing, Part 1 2019–I explained the incredible amount of time lost to mandatory, high-stakes testing (and other matters) each year. In the second installment–The Tyranny Of Testing, Part 2 2019 –I focused on trends and damages caused by testing, particularly forcing mandatory tests on special education kids fundamentally unable to take those tests. The results of such “testing” are less than meaningless, and the results of testing normal kids are often the same.
To find any article in this series, enter “the tyranny of testing” in the SMM homepage search bar.
American schools lose more than 40% of available teaching/learning time to mandatory, high stakes testing. I would hope that any parent or citizen would be outraged by that simple, indisputable fact. The most precious commodity a teacher has is time, and even five minutes a day taken from a class amounts to a loss of about 20 classes of the school year–a solid month of instruction.
The obvious questions: is mandatory, high-stakes testing worth the loss of more than 40% of the school year? Is having a few data points—the results of a few tests—at astronomical cost (see part 2) worth the loss of all of that learning and growth opportunity? Will taking those tests, rather than reading, writing, and making untold neural connections, better prepare students for the world of work, for college, for life?
Commenting on the series of articles on which this updated series is based, reader labrat wrote:
I was looking forward to your ‘test’ post, but I’m a bit disappointed. You usually make a very clear argument. In this case, I’m not sure what your argument is. Why do teachers feel compelled to ‘teach to the test’? It seems to me a test should be designed to measure a student’s mastery of a certain curriculum. Are you arguing that these tests do not achieve this? Are teachers not given a set of goals a class is supposed to meet and a teaching plan developed to reach those goals? So is your argument against the metrics these tests are measuring or against using a test to measure whether your teaching plan achieves the goals of the class?
If teachers didn’t ‘teach to the test’, but actually taught the material the children should master, then would these children fail these tests? Why would that be?
Good questions all. Obviously, if a teacher knows the content of a test in advance, and particularly the types and structure of the questions that will be used on that test, they can, by teaching to the test, produce a very high degree of student success. But is that truly the kind of success that should be the goal of education? Is learning a few narrow techniques and memorizing and spouting back a set of answers truly learning, particularly when most people will forget most of them almost immediately after the test is completed?
Why Do We Test?
As a teacher of English, I use tests for limited purposes. The primary value of testing is not—as most people think—to serve as a reliable measurement of what a student knows and will know forever. Notification of an upcoming test tends to encourage students to pay closer attention to the material to whatever degree is necessary to pass the test. Unfortunately, this is not universal; some students could not care less whether they pass. Others have a very broad base of general knowledge and have to expend relatively little effort to pass. Many of the concepts and ideas covered are already familiar to them, so they have less to assimilate than some. Others are simply very good at the tactics necessary to pass tests of all kinds, and once they understand my usual methods, they’re more or less golden for the year. After the test, they’ll tend to forget most of the material. Most educators would agree that people tend to retain only about 10% of what they learn, at least in the form they were required to know for the test.
This formulation of “10%” is commonly known, but it is, of course, not absolutely accurate. It’s merely an aphorism of sorts for the observation that people tend to learn what’s necessary for the moment, and forget much of it thereafter.
That’s why I use very few multiple choice or true/false items on my tests. Though such items are very easy and fast to grade, they reveal little other than the development of a student’s short-term memory. Most of my tests require analysis, interpretation and writing, all long-term skills that develop over time. It takes much more time and effort to grade such tests, but the results are far more indicative of long-term and lasting trends and abilities. And because students actually produce a product, something unique to them, they tend to have greater pride if they do well and tend to want to have that feeling again.
It is interesting to note, circa 2019, that the Education Testing Service, producers of the English test for Texas, have moved away from analysis and writing, and focused far more on multiple-choice items. This produces much higher profit margins, and the data points produced by the tests mean less than ever before–not that they ever meant very much.
Some teachers use tests for a respite. When students are taking a test, a teacher has time for grading, writing lessons, or simply vegetating. Many textbook packages come with all manner of exercises and assignments, including prepackaged tests, which require no effort or intellectual energy from the teacher. They also have some advantages in terms of appearances-–the teacher appears to be doing something significant–-less so in terms of actual learning and the intellectual development of their students.
Another significant factor is that of the seven separate classes a high school teacher sees each day, each has its own unique personality and needs. Prepackaged curricula ignore this important factor and make it difficult for teachers to adjust to meet the needs of their students. Of course, those selling these materials will claim the materials are so perfect, so inspiring, so supernaturally effective, no adjustment is necessary. The check is also in the mail, you can keep your doctor, and if the media says it, it’s true.
If a student spends all their time with a prepackaged curriculum, they are, by definition, being exposed to little more than teaching to the test. They will usually do pretty well on the tests, because the entire curriculum is focused toward them, commonly in a simplistic way. If all you did in a given class was work on material aimed at the test at the end of a given unit, would it be surprising you did reasonably well on that test? But what would that test score tell anyone? That drilling to pass a given test tends to produce results in passing that test?
Any competent teacher will know far more about the intellectual abilities and needs of their students within two weeks of the beginning of a school year than they’ll learn from any test. By the end of each school year, my students will have completed about 120 assignments, an average of something less than four per week. Of those assignments, only about 10% will be conventional tests of any kind. The rest will be assignments far more revealing of their actual development and abilities.
Since before the time of Aristotle, human beings learned in exactly the same ways they learn now: through correct and repeated practice of meaningful concepts and skills, guided by skilled, experienced and dedicated teachers. This will never change, and there is very little or nothing new in education. This does not stop people from make millions selling the same old failed concepts over and over again, usually by making up new acronyms and packaging.
For example, details in writing are now “golden bricks.” Scissors, staplers, rulers and similar items are now “manipulables.” Isn’t that much more educational? Don’t you see the inestimable value? In my classes, details remain details, and we cut paper with scissors.
What is more revealing of a student’s actual learning and abilities, the knowledge of a teacher that has read and remarked upon 120 assignments over a year, that has discussed those assignments with that student, that knows their strengths and weaknesses, and that sees them daily, or the scores of a few tests given once in their life? Unless you believe that the score of a single test—you also have to believe it to be perfect–can reveal anything meaningful about a person, to the extent that test should determine whether they graduate from high school, and should take 40%+ of their available time for learning, the answer should be obvious.
Let’s consider Labrat’s questions in more detail. “Why do teachers feel compelled to ‘teach to the test’? It seems to me a test should be designed to measure a student’s mastery of a certain curriculum. Are you arguing that these tests do not achieve this?”
One of the basic assumptions behind mandatory, high stakes testing (MHST) is that there is great value in being able to compare the test scores between not only individual students, but between schools, school districts and even states. For those that live and die by data—state and federal bureaucrats—this makes sense. For those that live in the real world, it makes none.
If I live in the Johnsonville school district and I discover that the Smithburg school district scored 4% higher on the MHST, will I immediately sell my home and move to Smithburg for the sake of my children? Education is not the delivery of a product 4% more effective than another. Relatively few people chose their place of residence based on schools. Surely, most people want the best possible schools for their children, but it’s rarely their most important consideration in choosing where to live.
The assumptions underlying such testing ignore the fact that all schools can do is provide the best possible educational opportunity. Education is always the responsibility of the individual student and their parents. If they’re not serious about the process, the process established in large part for their benefit, the best teachers and schools in the world avail them little.
The results of any year’s testing are utterly predictable. Most pass, but many kids who can barely string together a sentence also pass, while some truly excellent students, kids far above their peers in intellect and ability, genuinely excellent writers, fail.
Consider too that many students pass the 10thgrade English test, but cannot pass the 9thgrade test, which theoretically should be easier. In reality, passing such tests, for many students, is a dice roll.
Mandatory, high-stakes tests are massively expensive and contribute to the inexorable expansion of useless and intrusive government. Parents, students and teacher see them for what they are, so states withhold grade promotion and high school graduation to force them to pretend to see their value. For teachers, administrators and school districts, failing to demonstrate sufficient deference to their educratic betters can result in losing a career, or having the state take over a school district. Imagine the improvement that would cause.
This is why teachers must teach to the test. The scores determine, in many cases, whether they’ll be able to feed their families. Because of the perverse incentives established by the states, administrator’s careers are also on the line, and they become data lovers and slaves of the state education bureaucracy whether they like it or not.
State testing regimes are more about grading school buildings than grading students. I know of school districts where elementary schools do almost nothing other than test drills, all year long.
And because administrators don’t want to be surprised, they impose “benchmark” tests on teachers and students so that they can hopefully know whether students are improving in their ability to take the MHSTs, and will not cost them their jobs when test-taking time comes around. Unfortunately, this process usually costs a minimum of 10-20 class days a year to prepare for and take benchmark tests, and the results are even less useful than the results of the MHSTs because kids know they don’t matter.
Why are the results less useful? MHSTs usually are not given until April/May. Benchmark tests are given throughout the year. Unless kids are drilled throughout the year, they’ll be tested before they’re taught how to take the tests. The results cannot possibly give accurate forecasts about how the kids will do on the MHSTs, yet substantial time is wasted giving them anyway because they produce data, and more and more, data is all that matters to educrats, not learning, not good teaching and not the welfare of children.
Some states—Texas being one—recognize, finally, the lunacy of benchmark testing, and it has been, for several years, illegal to give more than two a year. Unfortunately, so addicted are many educrats, they knowingly violate the law and think themselves clever in calling benchmark tests “common assessments,” or “periodic tests,” or some other juvenile subterfuge. The state education bureaucracy knows the law is being violated, and does nothing.
When I taught 10thgrade, 95%+ of my students, year in and year out, passed these tests. What does this mean? I knew how to teach the kids to pass these specific tests, but it took me about two months of the year to do the mind-numbing repetition and drill necessary to stuff the very specific methods and tricks into my student’s heads. It also means that with two months of exposure to these drills, my students were able to perform at that level—on those specific tests.
That time, however, is lost to the kinds of writing, thinking and analysis absolutely vital to success in college and the adult world. Bad writing, once learned for the tests—and make no doubt, it is bad, formulaic writing—takes time to unlearn. Ask college English teachers whether their freshman students can write, and remember that the MHST mandates have been in effect for more than two decades. They’ve wrought great harm.
Consider too, that by the time the MHST is done, less than two months of school remain, and the kids are less then enthusiastic about expending enormous intellectual energy to unlearn what has been beaten into them in favor of real academic writing. They’re just drilled out. Many simply won’t try. For a great many students their sophomore year is a loss, thanks to testing.
Do these tests reveal mastery of a curriculum? Remember that students drilled in this way will retain about 10% of what they learn. Texas, like all states, has voluminous specific standards for each class, and teachers are supposed to teach each of them. In many cases, there are so many it’s actually impossible to teach them all, and if a teacher is silly enough to try, they might have one or two classes to devote to each. In English, we deal with about 80% of the standards each and every day, yet the state maintains the fiction that a given test question might deal with a single, specific, individual standard. The state will also swear that all a teacher need do for their students to pass a MHST is to teach the standards. This indicates they’re lying or have no clue of the reality of education or of the intent and construction of the tests.
The first year the last test series—the TAKS tests—was introduced, our test scores were far lower than usual. The same thing was true the first year of the new, improved and absolutely perfect STAAR tests, the tests that replaced the old, absolutely perfect TAKS tests. But why? Did we change the way we taught the standards? Did kids and teachers suddenly get stupid over the summer? The difference is we had no idea of the ultimate content of the tests, and the new tricks necessary to pass them. Such things are far more secret than nuclear weapons security.
I’m exaggerating–a little. Test security requirements are so bizarre teachers are told they’re not allowed to so much as look at the tests, nor may they discuss them in the slightest way, with students, each other, or with another living being. Yet they are required to handle them, somehow ensure that students are working on the right parts of the tests, and have filled out the right parts of the answer sheets in the right ways. We’re required to do “active monitoring,” which consists of wandering endlessly about the room during half day or full day testing sessions not looking at the tests we’re supposed to look at while not looking at them. This apparently makes sense to educrats.
And that’s not all. There are also mandatory “training” sessions where local school officials must actually read directions aloud to teachers who are, presumably, able to read the same manuals themselves. Fail in any of this and more and the state threatens one’s teaching certificate, without which one may not teach anywhere in the state.
Eventually, the state will release prior tests, and only then, with careful detective work, can we piece together what is actually required to pass the tests, and the best tactics. Unfortunately in Texas, since each year’s tests differ greatly, we are significantly handicapped in passing on effective test-taking tactics to students. This is why there is a multi-million dollar industry aimed at improving SAT test scores. Simply by knowing the best tactics for each portion of the SAT—and by engaging in correct practice–one can dramatically increase their score.
“Are teachers not given a set of goals a class is supposed to meet and a teaching plan developed to reach those goals? So is your argument against the metrics these tests are measuring or against using a test to measure whether your teaching plan achieves the goals of the class?”
Thankfully, most teachers are not forced to use pre-planned curricula and materials. In the same way that a plumber is not required to read from a manual whenever he installs a toilet, teachers are presumed to actually know something about their jobs. English teachers meet and agree on which novels they’ll teach in a given grade, about the kinds of instruction they’ll do, the levels of writing they must attain in each grade, the kinds of performance students will be held accountable for reaching each year, etc.
We do lesson plans, and generally follow a script about what we’ll do and in what order, but teachers must be flexible. Every year I was responsible for high stakes testing–I now teach Junior English; thank God there is no English testing for juniors–I lost easily 50% of my curriculum, probably more. I ended up cutting short many things and eliminating others entirely.
What was my ultimate teaching goal all those years? To cover the materials we agreed tenth graders needed to deal with to build bigger, better brains and to be prepared for 11th grade. To do this, I needed sufficient time to cover those materials and to do those assignments, the practice necessary for students to build the neural connections vital to the development of capable, functional, productive, tax-paying adults.
What’s my argument against these tests? Take the time away from the practice students need to develop properly from year to year merely so they can pass a single test, and they will not develop properly. That time can never be replaced and the opportunity is lost. There can be no corner cutting. It takes time for students to learn and grow. Remember, we’re talking about actually failing to make the neural connections that make us more capable and intelligent, all for the sake of producing a single data point, the result of a single test, very expensive in more ways than money. Take away 50% of a curriculum, take away 40% of class time—for any reason—and no teaching plan on earth, no magic curriculum, will prevent the destruction of the learning opportunity kids must have.
If teachers didn’t ‘teach to the test’, but actually taught the material the children should master, then would these children fail these tests? Why would that be?
Remember that the state doesn’t mandate the materials we use–-at least not yet; there are those who want to do that–-it mandates standards, which, if written properly, any competent teacher uses every day without having to think about or refer to a chart. The problem is it’s really difficult to grade more intellectually challenging assignments, particularly if you’re going to hire temporary test graders essentially wherever you can find them once a year. More on that next time.
Coming Next Week:
In the next installment of this series, to be posted next Monday, I’ll discuss how the business model of education figures into all of this, how teachers are ignored and infantilized, and how education bureaucracies are not only lying to the public, but are decidedly not its friend. I’ll also expose one of the dirty little secrets of this whole business: the tests are both dumbed down and written above grade level. That’s one of the primary reasons for all the secrecy. I hope to see you there.