In the first article of this series—“Testing Time…”, I explained the incredible amount of time lost to mandatory, high-stakes testing (and other matters) each year.  As I noted, my school lost as much as 42% of the available teaching time this year alone.  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 every class amounts to a loss of about 20 days of the school year.

The obvious questions: is mandatory, high-stakes testing worth the loss of 42% of the school year?  Is having a few data points—the results of a few tests—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 about “Testing Time…” 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?

I’ll see if I can answer those questions.

Why Do We Test?

As a teacher of English, I use tests for very limited purposes.  The primary value a test has is not—as most people think—to serve as a reliable measurement of what a student knows.  In reality, 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 care less whether they pass.  Others have a very broad base of knowledge and have to expend relatively little effort to pass.  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.

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.  True, 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 produced 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.

My school district—like most—requires that a certain percentage of a given student’s average be tests and a certain percentage be projects of some kind, so I am more or less bound to use tests from time to time as are most teachers everywhere.  However, a given test grade might have little resemblance to what most people would consider a test.  A test grade might be, for example, a critique of a short story or an essay about some facet of a book or movie (or both).  A history teacher might have more multiple choice or true/false tests, but better teachers require more writing and interpretation of history rather than simple recitation of dates and facts.

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 curriculums come with 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 intellectual development.

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 pretty 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?  This is surprising?  This is a genuine achievement?

Any competent teacher will know far more about the intellectual abilities and needs of their students within two weeks of the beginning of school than they’ll learn from any test.  By the end of each school year, my students will have completed more than 150 assignments, an average of a bit more than four per week.  My advanced placement (honors) students will average a bit more than five per week.  Of those assignments, only about 15% will be actual 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 really very little or nothing new in education, though many people make millions selling the same old concepts over and over again, usually by making up new language and acronyms.

Now we return to the reality that mandatory, high stakes tests cost 42% of available teaching time.  What is more revealing of a student’s actual learning and abilities, the knowledge of a teacher that has read and remarked upon 150 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 can reveal anything meaningful about a person, to the extent that test should determine whether they graduate from high school, and should take 42% of their available time for learning, the answer should be obvious.

Let’s consider Labrat’s excellent questions.  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 who 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?  Of course not.  Education is not the delivery of a product that is 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.  And 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 will avail them little.

The results of this year’s testing were utterly predictable.  Most passed, but many kids who can barely string together a sentence pass, while some of my truly excellent students, kids far above their peers in intellect and ability, fail.  In such cases, test scores are, at best, completely misleading.

Mandatory, high-stakes tests create a plethora of unintended consequences, and not a few horrendous, yet intended consequences.  They are massively expensive and contribute to the inexorable expansion of useless and intrusive government.  Because most students, and not a few teachers would simply ignore them, states must impose substantial consequences for those who fail to take the tests, and those that love them and their data, seriously.  For students, this usually means withholding graduation from high school.  For teachers, administrators and school districts, failing to demonstrate sufficient deference to their betters can result in losing a career, or having the state take over a school district.  This seldom results in actual improvement.

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 extensions of the state education bureaucracy whether they like it or not.  I know school districts where elementary schools do very little other than test drills, all year long.

And because administrators don’t want to be surprised, they always 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 therefore, not embarrass them or 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 usually, the results are even less useful than the results of the MHSTs.

Do these tests reliably measure mastery of a certain curriculum?  No.  Remember the limitations of tests compared with a year of first hand experience in determining a student’s development and abilities.  And they deliver this vastly inferior data at enormous expense, not only in money, but in time lost to more valuable endeavors.

My students, year in and year out, pass these tests at or near the 99% level.  What does this mean?  I am obviously capable of teaching the kids to pass these specific tests at that level, but it takes 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 at my hands, my students are able to perform at that level—on those specific tests.

Do these tests reveal mastery of a curriculum?  Remember that students drilled in this way will retain about 10% of that they learn.  Every state has standards of one kind or another.  Texas, for example, 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 specific, individual standard.  The state will also swear that all a teacher need do for their students to pass the test is to teach the standards.  This indicates that 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 this year, the first year of the new, improved and absolutely perfect STAAR tests, the tests that replace the old, absolutely perfect tests.  But why?  Did we change the way we taught the standards?  Not at all.  The difference is that we had no idea of the ultimate content of the tests, and the tricks and focus necessary to pass them.  This is all, you see, a secret, far more secret than nuclear weapons handling.

I’m exaggerating only a little—a very little.  Test security requirements are so bizarre that teachers are told they’re not allowed to so much as look at the tests, nor may they discuss them in the slightest way, among students, themselves 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.  In Texas, teachers are required to endure mandatory Internet videos about test security and “active monitoring,” which consists of wandering endlessly about the room during half day or full day testing sessions not looking at the tests they’re supposed to look at while not looking at them.  These videos are produced by people who obviously believe that teachers are not only immoral, but stupid.  And that’s not all.  There are also mandatory “training” sessions where local school officials must actually read directions aloud to teacher who are, presumably, actually able to read the same manuals themselves.  Fail in any of this and more and the state threatens your teaching certificate, without which, you 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.  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.  This is true for MHSTs too.

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?

Most teachers are not forced to use pre-planned curriculua 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.

Surely, we do lesson plans, and generally follow a script about what we’ll do and in what order, but teachers must be flexible.  This year, I lost easily 1/3 of my curriculum, probably more.  I ended up cutting short many things and eliminating others entirely.

What is my ultimate teaching goal?  To cover the materials we’ve agreed tenth graders need to deal with to build bigger, better brains and to be prepared for 11th grade.  To do this, I need sufficient time to cover those materials and to do those assignments, the practice necessary for students to build those neural connections vital on their path to becoming capable, functional 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 learning is lost.  And this goes deeper than students failing to learn a given body of facts.  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 1/3 and more of my curriculum, take away 42% of my class time—for any reason—and no teaching plan on earth will ensure that I achieve 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?

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 it 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.  But more on that next time.

Coming Next Week:

In the next installment of this series, to be posted next week, I’ll discuss how the business model of education figures into all of this, how teachers are ignored and infantilized, and how the state and federal 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 remarkably dumbed down.  That’s one of the primary reasons for all the secrecy.