One of the arguments against greater school accountability and performance-driven school reform is that the introduction of more tests to measure student performance will lead to more teachers “teaching to the test.”

I’ve been thinking about this a lot lately as our district’s first periodic assessment is coming up this week, and us teachers were given a copy of the exam so that we can prepare our students for it. I’m a bit nervous about my students taking the exam as we are behind where most other students should be in an Algebra class or a Geometry class. After all, we lost almost a month of instructional time at the beginning of this school year due to constantly shifting enrollments.

Do I prepare my students for the test or not, and if I do, how will I prepare them? It’s not just a matter of wanting my class or my school to have higher scores–if students see their low scores they will lose some of the precious confidence in their math skills we’ve been building so far, and confidence (specifically, belief about self-efficacy) matters a lot.

So I’ve been wondering: what exactly is “teaching to the test” and what is it about “teaching to the test” that carries negative connotations? Standardized tests often ask students to think about mathematics or to decode mathematics presented in peculiar ways. If the content of these questions is something far removed from a teachers’ usual curriculum and the teacher is compelled to help students acquire that content, is that bad? What if that content is potentially helpful or useful for students’ mathematical development?

Consider one of the test questions on this week’s periodic assessment: (I’ve modified it slightly for this blog.)

Which is greater, 5x+1-(1-2x) or 6+2x-1-(2x-4)?

A. 5x+1-(1-2x)

B. 6+2x-1-(2x-4)

C. Not enough information

D. They are equal

This question requires students to be able to combine like terms, to be proficient with positive and negative numbers, and to reason about which of the two expressions is greater. We’re still working on combining like terms and adding and subtracting positive and negative numbers in class so would that be considered “teaching to the test”? (Doesn’t seem like it.)

To help my students answer this question successfully, I would probably have to start them off with questions such as

Which is greater, 7x or 9?

A. 7x

B. 9

C. Not enough information

D. They are equal

If I pose questions like this, am I “teaching to the test”? Seems to me a moot question. If I find the content compelling for students’ mathematical development, then I will help them learn it. If not, I’ll let students struggle with the question and let the test question do its intended job: give me information about my students’ understanding and skills.

I think what I’m trying to say through this post is that “teaching to the test” is a catch phrase that galvanizes people in educational debates, not unlike the phrase “public option” in present U.S. debates about health care reform. It seems to me that the kind of teaching that tests encourage is mostly determined by the content of the those tests. Test writers have a great deal of power as they can shape what is taught in classrooms. A bad test could potentially fill math classrooms everywhere with useless, unimaginative content. A test could even be used to circumvent state and district standards for mathematics as it can establish a de facto curriculum. And let’s not forget that developing a good test can be expensive.

Closing side note: What I find fascinating is how teaching happens on so many scales at once. The public and academics mostly talk about teaching at macroscopic scales (districts, curricula, schools, classes) but when you think about what actually happens every day in every classroom, it’s what the teacher does and says and how students process it that really matters. How will Teacher X’s beliefs about what mathematical topics are important affect the way she or he prepares students for a standardized test? How does Teacher Y’s beliefs about the way students acquire knowledge affect the way she or he implements a curriculum? Will Teacher Z maintain the high level of cognitive demand in a rich curriculum like CPM? And who is watching/can watch what happens in every class every day?

UPDATE: A friend astutely reminded me that in the “backward” teaching design process, we (1) think about what it is we want students to learn, then (2) determine acceptable evidence for students achieving those goals, then (3) plan learning activities and experiences. (I believe strongly in this design method and try to follow it when I can.) This design process necessarily involves preparing students in step (3) to do well on the assessments determined in step (2) so if that is considered “teaching to the test” I would be guilty as charged.

There was one year in elementary school when the most common thing my teacher said was “And you had better remember this because it is going to be on the TAS test.” (TAS being the Texas equivalent of the STAR test given in CA). I have no idea what we did that year, but I know it must have been on that test.

We spent I don’t know how much time in HS English classes learning /how to take multiple choice tests/. Not learning the material that was going to be on the tests, but strategies for taking the things. I was pulled out of classes in HS (despite my protestations against it) and sent to what amounted to PSAT/SAT preparation seminars where we talked about test taking strategies. Not about the material that we would be expected to know, but about how to take the test.

The AP Stats class that I took focused on doing old AP stats problems rather than on actually understanding what we were doing and how to do it and why it worked. The focus was on the kinds of problems that were expected to appear on the test rather than on the concepts. I was interested in concepts and figured them out mostly on my own. Most of the class didn’t. I got a 5 on the exam, most of the class got 1’s and 2’s. (so in this case the teacher was trying to teach to the test and failed miserably even at that)

If the outside tests that your students have to take are over material that is actually relevant to what they are supposed to be learning and you talk about the kinds of problems that will show up and the concepts behind them, that’s great. If you were to drill them on the kinds of problems that they should expect to the point where the concepts are ignored, that is less ok. If covering the material that they need to know for the test means that you will not be able to cover other material that they should get out of your course, that is also less ok.

Overall I agree with your view on the tests and if the tests were honestly well written and all teachers approached them the way you sound like you will, I might have come out of the public school system less cynical about standardized tests. Unfortunately, my experience was that the tests are often not that well written and that there are far too many teachers who approach them poorly.

A great post for an unenviable quandary. If there’s one thing I’ve learned in my informal education on policy, it’s that certain phrases and words – often poorly defined, and often deliberately left so – do become triggers to short circuits in reasoning. I’d never thought about “teaching to the test” and what it actually means, so thank you for bringing it up.

It sounds legalistic, but I think intent of the teacher plays a role in whether it’s perceived well or poorly. I think one teaches to the test if one sees the test scores as the ends, and not simply a metric that simply has the virtue of being standardized and quantifiable.

I’m guessing there’s a group of teachers who teaches to the test because of aligned incentives for having students do well (or not do poorly). There may be a (hopefully larger) group of teachers who do want their students to learn, but have internalized a belief that the test dictates what is and is not essential material.

A very small class of teachers – often the best, and often the most burned out – see the test as a legitimate measure, but not the final word either in course content or in student comprehension.

And then I suppose there are teachers who write off the value of a standardized test simply because it is standardized and a test. I’d have to think about whether these teachers do more harm than teachers who teach to the test.

I’d say the first two teach to the test, while the second two do not – again, using intent. In this sense, I understand its pejorative meaning.

Based on my experiences in the Fourier class I took from you, I suspect you strive to be here, and suffer for it. I admire your effort, and hope that you find a way to measure what is no doubt considerable success relative to where your students would be if they had a less talented and passionate teacher.

Best wishes.

This comment really resonates with me. Here you prove that carefully constructed multiple choice questions can be a useful tool in developing and assessing mathematical thinking.

One of the things I love about talking about pedagogy with you is how deeply you think about *why* we choose to teach what we do and then *how* we might approach the subject. I think in the past, we haven’t talked as much about how to assess whether students have acquired the skills we are trying to impart. I look forward to thinking about this more in the context of courses we teach.

I was thinking a bit more about the incentives for teachers to “teach to the test,” and how this external incentive can short-circuit other ideals teachers should strive for. In my experience/reality, teachers “look good–” and are sometimes paid better– when their students score highly on tests designed from some outside source. One way to get results is to over-prep students for these tests like Sarah mentioned in her AP Stats class. But such an approach comes at a cost: students’ appreciation of the subject, their long-term understanding of the subject, their interest, motivation, and engagement with the important ideas. I think, however, that the best teachers can integrate a thorough knowledge of both the subject matter AND the nature of standardized assessments to design experiences that motivate, engage, educate, and prepare students. This takes a lot of preparation on the part of teachers: subject preparation, understanding of test makers’ intentions/perspective, and some design ingenuity.

My idea of what “teaching to the test” is seems to differ somewhat from yours, probably because to me it does carry a bit of a negative connotation and I feel like a certain amount of teaching to the test is really necessary for people to do well on tests.

I agree with a lot of what Sarah said… namely the bits about AP Stats (and other AP classes) putting more emphasis on working previous AP questions than learning the actual material. The same goes for PSAT/SAT prep sessions. I suppose multiple choice test taking is a skill, but if a student is taught to be able to pick out answers based on similarites in answer choices rather than the actual material, that is sad. Also incredibly frustrating — I had a friend in high school who was very good at taking tests, and as a result passed out of AP Calculus. He never actually learned the material, however, and was constantly asking me for help with even the simplest integrals for his homework.

To me, teaching to the test (in terms of material rather than mechanics of the test) applies more in non-math subjects. In math you could memorize formulas and such, but there is still a level of critical thinking you have to be able to apply to answer questions on the test. On the other hand, one of my distinct memories from high school is taking a Latin examination with others from other schools in the area. My school did very well, and the popular opinion was that it was all the more impressive because our teacher did not “teach to the test” as the teachers from the other schools did. Really what that meant was that they were drilled and made to memorize facts that were useful only really in the context of the tests. Such rote memorization can only get you so far in math, I feel, and thus perhaps at least my definition of teaching to the test applies less to math.

I guess what I really mean to say is that I think you’re right that as long as the test is good, teaching to the test (at least in the manner you are implying) is perfectly fine. I probably could have said that much more concisely…