Congress stomps on teaching to the test

I’ve mentioned more than once that my own somewhat positive attitude toward “teaching to the test” is driven, in part, by my experience teaching Advanced Placement courses.

All successful AP teachers – and by that I mean not only teachers whose students earn decent scores, but also teachers who inspire real learning – teach to the test. We also all grouse about the tests, and the College Board test writers. They require us to cover too much material. They surprise us with what seem to be peripheral or ambiguous questions. They are too politically correct, or too oblivious to urgent social issues.

Still, as tests go the AP tests are pretty good. The questions don’t just ask students to regurgitate information, but to also to analyze data, language, historical events, etc. Essays count for half the grade, at least in the subjects I’ve taught, and the essay questions require students to apply and not just download what they’ve learned. In my experience, and I think most AP teachers would agree, there’s a pretty strong correlation between how much a student has learned (and how hard he or she has worked for that learning) and the final score on the test. As we grapple with the tough issue of how to assess the common core, we could do worse than to look at the AP model.

As Washington Post education blogger Jay Mathews points out,

Those difficult exams, full of essay questions, not only give high- schoolers a taste of college trauma but also motivate harder work in the course by them and their teachers. The exams are written and graded by outside experts and cannot be watered down to save the reputation of students who didn’t do their homework or instructors not up to the job.

Still, AP courses pose a dilemma for teachers, and for schools. There’s considerable evidence that taking AP classes benefits even students who don’t end up passing the course. But students who aren’t obviously “AP material” are a lot of work for teachers, and they can also drag down class scores. Just how narrow should the AP gate be? I come down quite strongly on the “you’re welcome as long as you recognize and accept what you’re getting into” side of the debate, but I also work for a principal who doesn’t assume that every student who takes the exam is going to earn a passing grade. Not every teacher is as fortunate.

But there’s another gate that winnows out potential AP students: the $87 test fee, which students and their parents usually pay. It’s a bargain price for college credits . . . but also a barrier for many families.

Late last year, as Jay Mathews recently reported,

Congress sharply reduced funds to pay Advanced Placement and International Baccalaureate test fees for low-income students. The legislators chopped the allotment from $43 million to $27 million, a gross lapse in legislative insight.


I completely agree. These kids deserve a chance to be taught to the test.


  1. Richard Donaghey

    AP courses will become a dinosaur in the world of education. The paradigm is changing to what is called Concurrent Enrollment or Dual Enrollment. This where a student takes a college course in high school and recieves college credit for that course. The argument is why take an AP course, when you can get college credit in the first place for taking that course while still a high school student. There are some students who by the time they graduate from high school, will have already completed their Freshman year of college. Thats over 30 hours of college credit that they can transfer to the college or university they decide to attend.

  2. Mary McConnell

    If AP is going to become extinct, it’s a shame. I teach concurrent enrollment as well as AP courses. Sorry if I’m stepping on toes here . . . but the AP test sets a much higher bar than most concurrent courses, which are essentially entry-level community college classes. Both are great, since they help students both prepare for college and earn some college credits inexpensively. But the rigorous outside assessment that Advanced Placement provides drives excellence.

  3. Yak_Herder

    I agree with both comments. Concurrent Enrollment is swallowing up Advanced Placement, and AP is far more rigorous.

  4. Fred 44

    Only in the world of public education would someone view “teaching to the test” as a bad thing. I have a dream that sometime somewhere a rationale person will stop for just a minute and think through what “teaching to the test” means, rather than get all emotional because some politician or talk show host said it was another evil thing that teachers and their unions do to promote the ruination of the public school system. Instead of “teaching to the test” being something evil, we should actually fire teachers who don’t teach to the test! Mary I applaud you for teaching to the test, that is an excellent teaching strategy.

    What is the purpose of a test? Isn’t it to measure what a student has learned? Shouldn’t that test be based on a what a student SHOULD HAVE LEARNED? With those two thoughts in mind, shouldn’t the curriculum be aligned so that a teacher is teaching students what they should know to pass a test on what they should know? I know that sounds like a bad sentence, but I know my teacher friends and probably many others will understand what I am trying to say. My point being the curriculum and the test should be aligned, which means that the teacher is teaching to the test.

    Many if not most of our year end tests do not have testing aligned with the curriculum. When students do poorly, teachers are given reports with limited and very abstract information on what students did well on, and what students did poorly on. As a teacher I am forced to guess much of what students did not know without any help in figuring out why they did not know it. Did I not cover the material, did I not drill down deep enough, did I not use the right the same terminology that the test designers used? Syntax can often be a crucial difference in a students success or failure on a test. The test designer may use terms that have the same or similar meaning, but are different than the ones I use as a teacher. If I can’t see the test or least a test pool of questions, I can’t see the terms the designer used, the way that they constructed questions etc. Improving scores could be something as simple as changing a couple of terms that I use and replacing them with the terms the assessment uses.

    Sometimes these tests become more chance and luck of the draw than anything else. For example, I teach a Career and Technical Education course (CTE)in the state of Utah. The course consists of 12 Units each Unit consisting of approximately 15-18 objectives. For the year I am expected to teach over 180 objectives. The final test consists of 80 question (not even one question per objective). I would argue that this is not a valid test of what my students know because there should be multiple questions on each objective. This is not my curriculum, it is a curriculum provided by the state. In one Unit there are 18 Objectives, over 150 vocabulary words and PowerPoint presentation with 198 slides. The final test (that I will fired if I look at)has 4 questions from this unit. I would ask you is that measuring what a kid knows about that unit or simply whether I got lucky and emphasized the 2 or 3 things that the test designer thought were important.

    I wish the above scenario were an exception, but my 25 years of experience tells me it is the rule not the exception. What if we gave teachers twenty objectives to teach for the semester/year, and then had 4-5 questions from each objective? Wouldn’t that be a better measurement of what a student had learned and what a teacher had taught.

    One of the reasons I am opposed to standardized testing being part of teacher evaluation at this point of time is that quite honestly we have very few quality standardized tests in any subject, let alone across the board that we use to get an accurate reflection of what is being taught and what is being learned.

    The solution seems so simple to me. We must first make a determination as to what I student should know when the leave a course. We must then create an assessment that will measure what we expect students to learn that involves multiple questions and or essay questions on each standard and objective. We must then create a timeline for teaching the material. At that point will we be teaching to the test? Absolutely and if a teacher is not teaching to the test fire them!

  5. Winglish

    What in the world is a test supposed to be measuring if not the material that has been taught in class? Of course teachers should teach to the test! They should teach to every test. No test should be a surprise to students. Tests are assessments of knowledge acquired in the classroom. It seems only too obvious to me.

  6. Yak_Herder

    Teaching to the test shortchanges the students if:
    1) The test and the standards it is derived from are poor (to say many of them are poor is being kind), and
    2) The MINIMUMS established by those standards (and test) devolve into the MAXIMUM.

    Fred 44 wrote:
    “The solution seems so simple to me. We must first make a determination as to what a student should know when they leave a course.”

    We haven’t done an adequate job of making that determination. Utah’s standards for most courses aren’t anything to brag about. A significant debate is underway right now with respect to a Common Core for science. A significant stumbling block to its development is an issue over how to address evolution and natural selection. Statistics show a huge bias among members of the scientific community in regard to their personal beliefs regarding God. People living in Bible Belt States in particular are pushing back on that. Cue Spencer Tracy and the movie “Inherit the Wind”. Oddly, scientists (you know, those pesky folks known for establishing “fact” based on the evidence) are falling back on a strategy of “consensus”. Simply put, they want something included in the common core because the majority of scientists think one way. Welcome to the dark ages.

    That (poorly developed/written standards) is just half of the problem. The next problem is to figure out how to measure performance towards the goal (increasing learning) without the measure itself becoming the goal (raising a score). The latter so easily displaces the former. As attention shifts towards scores on standardized tests and school averages, I already see that shift happening. As it does, the pressure increases to “Teach to the Test”, streamlining the curriculum to maximize efficiency to that end.

    That’s another way of saying we’re stripping out all of the “good stuff” that high caliber teachers add to the curriculum as a matter of course (pun intended).

    Stop and think about that.

    An inspiring teacher and what they “add” is THE biggest factor in a classroom promoting good learning. In our efforts to improve things, we are actually stirring things up in just right manner to neutralize our best asset. Once we tie salaries to that performance, we will have sealed our fate.

    Let’s not focus on the tests until after fixing the standards. Then, let’s be careful about how we employ them.

  7. Carolyn Sharette

    I am so happy to agree with BOTH Fred and Yak! The core issue is that we don’t know what we want students to learn, to what depth and level of mastery, and when (clear, rigorous standards). Standards are most often too broad, too numerous, vague, etc., just as Fred outlined.

    I don’t believe it is possible (or even necessary) to get to standards that everyone can agree upon. The current heating up of the fight against the common core is evidence that NO MATTER WHAT – even with standards that are clearly better than the ones we currently have (in Utah the language arts standards are quite a bit less rigorous than the new CCSS so it will be an improvement for Utah students to be moved to the new CCSS) people are going to find reasons to argue and fight over standards and objectives FOREVER. I really mean that. I don’t believe there will EVER be a consensus by parents, teachers, districts, and the state on what students should learn by when and what mastery means in those areas.

    That is why school choice is critical. A school needs to be able to set its standards, advertise them, and implement them and then publish their failure or success (test scores, college entrance rates and locales) in helping students reach them. And parents should be forced to choose the school their child will attend.

    Were this our model, standards would quickly evolve based upon their effectiveness and nothing else (politics for one thing would be left out of it). Schools with students who are more able to compete for college spots than others (or have more students entering the workforce successfully into apprenticeships, if that is the school’s focus) will find their enrollments growing; schools with poorer standards and programs will find their enrollments shrinking until they improve. I see no other way other than the private sector – parents choosing the education “product” they want for their child – to accomplish having rigorous, effective standards in the majority of schools.

    I don’t think AP will be going anywhere soon. The students most ready for college rigor are still the ones that can pass the AP tests with a 4 or 5, and the colleges know it. Concurrent enrollment tells you little or nothing about a student’s capabilities with regard to college rigor (because the concurrent courses are not necessarily rigorous). So I believe colleges will keep the AP program as an important enrollment piece, which will keep the program in the high schools.

Leave a comment encourages a civil dialogue among its readers. We welcome your thoughtful comments.