If you are at all involved in education, whether you're a teacher, a student teacher, a student, or a parent, you have probably heard the popular alliterated phrase "teaching to the test." Until World War II, standardized tests didn't even exist. The SAT was first created to filter the surge of college applicants resulting from the GI Bill. When my mother was a child, elementary school students did not have standardized tests. Until the 1970s, publishing test results in newspapers just wasn't a thing.
We went crazy over standardized testing when George W. Bush signed No Child Left Behind into law. Now, when you hear "teaching to the test," you know that it is supposed to be bad, even if you don't know why. Logically, if we test students on what they are supposed to learn, teaching to that test should not be a problem, right? And we want to measure how well teachers do because we don't want them to slack off, right? Unfortunately, it's not that simple. Let me break it down and explain why teaching to the test is bad for our education system and bad for the future of America's economy.
What do these test measure?
The first thing to understand is that tests don't have validity, inferences do.
Validity is how well a test consistently measures what it is supposed to. An inference is a conclusion drawn based on information or data. A set of test results is just data. So the test, being pure data, cannot be valid. Instead, the inferences we make based on that data are what have validity. Example:
A multiple choice test on the characters of Harry Potter is not valid by itself. If the test covers a wide range of characters in a balanced way, the inference that students who score well on it have knowledge of Harry Potter characters is valid. The inference that those same students also have wide knowledge of the themes of Harry Potter is less valid. The inference that they are also good at biochemistry would be completely invalid.
Standardized tests-- like a Harry Potter test is not meant to measure knowledge of biochemistry-- are not designed to measure how well a teacher teaches.
Bad teachers.
The idea that a pencil and paper test is the magic barrier keeping bad teachers from teaching is absurd. Giving a list of facts to teach will not magically improve a poorly trained teacher, and the assumption that bad teachers are bad because they don't want to teach kids is insulting. Teachers have one of the lowest paygrades for the level of education required and work long hours planning and grading. People are teachers because they want to be teachers. If they aren't, the natural cycle of teacher turnover and administrative observations will weed them out far better than a scantron test created by a corporation.
Apples and oranges.
Classroom exams are made to compare students to content. The goal is for everyone who knows the material to score 100% and students who don't know it to fail. Mastery of the content and exam is ideal.
Standardized exams compare students to students (not content). These tests are not made to provide pass/fail results. The items that most students do well on are useless in differentiating between students, so they are omitted from the exam. These items are usually basics and foundations.
If we do not include foundations and basics on our test, then teaching to that test gives students superficial knowledge that lacks the roots of understanding and therefore will be shortly forgotten. We are already seeing this with the AP exam. In order to differentiate between a group of already over-achieving high school students, questions such as what is Calculus? are left off the test. Yet we can all agree, even if we are not math teachers, that students testing out of introductory level Calculus should understand what Calculus is. Because of reasons like these, college entrants trained to answer a bunch of questions without really tying the information together and building a foundation, many colleges are now deciding not to accept the AP exams.
Individualized inferences.
Standardized tests are fantastic for individual achievement inferences. They are not designed to test teachers. A combination of High School GPA and SAT scores are still incredibly good predictors of first-year college performance for individual students. Standardized tests have a point, but publishing scores in the newspaper to foster competition between schools is not the point.
Standardized tests with no consequences for students are unfair measurements of teaching because-- in many cases-- the students have no incentive not to blow off the test. In fact, it is much preferable to bubble random answers and take a nap. The types of questions we find on standardized tests often only test lower-level thinking. A student might correctly identify who wrote the Declaration of Independence, who signed it, and who made the famous ride to warn that "the British are coming!", but that does not guarantee that they will understand the underlying causes of the American Revolution. Which do you think is more important?
Then how do we test our education system?
So yes, these tests do have their purposes. A well-written reading comprehension test is valid for determining which students need remediation in reading. The point is, if we make the jobs of teachers hinge on their students bubbling correct answers on a scantron, the quality of our education system will suffer. Students will leave school with a memorized list of facts, but they will never really learn anything of importance. Memorizing the names of political figures won't help them decide who to vote for and knowing the meaning of "remediation" won't help them determine the meaning of the next unfamiliar word they come up against.
Observations and continuing education are vital to strengthening our education system. All of our focus is placed on the quality of our teachers, yet teachers aren't deciding the curriculum taught. As long as we're cutting funding to struggling schools and letting politicians decide what to cut from textbooks, our children will suffer. There are ways to evaluate the effectiveness of a teacher, but I guarantee you it has little to do with making dots with number two pencil.
