More than any other structure found in the classroom, testing reveals the working educational philosophy of the teacher and the school. For example, in a traditional classroom where the Curriculum Transfer Model dominates, and where the purpose of school is to transfer the course content into the minds of students, tests exist to measure how well that transfer has been accomplished. They are therefore designed and graded with the intent of sorting students along a scale of being more and less successful at that task.
Consider a typical test being given in this traditional manner. The test is written in such a way that only a few students are expected to answer all the questions correctly. That is because 1) so much curriculum has been covered that only a few students have actually mastered all of it, and/or 2) the test has been constructed so that students will be sorted along a bell curve and only a few students are intended to get all the questions correct. Both of these causes violate the Prime Directive — they have interfered with the self-directed learning of all the students who didn’t get A’s on the test.
Clearly, if we believe that the purpose of school is to prepare students to live their lives well, our testing system must change to serve that purpose. That preparation includes every student genuinely learning the knowledge and skills that he needs to know in life. To that end, tests must be designed to determine how well a student has independently mastered that content and only that content. A test that is used strictly for this purpose is called a “norm-referenced” assessment. The Prime Directive dictates that all tests should be designed to serve this purpose.
Notice that a comparison with how well other students have mastered the material plays no role in such a testing system. Grading to a curve to get the right distribution of A’s and B’s is forbidden. Each individual student’s assessment stands alone.
But if we are only testing the essential learning goals, aren’t we “dumbing down” the course? We would be if we were still requiring every student to learn the same content to the same depth at the same time. But differentiation frees us from that arbitrary constraint. Since some students — given their backgrounds, abilities, and motivation — are clearly capable of mastering all of content covered in the traditional test above, that full range must be made available to them.
In this scheme, then, every student learns what is essential to know, and those students who are able to dive in deeper, for whatever reason, are free to, and encouraged to do so.
And what of the argument made above that tests don’t measure learning? It rests on the premise that false positives — acing a test and not learning very much — is acceptable to students. This is, of course, true for a student who is doing school; the high test score is the highest priority, and if he gets an “A” he has met it. But what if he is not doing school? When self-directed learning becomes the bedrock of the classroom culture, when it is the common purpose of students and teacher alike, then the fact that a high test score can be false positive becomes much less important. It is only one source of feedback among many, and when it is inconsistent with other, more realistic feedback, the student knows that he still hasn’t mastered the learning goals, even when he does well on a test. If he cares about learning, a false positive won’t deceive him because he no longer wants want to be deceived.
Many of the problems described in the last section arise from the way tests are traditionally used. They typically follow a unit of study and are considered results. Once the test has been taken, the class moves on to the next unit. The teacher may review the test with the class, but there is very little opportunity for a student to do anything about a poor grade. In educational jargon, these are called summative tests. They serve little or no purpose in the learning process itself. They are merely reporting the status of each student’s learning. Unfortunately, as we have seen, this leads to a number of unspoken consequences that have serious negative effects on students and how they learn.
There is an alternative. Tests which are integrated into the learning process can serve as important feedback that helps a student steer his learning. These are called formative tests.
A good analogy for this approach is how Girl and Boy Scouts earn merit badges. They are given a specific goal and some tools for reaching it. Once they feel they are ready, they are tested. If they don’t succeed in showing mastery of the learning goals, they aren’t given a grade. Instead, they now know specifically what they haven’t learned yet, and they work on that. When they are ready, they are tested again. If they are successful, they get a patch representing that goal, and they move on to the next merit badge topic.
Similarly, if tests are to be integrated into the learning process, they must not only measure how well a student has mastered the material, but must also serve as feedback about what a student still needs to learn. Unlike summative tests, which are given at the end of the learning process, formative tests are built into the learning process. They are the starting point of helping a student to learn from his mistakes. Tests identify specifically what he has not yet mastered.
The key word in the last sentence is “yet”. When tests are used in this way, the student knows that if he is struggling at the time of the test, he will be able to continue working on the same material until he is no longer struggling. For a student who hasn’t yet mastered the material, it is simply the first assessment, which leads directly to remediation, focussing intensely on the material he knows he must still master.