by John Jensen, PhD
Is there a solid, rational, evidence-based reason to limit high-stakes testing?
In my previous articles about it, I suggested one: Students' motivation to learn is the linch-pin of their progress. We should stop doing what undermines it. We should drop tests that discourage, embarrass, pressure, or threaten them.
If we need to collect information about their learning for decisions made outside the classroom, we can leave students anonymous. To make clear that a test is meant for understanding the group as a whole, they don't put their name on it. Details about personal progress remain between teacher, student, and parents, where diagnostic/formative tests help provide guidance and immediate steps can address needs,
But without the pressure of high-stakes tests, what does education look like?
The question may seem strange, as though classrooms hold a mystery. In fact, for years testing has defined much of what we do. Without it hammering at us, we may be at loose ends and call to the teacher in the next room: "Hey Victoria. Can you remember what we were doing before tests took over our classroom?" What else is there?
The best alternative to test-organized learning, I believe, is explanation-organized learning. Explaining fuses together memory and sense-making, both of them standard goals of education. To appreciate the difference between test-organized and explanation-organized knowledge, we can use a section on the Civil War. Imagine a fifth grader standing up and saying:
Grant and Lee had this thing going of trying to figure out what the other would do. This head game was especially important for Grant because Lee was usually a brilliant tactical commander and had carried the fight north. We think of the south as Mississippi, but big battles like Fredericksburg, Antietam, and Gettysburg were all within a hundred miles of Washington D.C.! Lee knew he didn't have forever. He had to knock out the Union Army, whereas Grant felt that with the greater resources of the North, he could eventually grind down the Confederates.
Contrast that with test-structured learning:
1. The top commander of the Union forces in the Civil war was_________ and of the Confederate forces was________.
2. The objective of the Confederates was__________, and of the Union forces was_____________.
3. Name three battles (__________, __________, and ________) within a hundred miles of an important northern city (_____________).
With nearly the same information in both, note the personalness and integration of ideas in the first compared to the sterility of the second, and think which would appeal more to students. The first implies that students can think and talk about a network of ideas, which is a clue to genuine knowledge–the ability to think about something without help.
On Wednesday you might have said to them, "Today and tomorrow I want you to explain back and forth to a partner everything we've studied about the Civil War. On Friday, I'll draw names at random and ask you to stand and explain parts of it." Using only classroom time, you organize knowledge so they (1) integrate their thinking, (2) sustain interest, (3) practice to gain mastery, (4) demonstrate mastery, and (5) develop ownership. Paring down the same ideas into isolated answers breaks up the natural links between ideas.
The shift away from narration of knowledge and the consequent loss of interest came home to me one day at a public library while browsing through books about math instruction. An old text explained concepts through the history of their development. Challenges in their lives moved mathematicians to pursue particular ideas. As I read into it, interest spurred by real-life time and place worked like a current carrying concepts along. People's experience constituted a framework within which mathematical ideas could easily be recalled and developed.
So why would curriculum designers ignore such interest-generating details? My guess is due to objectives too limited. When you believe students are in serious deficit and you want to give them at least something, you first provide them basic answers–doing problems according to formulas. For mathematicians, these would be only the conclusions of a long experience of effort. But upon noticing that answers installed to pass a test are not an adequate foundation for adult thought, we raise our standards and aim to teach more than bare-bones conclusions. We need them to think mathematically—i.e. able to explain what they learn.
In the same library, I found a book with several dozen short chapters, each a conversation with someone about the challenges in their occupation, and how they met them daily.
The narratives were fascinating. People's response to their job expressed heart-felt values, and their thought processes were intriguing. Every occupation became interesting as people explained how they used their imagination and energy to work past obstacles. To interest students in a job they might pursue, we can assign them to learn someone's explanation of their job and explain it to another student. We stimulate their thought by arranging for them to explain what they understand, which is a critical point. Knowledge comes alive when students put to work their own ability to assemble and express it. .
Contrast this with test questions that supply information boiled down to four answer-options and the student places an X by a choice in the array instead of explaining the array itself. Students are robbed of appreciating how humans use knowledge to form their world. They are constantly left to react to others' ideas.
Explanation is a better framework for student effort than are simple memory, selecting from options, or transferring knowledge from one paper to another. Expressing ideas builds a lifetime skill, sometimes drawing more on memory and sometimes more on sense-making, and also changes the social atmosphere of a school. If you present an idea to students and tell them, "Get a partner and explain this to each other so you both know it," explaining becomes a bridge to relationships. Having another listen to our ideas cues our social instincts, stimulating us to elevate the quality of our thinking so the other grasps what we say.
Upon designing instruction so students explain what they learn, we need to grade it without creeping back to the very testing we tried to escape.
In our first article, we noted the solution of counting up sections completed. We can divide the curriculum into lessons, each having defined effort and a specific signal of completion. We check these off one at a time, and students graduate when their list is finished. Because the actual residue of learning retained varies among students, however, we could use a more exacting standard.
Three conditions solve the problem: (1) Credit effort for points learned instead of subtracting credit for points not learned. Stop tallying mistakes against an arbitrary standard. (2) Count up maintained knowledge instead of temporary knowledge. An idea learned receives one point of score the student keeps by retaining the learning itself. (3) Award a point of score for every new point of knowledge that took independent effort to learn. In sum, maintain positive points-of-knowledge, and count them up.
Maintaining knowledge constitutes a fundamental redirection of U.S. education. The transitory nature of what it purports to measure is an inherent problem with current testing. Nearly unquestioned in the education community is the assumption that knowledge appears and disappears. Once accepting this, we then depend on testing to drive students sporadically to "study hard" so we can catch at least some of their knowledge on the fly.
A student takes a 10-question quiz on a new section, and gets 2 wrong. On a monthly test on several sections, he corrects some errors, makes more, is preoccupied with other activities, and scores 75%. For a "final" (which he anticipates because then he can drop the subject), he knows he must bring up his scores. Some knowledge appears, other disappears, and he scores 85% with his knowledge-wave at its peak. But if he were given the same test a month later, his score would drop to 65%, from a low B to an F.
Once presuming that we cannot fix knowledge (like "fixing" a photo), we must continually retest as the wave of knowledge rises and falls, and constantly repeat even superficial knowledge that soon dissipates. The alternative is designing instruction so material is retained permanently from the start (I explain how in detail in my books, cf. below).
For scoring this accumulating knowledge:
1. You may wish to award points for verbal explanations like you grade a written essay question. In the Civil War illustration above, you might tell them, "In explaining this section, include eight details or factors, and you have a score of eight." The details each required independent effort to learn, and would have counted against the student if gotten wrong on a test. But instead of waiting for a mistake on a test, you award one point of score for every point of knowledge a student produces either for you or a partner. In the explanation above, a score of eight points reasonably approximates the work that went into constructing it.
Explaining to a partner helps piggyback learning on students' drive to show off their competence to peers, measure up to peer standards, and count up each other's gains accurately. Students let you know who really knows what.
2. You can also score verbal explanation by time spent at it. When I first suggested this to several students, one came to me the next day. He had watched a TV science program the night before, and as I timed him gave me a 19 minute discourse on the chemistry of the sun. The length of time one takes for an explanation (minus undue repetition, etc.) is an objective measure you can post. It appeals to students the same way they already count progress on many personal activities. Explaining-time and the accumulation of points of knowledge are both objective measures that correlate with effort and hence are appreciated by students.
3. Accounting for precise details accords with the mind's natural bent. To learn anything, we focus on one aspect at a time even while organizing many. We nail down this piece and this and this, learning subjects in a step-wise fashion. All math is one step after another, each a new point of knowledge. The glossary of a middle school math text may contain 250 terms–a perfect year-long project one term at a time. Word meanings, spellings, rules of grammar, definitions of important words in all subjects, parts of wholes, steps of processes, formulas, key details, important dates and events are all learned one point-of-knowledge at time, and can be practiced, retained, and scored the same way.
4. By counting up points of knowledge and time-explaining, at the conclusion of a school year you can collect the data in an information-rich, one-page Academic Mastery Report summing up everything a student knows. Divide subjects into sections and report the maintained score of knowledge for each. (In my books noted below, I explain how to construct this report).
5. If such changes seem to require too much accounting, consider what testing does now. It counts up micro-details against students, while their natural orientation is to accumulate such details for themselves. Once you show them how to demonstrate their learning, score it, claim it, and monitor each other's progress through partner practice, they readily cooperate and relieve you of detail work you now supply yourself.
6. If this approach sounds too novel, you can observe its effect on students within a couple weeks. You'll note that they respond eagerly because it meets their needs and there is no emotional downside to it. Because students know the increment of effort they exerted in order to learn a new point, they appreciate the logic of receiving a point of score for it. We invigorate them by counting up their increasing knowledge point by point.
There is no defensible, pedagogical rationale for counting mistakes. Post everyone's cumulative progress on a wall chart so they can inspire each other, and set a quantity of points of knowledge achieved by all together that will earn them a party.
John Jensen is a licensed clinical psychologist and author of the three-volume Practice Makes Permanent series (Rowman and Littlefield). He will send a proof copy of the volumes to anyone on request: [email protected]