by John Jensen, PhD
An insight about education in general lurks within testing — if people care to look. It concerns the degree of match between what is learned and what is assessed. To take practical examples:
- A pole vaulter practices pole vaulting a thousand times. The meet is his test. He pole vaults again.
- A marksman shoots a thousand bullets in practice. The meet is his test. He shoots again. He does what he has practiced.
- A chess enthusiast plays a thousand games. The meet is his test. He does what he has practiced.
People practice beforehand exactly what they want to reproduce later. While practicing, the three people above may concentrate on specific thoughts and actions needed in phases of their demonstration. It is not automatic, for instance, that a pole vaulter plants his pole precisely in the tiny groove in the track from which he pivots upward. Going at a dead run, he must apply attention and practice. In this and all sports, however, the final test of playing the game for real incorporates the nuances practiced. You don’t practice one thing and test something else.
This is our problem with high stakes testing. Prior to the test, instruction is off doing a smorgasbord of things. Then to find out the presumed value of any of these activities, we test—by doing something different from the learning activity, failing to incorporate the nuances that were practiced, and demonstrating poorly those that were.
John Wilson, former head of the NEA, commenting on the effects of high stakes tests, points out: “The unintended consequences of emotional distress among students, cheating by stressed-out teachers and administrators, suicide by even a single student or teacher, resource discrimination based on zip codes, and the dissuasion of great educators from accepting assignments in high-poverty schools are all prices too high to pay for what little value may come from these policies” (Stop the Madness of High Stakes Testing” Edweek online, October 22, 2012). Add to these effects the sheer difficulty of making the tests valid—reflecting what students actually labor to learn.
The instructive point about testing, the assumption on which it is based, is uncomplicated — that students’ expression of their learning validly reveals what they learn. We hear, read, and watch what comes out of them to find out what is in them. In the face of such an obvious principle, why is there any argument? The problem lies in the mismatch between what they practice and what is tested. Imagine, for instance, a high school football team that practices making touchdowns. They arrive at the big game to be informed that the winner will be the team that has the fewest penalties. Or a baseball team accustomed to practicing making runs is informed that the winner will be the team with most players reaching first base. If we teach for getting to first base, it’s inappropriate to test for making runs. There’s a mismatch.
So here’s a brain teaser. What would either testing or instruction do differently for a match between practice and testing that validly reflects actual learning? What piece is missing from instruction that could make it more like testing, or missing from testing that could make it more like instruction?
The unifying feature is that both depend on students expressing their learning. To the extent that 80% of student responses to teachers are in 30 words or less, this is not expressing their learning. To the extent that teachers do hours of talking every day and students do minutes, this is not students expressing their learning, yet our manner of testing students relies solely on student expression. We remove all help and put them on the spot to produce, while their “practice learning” is heavily comprised of listening to a teacher, responding in piecemeal ways, and moving information around—gather it, put it in a designated place, and turn it in—a process barely recognizable as parallel to testing. US instruction has a smorgasbord of things it wants to do that, individually taken, are not direct learning, but planners remain curious what proportion of learning might lodge in students’ brains anyway.
Our bottom line then is, What universal activity K-12 both deepens learning and validly tests it? We return to our starting premise. The activity is expressing learning. High-value practice presumes the input of approximate, raw knowledge, and then the output of it to integrate and deepen it. These two functions, input and output, occur most notably when students:
- write a summary of what they hear from the teacher.
- write a summary from what they already have in mind.
- write a summary of text material.
- answer questions posed by teacher or fellow students.
- explain to a partner, piece by piece, everything presented that day.
- explain to a partner, piece by piece, everything presented for the week, month, and year.
- explain to a partner the whole subject, beginning to end.
- randomly selected students stand and explain randomly selected pieces to the class.
- in a public performance, each student presents a randomly selected piece of the year’s material.
- students go into the community for public speaking opportunities about what they are learning.
Just monitoring the process of students doing these things cannot fail to provide teachers a refined picture of what students actually know. Students expressing their learning in quantity in various forms provides all the information adults need in order to guide their individual growth.
So if we know what we need to about individual students, what is the role of high stakes testing? It serves adults who are faced with managing education resources. Even given that they know the test does not match what has been taught, even given the fuzziness of its conclusions, and its downsides noted above, they may still want this information (accommodating to the cultural assumption that any idea that can be expressed in numbers is more reliable).
For them to obtain the information they think they need, 1) it should not threaten or undermine any of the ways students benefit from instruction; 2) It should not make instruction harder for teachers; 3) A snapshot of a school or district precise enough for management purposes is to randomly select ten percent of its student body and administer the test to them; 4) Remove pressure from the students by informing them that their personal score is not the point. Rather, they represent their school or district, and you appreciate their help.
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: firstname.lastname@example.org