Assessment with Purpose: Moving Beyond Memorization at AELO Swiss Academy

Every teacher knows the ritual. The night before the exam, a student hunches over a stack of flashcards like a raccoon guarding something shiny. Dates, definitions, the stages of mitosis, the exports of Bolivia. By morning the student can recite all of it with the fluency of a late-night radio host. Three weeks later, ask what mitosis is, and you will watch the knowledge leave the premises in real time, like a tenant who stopped paying rent.

I have taught long enough to stop finding this funny and start finding it instructive. Cramming works, briefly and narrowly, which is precisely the problem. It works well enough that students keep doing it, and badly enough that we should stop letting exams reward it. At AELO Swiss Academy, we have spent a great many uncomfortable meetings asking one question: if an assessment can be defeated by forty-eight hours of panic and caffeine, what exactly is it assessing?

The honest answer is usually: short-term memory and a tolerance for misery. Neither of those is what anyone means by an education.

What memorization is actually good for

Let me defend memorization for a moment, because it deserves a fair trial before we hang it. You cannot think critically about nothing. A chess player who has to re-derive the rules every game is not creative, they are exhausted. Facts are the raw material of thought, and a student who has to look up every term mid-argument is a cook running to the shop between each chop of the onion.

Research on retrieval practice keeps confirming something old-school teachers always suspected: the act of pulling a fact out of your head strengthens the memory more than putting it in does. So yes, we still quiz. We still ask students to know things cold, on demand, without warning. A Swiss school that gave up on precision would be a strange kind of Swiss school, roughly as plausible as a fondue made of water.

The trouble starts when recall is the ceiling instead of the floor. Memorization gets you the ingredients. Assessment that stops there is a cooking exam where nobody is allowed to turn on the stove.

The multiple-choice problem

Recognition is not the same as production, and life is stingy with options. Life hands you zero choices and a deadline. Yet a remarkable share of traditional testing measures the ability to pick the right answer out of four, which is a skill with a shockingly narrow professional application. Outside of game shows, nobody has ever handed me a problem pre-packaged with one correct option and three decoys, one of which is obviously too long to be wrong.

There is a student I think about often, the one who can recognize a correct definition of osmosis from across the room but cannot tell you what will happen to a raisin left in a glass of water. The definition is in there somewhere, filed and laminated. The understanding never checked in. A well-built multiple-choice question can catch some of this, but the format itself whispers to students that knowing the shape of the right answer is enough. It is not enough. It is the shape of enough.

Opening the caseback

Swiss schooling has long kept the oral examination in the family, and I suspect that tradition survives because it does something no written format can. Think of a watch: from the outside you see the face, and the face can be lovely. But you only learn what the thing is worth when someone opens the caseback and the movement is exposed, gears turning, tension visible, the whole mechanism either coherent or pretending.

An oral assessment opens the caseback. You ask a student to explain their reasoning, then you do the thing flashcards cannot do: you ask a follow-up. "Why did you assume that?" "What would change your mind?" "Where does this argument break?" The FAA pilot license types follow-up question is where the cramming stops and the thinking starts, and it is the single most efficient diagnostic tool I have ever used. It takes forty seconds and it separates understanding from rehearsal with the ruthlessness of a border guard.

At AELO Swiss Academy we kept orals and defenses not because they are efficient. They are not. They eat calendar time like a goat eats a garden. We kept them because a student who can handle "I don't know, but here's how I'd find out" has passed an exam that no answer sheet can administer.

Designing questions that refuse to be crammed

The craft of purposeful assessment lives in the prompt. You can usually take a recall question and rebuild it into something that demands transfer, and the difference is one sentence of intent.

| The old prompt | The purposeful rewrite |

| --- | --- |

| List three causes of the First World War. | A historian claims the alliance system alone made the war inevitable. Use two other factors to argue them out of it. |

| Define osmosis. | A gardener salts the path to kill weeds. Days later, the hedge beside the path wilts too. Explain what happened at the level of the plant cell. |

| Name the four Ps of marketing. | Your café in Lausanne has a great location and is still failing. Diagnose the problem using two of the four Ps and tell me which one you would fix first, and why. |

Notice what the rewrites have in common. Each one hands the student a situation rather than a slot to fill. Each one can still be crammed for, technically, but the cramming has to include actually understanding the material, at which point we have quietly won.

I will be honest about the cost, because pretending there is none is how bad ideas spread. These questions take longer to grade. A bubble sheet grades itself in seconds; an argument takes minutes and a cup of coffee. Multiply by a class set and a semester and you understand why the recall exam has survived so long. It is cheap. But cheap and worthless is a bad bargain at any price, so we pay in teacher hours instead.

Five questions we ask before any assessment

Before anything with points attached goes in front of students, it has to survive a small inquisition. If it cannot answer these, it goes back to the workshop:

• What decision will this result actually inform? If nobody changes anything based on the outcome, it is a ritual, not an assessment.
• What does mastery look like, concretely, on a good day, in a sentence a student could repeat back?
• Can this be defeated by cramming? If yes, redesign it or admit we are measuring caffeine tolerance.
• What will students do with the feedback, and when, and will there still be time to act on it?
• Does the format match the claim? If we claim graduates can construct arguments, an exam with no room to argue is false advertising.

Rubrics, the least glamorous document in the building

Somewhere between the fire evacuation plan and the staff-room rules about the dishwasher sits the rubric, and I would like to rehabilitate its reputation. A rubric shared before the assessment is a promise. A rubric written after, or never shown, is a guessing game, and students who spend the exam trying to read the teacher's mind are not demonstrating mastery of chemistry. They are demonstrating mastery of us.

A good rubric describes observable performance rather than vibes. Compare "excellent analysis" with "identifies the strongest objection to their own argument and responds to it." The first could mean anything, which in practice means it means whatever the grader had for breakfast. The second tells a student exactly what thinking looks like when it is done well, which means it also tells them what to practice.

The edge case deserves a mention, because it shows up every year: the brilliant essay that ignores the rubric entirely and is somehow magnificent. When that happens, grade the performance against the purpose, then go home and fix the rubric. The rubric serves the learning, never the other way around, and a checklist that produces paint-by-numbers essays has betrayed its own cause.

Feedback as coaching, grades as autopsy

A grade arrives after the learning is over. That makes it, structurally, an autopsy: cause of death, time of death, nothing to be done. Feedback during the process is coaching. Same information, radically different consequences.

Decades of research on feedback keep landing on the same awkward result: when a comment and a score travel together, students read the score and bin the comment. So at AELO Swiss Academy we sometimes hand back work with comments only and the grade banked until after revision. The first time you do this, students look at you like you have proposed paying rent in poetry. The second time, they revise. Turns out the grade was the thing standing in the doorway.

Timing matters as much as content. Feedback three weeks after the essay is a postcard from a country the student has already left. Exit tickets, one-minute papers, a teacher circulating during work time rather than appearing afterward like a verdict: these are the small, unglamorous moves that move feedback from autopsy to midwife.

Failure, retries, and the dignity of a second attempt

Switzerland is a country of via ferrata routes, those iron-cable climbs bolted into the Alps, and the mountain does not offer partial credit. But the whole point of the cable is that you can fall, hang there, breathe, and climb again. A course should work the same way.

Our policy on major assessments is that a retake is available, but it is not free and it is not identical. The student submits a short plan for what they will do differently, and the retake tests the same skills from a different angle. The reflection requirement is not bureaucracy for its own sake. It is the difference between a second attempt and a second spin of the same slot machine, and it filters out the strategic re-takers who were hoping to game the calendar. The trade-off is real: retakes cost time, and somewhere a teacher is sighing. We budget for the sigh.

What about the exams that exist outside our walls?

Students still face external examinations, standardized admissions tests, and university requirements, and pretending otherwise would be a disservice dressed as idealism. We have not declared war on exams. We have declared war on bad ones.

Preparing for a recall-heavy external exam is a skill, and we teach it openly: spaced retrieval, interleaved practice, timed drills, all of it. Cramming has an honest cousin called distributed practice, and the difference is mostly calendar placement and self-respect. Ten minutes today and ten tomorrow beat two hours of Sunday despair, and the research on spacing has been saying so since before any of us were born.

The introvert deserves a word here too, because orals can be cruel to quiet students if the school confuses volume with understanding. Preparation time, questions shared in advance, the option to submit a recorded response first: these accommodations do not weaken the assessment. They remove the measurement error. Fairness is not sameness, it is accuracy.

How to study when the exam cannot be crammed

Students ask the obvious follow-up question: if flashcards will not save me, what will? The honest study plan is short and slightly annoying in its simplicity:

• Explain the topic to someone who does not know it. The gaps announce themselves mid-sentence.
• Redo problems you got wrong, from a blank page, three days later.
• Change the context. If the practice problem is about trains, rewrite it about buses. Transfer is the actual exam.
• Write the question you would fear most, then answer it before a classmate writes it for you.
• Space it. Small and repeated beats large and desperate, every single time.

A confession, and a wager

I will admit something. I still write the occasional lazy recall quiz, usually late at night, usually when the good version of me has gone home. Purpose is not a plaque on the wall. It is a practice, and practices wobble. The difference at AELO Swiss Academy is that when a bad assessment slips through, someone notices and says so, which is more uncomfortable and more valuable than any policy document.

Here is the wager, and I am willing to make it publicly. Ask our graduates in ten years about the Congress of Vienna and they will not remember the date. Ask them to take a position, defend it against a follow-up question, and change their mind when the evidence deserves it, and they will manage all three before their coffee cools.

That is the assessment we are actually trying to pass. The students are just the ones taking it for us.