Better Feels Worse?

Within a week or two of a new student starting, every tutor, every school and every course hears the same sentence.

“I’m not sure this is working for me.”

Suppose your child said it to you at the end of their second week. How would you decide whether they were right? Most of us would start with how the lessons have felt: whether the tutor explains things clearly, whether our child enjoys the lessons, and whether they finish each one feeling that they understand more than before. It is the obvious test. In 2019, a group of physicists at Harvard found a way to check whether it works.

The class that was split in two

The Harvard study was designed to separate two things that are easy to confuse: how much students actually learn and how much they think they have learned.^[1] The physicists took a single fifteen-week introductory physics course and taught every student in it the same way for eleven weeks. Then, in week twelve, they split the class at random.

One half continued with lectures from an experienced and highly rated instructor. The other half was taught actively: they had to attempt the problems themselves, get them wrong, and be corrected as they went. Both halves had the same content and the same handouts, and the instructor made no effort to persuade anybody that either method was better.

Then the researchers measured two things: they tested what the students had actually learned, and they asked what the students believed they had learned. Before you read the result, make a guess. Which half do you think learned more, and which half do you think felt that it had?

The two results pointed in opposite directions. The students in the active classes learned more. The same students felt they had learned less. The researchers’ own summary is blunt: students in the active classroom “learn more, but they feel like they learn less”.^[1] So if those students had judged their lessons by how they felt, the lectures would have won.

That study was run on university physics students, not on fourteen-year-olds writing English essays, so we will not claim that it transfers exactly. We think the reason behind the result does transfer, because it has nothing to do with physics: effort feels like failure, and fluency feels like understanding. Watching somebody clever take a hard idea apart feels excellent. Being handed the same idea and made to attempt it yourself, badly, feels like proof that you cannot do it.

Why the better lesson can feel worse

Now put that finding next to the ordinary way families judge a new course. The researchers themselves warn that trying to evaluate teaching on the basis of how much students think they are learning “could inadvertently promote inferior (passive) pedagogical methods”.^[1] We think a family that judges teaching by how good it feels will keep choosing the teaching that works less well. The comfortable lesson wins the survey and loses the exam.

Think of a student who has spent two weeks being made to plan before writing a word, being marked strictly, and being told to write the same paragraph again. When that student says “I don’t know if this is working”, the Harvard result would lead us to expect exactly that feeling from effective teaching. The feeling cannot tell you whether the teaching is failing or whether it is succeeding.

One number points nowhere

Even if the feeling pointed the right way, an early judgement would have a second problem. Learning is a lasting change, while what anybody can watch in a lesson is performance: how well the student did that morning. A major review of the evidence makes the same point: what we can observe during teaching is performance, “which is often an unreliable index” of whether learning has taken place.^[2] The same review notes that improvements in performance can fail to produce significant learning, and that some teaching choices push the two in opposite directions.^[2]

A plainer point needs no research at all. A student writes one essay, it is marked, and you are holding a single number. Can that number tell you whether your child is getting better or standing still? It cannot, because a single number points nowhere.

The smallest thing that has a direction is a completed cycle: the student writes something, it is marked against the examiner’s criteria, the feedback names what to change, the student redrafts it, and the redraft is marked again. A second number exists only at the end of that cycle. Two numbers can make a line, while one number on its own can only make a feeling. So a week cannot answer the question, because in a week the second number does not exist yet.

Patience from you, evidence from us

“You cannot judge it yet” is a reasonable thing for a teacher to say in week two. It becomes an excuse the moment it is used to avoid ever being judged.

If we ask a family to wait for the first full cycle before drawing conclusions, then we owe them the evidence the moment that cycle closes. That evidence is the marked piece, the feedback, the redraft, the new mark, and an honest reading of the difference between the two marks.

Patience now in return for evidence later is a fair exchange, and you should expect it from any teaching, including ours. If nothing changed, that is a real answer and you are entitled to it. A grade with no explanation is not an answer, from us or from anyone.

The student is not a patient

Now go back to the sentence from the start, and this time look at the words instead of the worry. Underneath it is a question: “Does it work for me?” That question describes a student as somebody things are done to. A course arrives and either works on them or does not, the way a medicine either helps a patient or does not.

Tuition does not behave like a medicine, because nothing is being done to the student that could take effect or fail to. What a tutor hands over is a set of tools: a way to open a paragraph, a mark scheme put into plain words, a sentence crossed out with the reason written beside it. Those tools only turn into a grade when the student picks them up. That means attempting the work, using the method on their own writing, carrying the confusion back to the tutor, and rewriting the paragraph when the feedback names what was wrong. So instead of “is this working for me”, the honest question is: am I using this?

You might expect a tutor’s explanations to be what makes tutoring work. In one of the best-known studies of tutoring, researchers stopped the tutors from explaining and giving feedback, and asked the tutors to prompt the students instead. In the authors’ own word, the result was “surprising”: the students learned just as effectively. The researchers credited what the students then did, rather than the teaching that had been removed: the students built more of the understanding themselves and took more control of their own learning.^[3]

So a family can be investing in excellent teaching and getting very little back, with the marked essays piling up in a folder while the writing stays the same.

We think nobody has ever learned to write well just by being present while writing was explained. The student who improves is the one who attempts the paragraph, gets it wrong, listens to why, and writes it again — and then does that fifty more times. The work is theirs: the teaching decides what comes next, holds the standard and aims the next attempt at whatever went wrong in the last one, but it cannot do the writing.

That gives a student a more useful question than whether the course is working: am I working, and is my work getting better? In week two, nobody can say whether the course is working. In week two, your child can already say whether they are working. They can also say whether the work is getting better.

Four things to do in the first month

While you wait for the first full cycle of marking and redrafting to close, there are still four things you can do.

1. Ask to see the criteria the work is marked against. Ask for the examiner’s own wording rather than a description of it, and for the line your child’s paragraph is being compared with. A tutor who cannot show you that wording is marking by instinct.

2. Ask for one named weakness instead of a grade. “Her topic sentences state the topic instead of making a point” is something a student can work on tomorrow. “She needs to go deeper” is not.

3. Wait for the second mark on the same piece of writing. One essay gives you a number, but the same essay redrafted and marked again gives you a direction. Judge the direction.

4. Ask your child what changed, rather than how the lesson felt. “What did you get wrong last week that you can do now?” is a question a fourteen-year-old can answer honestly. “Is it working?” is not.

If you would like an honest starting measurement — which specific parts of your child’s writing are costing the marks, judged against the examiner’s criteria — our free assessment takes five minutes.

—

Sources