Session 4: Materials, Assignments & Quizzes

Most teaching materials are never opened, and the reason is that they are written as records of the lesson rather than as tools for the learner. A forty-slide deck sent after class is a transcript nobody reads. A one-page sheet with the steps, the shortcuts and one worked example gets opened the next day when the student is actually trying to do the thing.

So design materials for the moment of use. Ask: when will they open this, and what will they be trying to do? The answer shapes everything. A reference sheet needs the steps in order with the key commands bolded. A worked example needs the reasoning shown, not just the answer. A checklist needs to be short enough to hold in the head while working.

Then the constraint that makes materials good: one page. Not as an arbitrary rule but because a page forces you to decide what actually matters, and because a student will open one page and will not open twelve. If it genuinely needs more, split it into separate one-page tools for separate tasks rather than one document that tries to cover the course.

An assignment is not busywork, and the test is whether completing it makes the student more capable of doing the thing. 'Watch this video and summarise it' produces summarising, not competence. 'Total this column of real sales figures and flag any month that looks wrong' produces the skill the lesson was about, plus the judgement that goes with it.

The structure that works has three parts. The task, stated as something they would do in real life rather than as an exercise about the software. The materials, provided rather than left to be hunted — data files, starter documents, screenshots. The standard, stated in advance: what a correct submission looks like, so the student can check their own work before sending it.

Then make it slightly harder than the lesson, for the same reason practice must be. An assignment that repeats the demonstration exactly can be completed by copying and teaches nothing. Change the data, change the goal, add one complication. And size it honestly: an assignment taking three hours in a class of working adults will not be completed, and an uncompleted assignment teaches less than a shorter completed one.

A quiz has two possible purposes and they produce different questions. Checking recall — did they retain the facts? — is occasionally useful. Diagnosing misunderstanding — where exactly does their thinking go wrong? — is what actually helps you teach, and it is what most quizzes fail at.

The technique that produces diagnostic questions is writing the wrong answers from real errors. For a formula question, the distractors should be the mistakes students actually make: referencing one cell instead of the range, using an absolute reference where a relative one was needed, forgetting to handle blanks. A student who chooses one of those has told you precisely what they misunderstand, and you can fix that specifically. Random plausible-looking wrong answers tell you only that they did not know.

Then the format. Short, frequent and low-stakes beats long and heavily weighted, because a quiz at the end tells you too late to help anyone. Three questions at the start of the next lesson, based on what confused people last time, tells you in the moment whether to move on or go back. And explain every answer when you review it, including the right one — the explanation is where the learning happens, and a score without an explanation is just a number.

Most student confusion is not about the subject; it is about what they are supposed to do. Unclear instructions produce silent non-compliance — the student does not ask, they simply do not do it, and you conclude they are uncommitted. The fix is to write instructions that remove every guess: what to do, using what, producing what, by when, and where to send it.

Then one channel, checked predictably. Scattered instructions across WhatsApp, email and the meeting chat guarantee that someone misses something. Choose one place for announcements and one for questions, state which is which, and check them at stated times. Students do not need instant replies; they need to know when a reply will come.

And answer the question everyone has. When one student asks something in the group, answer it publicly rather than privately, because three others had the same question and only one asked. The same applies to a mistake you made in the materials: correct it to the whole group rather than quietly fixing the file. Students forgive errors readily; they resent discovering them alone.

Handling late work, absence and non-completion consistently matters more than the specific policy. A strict policy applied evenly is respected; a lenient policy applied unevenly is resented, because students can tell who got the exception and why. Decide your rule before the first late submission, state it at the start, and apply it to everyone including the sympathetic case — with a genuine emergency handled as an exception announced as such, not as a quiet favour.

For non-completion, the useful response is diagnostic rather than disciplinary. Ask what got in the way, because the answer is usually one of three things: the task was too long, the instructions were unclear, or the student is lost. All three are yours to fix, and treating them as laziness loses the student and the information.

Finally, build the materials as a reusable system rather than per class. A template for the one-page reference sheet, a bank of quiz questions with real-error distractors, a standard assignment structure with a slot for the task and the data. It costs more the first time and much less every time after, and it is what makes teaching sustainable alongside everything else you do.