Choose your language
Creating assessments and exam success — Part 1: Creating fair assessments
Most teachers create some kind of test. A short quiz on Friday. A unit test at the end of a chapter. A monthly progress check. Even when big exams are decided by the education authority, teachers still build their own assessments most weeks.
But here is the question many of us never get asked: was this a good test?
A test can look fine on paper and still measure the wrong thing. A test can give clear marks but be unfair to half the class. A test can have nice questions but tell us nothing about what students actually need to learn next.
In Part 1 of this series, we will look at four practical principles for creating assessments that are fair, that measure what they say they measure, and that actually help students learn. None of this needs new equipment or training. It just needs you to ask better questions before you write your next test.
Q2: Which of these are real challenges in your assessment work? (Tick all that apply)
- Most teachers were never formally trained in assessment design. The good news is that the principles are simple and learnable — you do not need a degree in measurement
- If the same students always fail, the test may be unfair, not the students unable. A poorly designed test punishes students who missed classes, who are weaker readers, or who do not know the question type
- Instructions that are harder than the language tested are a hidden trap. If a student fails because they did not understand the instructions, you have measured their reading of instructions, not the skill the test was supposed to assess
- If you have to mark on a curve where some students must fail, that is norm-referenced assessment — and there are fairer alternatives. We will look at three options below
- Marks without feedback are the biggest waste in education. A student who fails and does not know why will fail again. We will return to this in Section 3
An assessment is not just a tool for giving marks. It is a tool for learning.
A good assessment tells you what your students can do, where they need help, and what to teach next. It tells the student where they stand and how to improve. A bad assessment just produces a number that ranks people — without helping anyone learn.
Below are four principles for designing assessments that are genuinely fair and useful. They come in the order you should think about them: before you write the test (purpose, type), while you write it (validity, good tasks), and after the test happens (marking and feedback).
First — what kind of comparison?
There are three ways to make sense of a student’s test mark. Each has a very different impact on motivation.
In challenging settings, a mix of criteria-referenced (for big tests) and ipsative (for tracking individual growth) is far more useful than norm-referenced grading. Even when the official exam is norm-referenced, your classroom assessments do not have to be.
Four principles for designing fair assessments
Most teachers find one or two principles they did well and one or two they did not even consider. That is normal. The first step is noticing.
- The most commonly missed principle is Principle 1 — clear purpose. Many tests are written because “it is the end of the chapter”, not because the teacher has decided what they want to find out
- The second most missed is Principle 4 — clear marking scheme for subjective tasks. Without one, marking varies by mood, time of day, or which student went before
- Validity (Principle 2) is often broken by tests that accidentally measure something else — reading speed instead of comprehension, or memory of the textbook story rather than language ability
- If your test instructions use harder language than the test itself, weaker students fail before they even reach the questions. This is one of the most common — and most fixable — problems
- You do not need to redesign every test. Pick one principle to apply to your next assessment. Add another to the next one. Over a term, your tests get steadily fairer
These are real moments from teachers’ experience. The right answer is the one most likely to make assessments fair and useful.
This is a planning template. Fill in what you can. Where you cannot, that is the part to work on.
| Design step | My plan |
|---|---|
| The purpose: what am I testing and why? | |
| The type: norm / criteria / ipsative? | |
| Validity: does it really test that? | |
| Instructions: are they simpler than the language tested? | |
| Marking scheme: what counts for each mark? |
Example: a unit test on the past simple tense.
| Design step | What good looks like |
|---|---|
| Purpose | To check if students can use the past simple tense in a short personal account about last weekend — and to identify which students still need help with regular vs irregular verbs. |
| Type | Criteria-referenced. The criterion is “can use past simple in a short personal account.” Three categories: Competent / Nearly competent / Not yet competent. |
| Validity | The test asks students to write about last weekend (real production), not just fill in gaps. This actually tests what we want — using the tense, not recognising the form. |
| Instructions | “Write about your last weekend. Use the past simple tense. Write 5 sentences. You have 15 minutes.” Simple words, all the information needed, no surprises. |
| Marking scheme | Competent: 5 sentences, mostly correct past simple, clear meaning. Nearly competent: 4–5 sentences, some past simple errors but clear meaning. Not yet competent: fewer than 4 sentences, or major past simple problems. |
Notice how every step connects: the purpose drives the type, which drives the marking scheme. A test designed this way is fair, useful, and tells you exactly what to teach next.
Q6. Watch the video below. Think about which change is easiest for you to try first.
Host: We have just looked at four principles for designing fair, useful assessments. Now listen to three teachers. They share their problems first, then the changes they made.
Teacher 1: I used to write tests because the textbook ended a chapter, not because I had a clear purpose. The marks went into the gradebook. The students got their numbers. But I never asked myself: what am I trying to find out? My tests were a routine, not a tool for learning.
Teacher 2: My tests were always norm-referenced. The top of the class passed, the bottom failed. Every term, the same students were at the bottom. They expected to fail. They had stopped trying. I did not see how my tests were part of the problem — they were keeping those students stuck.
Teacher 3: I had no marking scheme for writing tasks. I marked by overall feeling. Sometimes I gave 7 out of 10. Sometimes 8. The same student would get a different mark depending on my mood, or which paper I had marked just before. It was not fair. But I did not know how to fix it.
Teacher 1: I started writing the purpose at the top of every test I drafted. “To check if students can…” One sentence. Suddenly questions that did not serve the purpose got cut. Tests got shorter and more focused. I learned much more from the results because I knew what I was looking for.
Teacher 2: I switched to criteria-referenced grading for my own classroom tests. Three categories: Competent, Nearly competent, Not yet competent. I also started recording each student’s growth from term to term — ipsative tracking. The students who used to fail every test now had something to be proud of. Their attendance went up. Their effort went up.
Teacher 3: I started writing simple marking schemes for every subjective task. 3 for clear and accurate. 2 for clear with some errors. 1 for limited. Just three lines. My marking became consistent. Students could see why they got their mark. They could see what to improve. The marking actually helped them learn.
Host: None of these teachers had different students or new equipment. They thought differently before they wrote each test. The result: tests that were fair, useful, and that genuinely helped students learn — rather than just sorting them into pass and fail.
Q7. For each principle, choose the option that best describes where you are now.
One test, one change. Pick the principle that needs the most attention. Real change in assessment habits comes one test at a time.
- An assessment is a tool for learning, not just for ranking. The best test tells you what to teach next and tells the student where they stand and how to improve
- Three types of comparison: norm-referenced (against classmates — demotivating), criteria-referenced (against a clear standard — fairer), and ipsative (against past self — tracks growth). Use a mix of the last two for your classroom tests
- Four design principles: clear purpose, valid & reliable, good tasks (with simple instructions), clear marking scheme. Apply them in order, before you write the test
- The 5-question check: focus on what was studied? Will students think it is fair? Does it test what it says? Can absent students still show what they know? Would different markers give the same mark?
- Marking schemes are not bureaucracy — they are how you stay fair and consistent. Even a simple 3-2-1-0 scheme is better than no scheme at all
How useful did you find this lesson? Leave a rating and a comment to help other teachers.