Formative vs Summative Assessment: Getting the Balance Right

Assessment is often reduced to one image: a room full of students bent over an exam. But that is only half the story. The most powerful assessment happens quietly, every day, in the flow of teaching.

Two questions, two tools

Formative assessment asks: how is learning going right now? It is low-stakes, frequent and designed to inform the next lesson, a quick quiz, a show of hands, an exit ticket.

Summative assessment asks: what has been learned overall? It is higher-stakes and periodic, an end-of-unit test, a term exam, a final project.

Why the balance matters

Lean too far into summative assessment and you measure learning without improving it. Lean into formative assessment and you get a steady stream of signals you can act on, but you still need summative checkpoints to certify progress.

Formative assessment is for learning. Summative assessment is of learning. Great teaching needs both.

Ten formative techniques that take under five minutes

1. Exit tickets. One question on a slip as students leave. Sort the slips into three piles: got it, nearly, needs re-teach.
2. Show-me boards. Every student writes an answer and holds it up. You see the whole class in one glance.
3. The mini-quiz. Five questions on last lesson’s content. If the class average drops below seventy percent, reteach before moving on.
4. Thumbs check. “Thumbs up if you could explain this to a friend, sideways if unsure, down if lost.” Crude but fast.
5. One-sentence summary. Students summarise the lesson in a single sentence, their sentence reveals their understanding.
6. Wrong answers only. Students propose the most likely mistake on a problem. Understanding errors is deeper understanding.
7. Traffic-light self-rating. Students mark their own confidence on the objective; compare with their actual quiz score to build honesty.
8. Two-minute diagram. Sketch the process, the water cycle, the sentence structure, the algorithm, from memory.
9. Peer explain. Student A teaches the step to Student B in ninety seconds. Teaching is the strongest check of learning.
10. Anticipation stakes. Before revealing an answer, every student commits to a prediction. Commitment makes the reveal stick.

Designing a summative assessment that is fair

A fair summative paper is boringly predictable, and that is a virtue. Its blueprint should be fixed before any question is written: which objectives it covers, in what proportion, at what difficulty mix. Every objective taught should appear; no objective untaught should appear. The command words should match what was practised, if lessons never asked students to “evaluate”, the exam cannot suddenly demand it.

Give the paper to a colleague and ask one question: “Which objective does each question test?” If they cannot answer in a sentence per question, the blueprint has drifted, fix the paper, not the students.

Feedback that actually changes outcomes

Marks alone change nothing; feedback changes outcomes only when students do something with it. Keep it proportional: one strength, one fix, one task. “Your method is organised, your weakness is converting units. Redo questions three and five converting first.” That is feedback a student can act on in minutes, and acting on it is what consolidates the learning.

How technology helps

• Speed. Auto-graded quizzes return results instantly, so feedback arrives while it still matters.
• Insight. Analytics reveal not just who struggled, but which concepts and questions caused it.
• Reduced load. Automatic marking of objective questions frees teachers to focus on the human feedback only they can give.

A practical weekly rhythm

Monday: five-question retrieval quiz (formative). Wednesday: exit ticket on the week’s hardest concept (formative). Friday: ten-minute marked mini-test covering the week’s objectives (a bridge between formative and summative). At unit’s end: the full summative test, blueprint-checked. None of this adds workload, it replaces vague impressions with signals, and signals with decisions.

With LexsEdu Assessments you can build both formative quizzes and summative exams, deliver them online, and let the platform grade and analyse the objective sections automatically. If workload is your worry, pair this with our guide to cutting teacher workload without cutting corners.

Recording formative evidence without drowning

The trap with formative assessment is bureaucracy: teachers start logging every exit ticket and the practice collapses under its own paperwork. The fix is to record decisions, not scores. A simple grid per class, objectives down the side, a tick for secure, a circle for partial, a dash for reteach, is enough. The grid exists to answer one question on Sunday night: what does Monday’s starter need to be? Anything that does not change a future lesson does not need recording at all. Summative results, by contrast, deserve full records, because they travel, to parents, to leadership, to the next teacher. Keep formative light and summative thorough, and the balance maintains itself.

Two mistakes schools make when they automate

First, automating the wrong questions. Auto-grading shines on objective items, facts, calculations, structures, but a platform graded essay is not yet a good essay grader; keep extended writing with the human, and give the human time to actually read. Second, letting frequency become surveillance. Ten auto-quizzes a week that students resent produce compliance, not information. The rhythm above, three short check-ins a week, stays under the radar of test anxiety precisely because it is predictable and low-stakes. Automation should buy back teacher time and sharpen signals, never turn a classroom into a checkpoint.

What this means for exam years

Final-year classes live under the shadow of one enormous summative assessment, and the classes that fare best are paradoxically the ones that treat most of the year as formative. Past-question drills marked purely for score teach students what they already know. The same drills run formatively, students marking with the mark scheme, diagnosing why the lost marks were lost, keeping an error log by topic, convert every practice paper into a personalised revision plan. By the time the real examination arrives, the summative event is simply the last of a hundred familiar exercises, not a first encounter with pressure.

One final nuance: the same task can play either role, and the difference is what happens next. A class test followed by a grade and a new topic is summative. The identical test followed by reteaching of the two weakest questions is formative. The paper does not decide, the teacher’s response does.