Assessment for Learning: Moving Beyond Tests and Grades
The most powerful assessment tool is not a test — it is a well-timed question. Assessment for learning means using information about what students know to improve teaching in real time, rather than to rank students after the fact.
The Difference Between Assessment OF Learning and FOR Learning
Traditional assessment — end-of-unit tests, examinations, graded essays — is assessment of learning. It measures what has already been learned and communicates that measurement to students, parents, and administrators through a score or grade. It is backward-looking, summative, and largely disconnected from the teaching that produced the learning it measures.
Assessment for learning — also called formative assessment — is fundamentally different in purpose, timing, and effect. It asks 'what do students understand right now, and what should I teach differently tomorrow?' It is forward-looking, iterative, and embedded in the daily fabric of teaching. When done well, it is invisible to students as 'assessment' at all — it simply feels like good, responsive teaching.
Dylan Wiliam, who has reviewed more than 4,000 research studies on assessment, concluded that formative assessment — when implemented consistently — is one of the most powerful tools for improving student achievement ever documented. The effect sizes are comparable to reducing class size by a third. It costs nothing beyond a change in practice.
Five Formative Assessment Techniques
Exit tickets are simple written responses to a prompt at the end of a lesson — 'write one thing you understood and one thing you are still confused about.' They take three minutes, give the teacher precise information about which concepts landed and which need revisiting, and communicate to students that their thinking matters. Reviewing exit tickets before the next lesson transforms what might otherwise be a continuation of ineffective instruction into a targeted, evidence-based response.
Think-pair-share — where students think individually, discuss with a partner, and then share with the class — reveals student thinking in a low-stakes, supportive environment. The partner discussion stage is particularly valuable because students often articulate their understanding (and their confusions) more freely to a peer than to a teacher. Circulating during the pair stage gives the teacher rapid insight into the range of understanding in the room.
Traffic light self-assessment asks students to rate their own understanding using a simple system: green (confident), amber (somewhat uncertain), and red (confused). When students hold up coloured cards or write coloured dots on their work, the teacher gets an instant overview of the class's self-reported understanding. Combining this with targeted questioning — asking a 'green' student to explain to an 'amber' student — creates productive peer teaching that benefits both parties.
Feedback That Moves Learning Forward
The most important finding in assessment research is that feedback is only useful when students can act on it. A grade of 65% on a test tells a student how they performed but gives them no information about how to improve. A comment that says 'your analysis is strong — now push further to evaluate which factor was most important and defend your choice with evidence' gives the student a clear next step and the motivation to take it.
Effective feedback is specific, forward-looking, and action-oriented. It identifies one or two areas for improvement rather than exhaustively cataloguing every error. It uses language that maintains the student's sense of competence — 'this is good thinking that will be even stronger when...' rather than 'this is wrong because...' And it comes quickly enough that the student can still remember the thinking that produced the work.
AI tools are beginning to support feedback at scale. A teacher who uses AI to generate differentiated question sets can use the same system to generate targeted feedback prompts for common misconceptions. When 15 students make the same error on a worksheet, an AI-generated explanation of the misconception — personalised to the specific error — can be distributed quickly, allowing the teacher to focus one-on-one attention on the students with more complex or unusual difficulties.
Put this into practice with BrightBoard
Create AI-powered presentations, worksheets, games and more — designed specifically for teachers.
Get Started Free