A school leader sits at the back of a classroom with a laptop.
The teacher knows she is being observed.
The lesson is unusually polished. Resources are printed. Questions have been rehearsed. The difficult class has been told that a visitor is coming. The observer writes notes. Forty minutes later, everyone leaves with evidence that a lesson occurred.
What, exactly, has been appraised?
Perhaps the observer saw important teaching.
Perhaps the observer saw a performance created for the observation.
Perhaps one lesson revealed a real pattern.
Perhaps it revealed an exception.
Perhaps the teacher receives useful feedback tomorrow.
Perhaps the notes disappear into a form that is opened again only at the next review.
Teacher appraisal sounds simple because the visible event is familiar: someone watches a teacher and judges performance. The real mechanism is more demanding.
A useful appraisal system has to connect at least six things:
a trustworthy standard, relevant evidence, defensible interpretation, professional dialogue, a consequential next step, and later evidence that the next step mattered.
Break one connection and appraisal can become theatre.
If standards are vague, the judgement becomes personal taste.
If evidence is thin, one lesson becomes a verdict on a year.
If feedback is generic, the teacher does not know what to change.
If appraisal has no developmental follow-through, the process documents practice without improving it.
If it is only developmental and never protects minimum standards, serious weakness can remain unaddressed.
If it is only punitive, teachers may learn to hide problems rather than examine them.
OECD’s current TALIS 2024 reporting makes this tension unusually concrete. Teacher appraisal is widespread across participating education systems. The report describes two central functions: formative, in which appraisal helps teachers improve, and summative, in which it supports judgements about effectiveness and accountability. Classroom observation is the most common formal appraisal method reported by principals. Yet what follows appraisal varies substantially. Across OECD education systems, about 65% of teachers are involved in post-appraisal discussions aimed at improving teaching, while about 46% are offered development or training plans.
Singapore provides a useful local example. In the TALIS 2024 country note, principals report classroom observation as an appraisal method for 98% of teachers, assessment of content knowledge for 95%, and school- or classroom-based results for 82%. Common reported consequences include discussing measures to address weaknesses in teaching, developing a training or development plan, and appointing a mentor.
Those numbers describe systems, not causal effects. They do not tell us that frequent appraisal automatically improves students’ learning. They do, however, make the mechanism worth understanding.
The question is not whether teachers should ever be observed.
It is how an appraisal process turns observation into better professional action without mistaking measurement for improvement.
The 50-second answer
Teacher appraisal works when it closes a professional loop:
define good teaching → gather enough relevant evidence → interpret the evidence carefully → make a proportionate judgement → identify a small number of important next actions → provide support → inspect later evidence.
The appraisal event is only the middle.
A useful system asks:
What standard matters here?
What evidence is strong enough to support the judgement?
What alternative explanation could fit the evidence?
What should the teacher keep doing?
What should change?
What support makes that change plausible?
When will we know whether it improved?
A weak system ends with a rating.
A stronger system ends with a better next lesson—and later checks whether the improvement survived beyond the observed day.
Start on the classroom floor: one observed lesson is not one whole teacher
Imagine two observations.
Teacher A runs a calm lesson. Students answer questions. The board is neat. The class completes the task.
Teacher B’s lesson looks less smooth. Students struggle with a difficult problem. Several answers are wrong. The teacher pauses, changes representation, asks pupils to compare two errors and eventually gets a stronger explanation from the class.
Which teacher should be rated higher?
The question cannot be answered from surface smoothness.
A lesson may look orderly because the task is too easy.
A lesson may contain visible struggle because the teacher has set an ambitious intellectual demand.
A teacher may talk less because students can already work independently.
Another may model more because the content is new.
Appraisal therefore needs a model of teaching that distinguishes appearance from instructional purpose.
This is why observation rubrics and teaching standards exist. They try to establish common reference points: clarity, subject knowledge, classroom environment, responsiveness, assessment, cognitive demand, support, professional responsibility and other dimensions.
But a rubric is not a substitute for judgement.
“Students were engaged” is not enough.
Engaged in what?
“Teacher questioned effectively” is not enough.
What did the questions reveal?
“Pace was appropriate” is not enough.
Appropriate to which learning demand and which students?
Good appraisal needs evidence connected to mechanism.
Stage 1: establish the standard before observing the performance
A teacher cannot respond well to a judgement if the standard is invisible.
If appraisal depends on one leader’s personal image of a “good lesson,” teachers learn to perform for preferences.
One observer likes group work.
Another dislikes teacher talk.
One rewards colourful resources.
Another rewards silence.
The same teaching receives different judgement because the criterion lives in the observer’s head.
A professional system needs standards that are explicit enough to guide preparation and interpretation while broad enough to respect legitimate teaching variation.
OECD’s teacher-appraisal guidance emphasises the role of teaching standards as common reference points for appraisal and professional development. Standards can define expectations, support consistency and connect appraisal with wider learning goals.
The key is not to create a hundred-item compliance checklist.
A standard should identify what matters, not prescribe every visible behaviour.
For example:
“Teachers use evidence of student learning to adapt instruction” is a professional standard.
“Teachers must use mini-whiteboards exactly twice in every lesson” is a method prescription.
The first allows professional judgement while demanding evidence.
The second confuses one possible technique with the educational job.
This boundary matters because appraisal systems can accidentally narrow teaching. Once teachers know a visible behaviour earns points, the behaviour appears whether or not it fits the lesson.
The measure starts shaping the performance.
Stage 2: gather more than one kind of evidence
Teaching is too complex for a single snapshot to carry every judgement.
Observation is valuable because it reveals interaction: explanations, responses, transitions, checks for understanding, classroom climate, questioning, task design and adaptation.
It is also limited.
The observer may not know what happened yesterday.
A lesson may be atypical.
Student behaviour can vary by day.
The strongest learning may appear in work produced later.
A teacher’s curriculum planning may be invisible during the observed period.
Student outcomes can add information, but they also require care. Scores are influenced by prior attainment, attendance, curriculum alignment, tutoring, cohort composition, assessment design and many other factors. A number does not isolate the teacher’s causal contribution.
Student surveys can reveal experience but are not direct measures of instructional quality.
Planning documents can show intent but not enactment.
Professional discussion can reveal reasoning but can also reward verbal fluency.
A robust appraisal therefore uses triangulation: different evidence sources that answer different questions.
Observation: What happened in instruction?
Student work: What did learners produce and misunderstand?
Assessment: What patterns persisted or changed?
Planning: What was the intended sequence?
Teacher explanation: Why was a choice made?
Follow-up evidence: Did the change improve the problem?
The aim is not to collect everything.
It is to avoid making a large claim from a narrow source.
Stage 3: separate description from inference
Observers often jump too quickly from what they saw to why it happened.
Description:
“After the teacher asked a question, three students raised their hands. The same two students answered four of the next five questions.”
Inference:
“Most students were disengaged.”
The inference may be correct.
It may also be wrong. Students may have been thinking silently. The questions may have been rhetorical. The teacher may have checked understanding through written work later.
A better appraisal conversation begins with the evidence.
“I noticed most public answers came from two students. What evidence did you have about the rest of the class?”
Now the teacher may say:
“I did not have enough. I assumed silence meant they were following.”
Or:
“They had all written an answer first; I sampled those before taking volunteers.”
The observation has become richer.
This discipline matters because feedback becomes less defensive when it can be traced to shared evidence.
It also protects teachers from confident but weak interpretation.
An appraiser should be able to say not only what judgement they reached, but which evidence justifies it and what uncertainty remains.
Stage 4: distinguish formative and summative functions without pretending they never collide
Formative appraisal asks:
How can this teacher improve?
Summative appraisal asks:
Does this teacher meet required professional standards, and what formal consequence follows?
Both are legitimate.
The tension arises because the same conversation can feel very different depending on stakes.
If a teacher believes that admitting a problem may lower a formal rating, the teacher has an incentive to present competence rather than explore uncertainty.
If every appraisal is purely developmental, leaders may avoid necessary decisions about persistent underperformance.
Good systems make the function clear.
A routine developmental observation can create a safer space for experimenting, diagnosing and learning.
A formal performance review can use broader evidence and explicit standards.
Where the processes are combined, teachers need to know what evidence is used, how judgements are made, what rights they have to respond and what support follows.
Ambiguity is corrosive.
It produces the worst mixture: high anxiety with low clarity.
Stage 5: turn judgement into one or two actionable priorities
Suppose an appraisal identifies seven areas for development.
The teacher leaves with:
improve questioning, increase pace, differentiate more, strengthen behaviour, improve assessment, develop student independence and reduce teacher talk.
This is technically rich and practically useless.
Improvement requires prioritisation.
Which problem most affects learning?
Which is foundational for the others?
Which is small enough to practise?
Which evidence would show progress?
A good post-appraisal action might be:
“Over the next four weeks, increase the proportion of pupils whose understanding becomes visible before independent practice. Use an all-pupil response method in the first two worked examples, sample common errors, and adapt the next example. A mentor will observe the first fifteen minutes of one lesson in weeks two and four.”
That plan has a mechanism.
It links problem, action, support and follow-up.
OECD’s TALIS 2024 figures on post-appraisal development plans matter because they highlight the gap between being judged and being developed. If appraisal identifies a weakness but the system provides no time, expertise or learning route to address it, the teacher has received diagnosis without treatment.
Stage 6: provide the support implied by the judgement
An organisation should be cautious about saying “improve” when it has not considered whether improvement is practically supported.
If a teacher needs stronger subject knowledge, provide subject-specific learning.
If lesson transitions are consuming time, observe a colleague with strong routines.
If curriculum planning is weak, give access to coherent materials and planning collaboration.
If the problem is a persistently disruptive class, do not pretend that one teacher’s technique is the only variable. Leadership, behaviour systems and student support may be required.
If workload is preventing careful feedback, examine workload.
Appraisal becomes unfair when every problem is individualised.
A teacher is accountable for professional practice.
A school is accountable for creating reasonable conditions for professional improvement.
This is why teacher appraisal should connect to professional learning rather than operate as a separate compliance machine.
How Professional Learning Cycles Work | Why One-Off Training Rarely Changes Classroom Practice is a neighbouring owner. Appraisal can identify a learning priority; a professional learning cycle helps that priority become changed practice.
Why calibration matters when more than one person appraises
A school can have a clear rubric and still produce inconsistent judgements.
Two appraisers may use different thresholds. One may treat a quiet class as evidence of strong management; another may look for whether pupils are intellectually active. One may heavily weight lesson structure; another may focus on student work. If these differences are never surfaced, teachers experience appraisal as personality-dependent.
Calibration means appraisers examine common evidence, discuss what the standard looks like in different subjects and age groups, and compare the reasons behind their judgements. It does not require pretending every lesson should look the same. In fact, good calibration should make legitimate variation easier to defend because observers become clearer about the underlying educational function.
A mathematics lesson, a drama rehearsal and an early-years phonics session will not have identical visible forms. The appraiser’s job is to ask whether the relevant professional standard is being met in a way that makes sense for the content and learners.
Calibration also protects equity. If some teachers are routinely judged more harshly because of class composition, subject stereotypes, communication style or an observer’s expectations, the appraisal process can reproduce bias while claiming objectivity. Schools should therefore review patterns in ratings and consequences, not merely assume that a rubric removes human judgement.
Stage 7: return to the evidence later
Without follow-up, appraisal rewards intention.
The teacher agrees to improve questioning.
A development plan is filed.
Six months later nobody knows whether questioning changed.
The loop stayed open.
Follow-up should be proportionate. Not every target needs another full formal observation. A short visit, student-work review, video segment, planning conversation or assessment sample may be enough.
The key is to return to the original problem.
If the issue was that only volunteers revealed their thinking, has participation become more representative?
If the issue was weak explanation of causation, do student answers now contain causal reasoning?
If the issue was slow lesson starts, has time-to-task improved?
If the issue was excessive scaffolding, can pupils now perform independently?
A follow-up check protects appraisal from becoming annual theatre.
It also gives the teacher evidence of progress.
That matters because professional confidence should come partly from seeing that a changed practice produced a better result.
A worked example: the observation that almost produced the wrong conclusion
A school leader observes a Secondary mathematics lesson.
Students spend twelve minutes working silently.
The observer initially writes: “Limited student discussion; increase collaborative learning.”
During the post-observation conversation, the teacher explains that pupils had spent the previous lesson comparing solution strategies in pairs. Today’s goal was to test whether each learner could independently select a method without peer support before the assessment.
The observer reviews the task and student work.
The silence now has a different meaning.
It is not evidence that collaboration is missing from the curriculum. It is part of a sequence moving from supported comparison to independent selection.
The appraisal judgement changes.
This is not a story about teachers always being right.
It is a story about why instructional context matters when interpreting visible behaviour.
The observer can still ask whether the independent task provided enough evidence, whether the difficulty was appropriate and whether some students needed additional support.
But the conversation now evaluates the learning design rather than rewarding a preferred activity structure.
Another worked example: when a rating should lead to support
A Primary teacher is consistently struggling with classroom transitions.
Three observations across several weeks show the same pattern: instructions are given while materials are still moving, pupils ask repeated procedural questions, and five to seven minutes are lost between activities.
The pattern is strong enough to support a judgement.
The appraiser does not simply rate “classroom management: needs improvement.”
The school identifies a development target.
The teacher observes two colleagues.
A mentor rehearses concise directions.
Materials are reorganised.
The teacher uses a visible routine.
A short follow-up observation measures transition time.
It falls to two minutes.
The appraisal has done real work because it connected evidence to practice.
If the school had only recorded the weakness, nothing important would have changed.
Appraisal is not the same as instructional rounds
Instructional rounds look across classrooms to identify patterns in an instructional system. They are not primarily designed to score individual teachers.
Teacher appraisal evaluates an individual’s professional practice against standards and can carry formal consequences.
Confusing the two changes behaviour.
If a system-level learning walk is secretly treated as individual evaluation, teachers will protect themselves rather than expose normal variation.
For the neighbouring mechanism, see How Instructional Rounds Work | See the Pattern Across Classrooms Before You Try to Fix the System.
Appraisal is not the same as performance pay
Performance pay links incentives or compensation to specified performance measures or judgements.
Appraisal can exist without financial consequences. Its primary functions may be development, standards, career progression or accountability.
The distinction is important because incentives can change behaviour around the measure itself.
How Teacher Performance Pay Works | Why Incentives Can Change What Gets Measured Before They Change Teaching owns that separate mechanism.
An appraisal system should therefore be explicit about what consequences attach to ratings. Hidden stakes distort developmental dialogue.
Appraisal and teacher autonomy
At first glance, appraisal and autonomy can look opposed.
Autonomy says: trust professional judgement.
Appraisal says: inspect professional practice.
In a mature profession, these can reinforce one another.
Autonomy without feedback can become isolation.
Appraisal without autonomy can become compliance.
OECD’s TALIS 2024 reporting explicitly examines the interaction. Across systems, teachers with substantial instructional autonomy may work under very different appraisal regimes. In Singapore, almost all teachers reporting high curriculum-implementation autonomy work in schools where formal appraisal occurs at least annually.
The useful idea is accountable professional discretion.
Teachers need room to choose methods and adapt instruction.
They also need standards, feedback and evidence that help calibrate those choices.
A strong appraisal system should not reduce expert teaching to “Did you follow the preferred script?” It should ask whether professional choices served learning and met agreed standards.
What TALIS 2024 adds
OECD’s Results from TALIS 2024 reports that formal appraisal is widespread internationally. According to principals, classroom observation is used for appraisal for 96% of teachers on average across participating OECD education systems. The report also shows substantial variation in what happens after appraisal. About 65% of teachers are involved in discussions about addressing weaknesses, and about 46% are offered development or training plans.
Singapore’s country note reports especially high use of observation and developmental consequences: classroom teaching observations are reported for 98% of teachers; discussions of measures to address weaknesses, development plans and mentoring are common consequences.
These figures should not be read as causal proof. TALIS is primarily survey evidence. Systems differ in definitions, frequency, culture and implementation. An appraisal labelled “developmental” may feel punitive in practice. A formal observation can be rigorous or superficial.
The value of the data is that they expose the architecture of appraisal: evidence, standards, feedback and consequences vary, and those design choices matter.
The common failure modes
The first failure mode is the showcase lesson. Teachers optimise one observed period rather than normal practice.
The second is single-snapshot judgement. One lesson carries a claim that requires a broader evidence base.
The third is rubric theatre. Visible behaviours are performed because they score well, even when they do not fit the learning.
The fourth is observer preference masquerading as evidence. “I would have taught it differently” becomes “you taught it poorly.”
The fifth is data worship. Student scores are treated as a clean measure of one teacher’s causal effect.
The sixth is feedback overload. Every possible improvement is listed, so no priority becomes practice.
The seventh is rating without support. The teacher is told what is weak but receives no protected time, modelling, training or coaching.
The eighth is development without follow-up. Plans are written and never checked.
The ninth is hidden stakes. Teachers do not know whether an observation is coaching, appraisal or both.
The tenth is individualising system problems. Curriculum gaps, staffing instability or weak behaviour systems are recoded as one teacher’s deficiency.
The eleventh is kindness without standards. Appraisers avoid difficult judgements so persistent weak practice remains unaddressed.
The twelfth is standards without humanity. Formal accountability becomes so threatening that teachers stop discussing genuine problems, which removes information needed for improvement.
A practical route for learners
Students should not be placed in the position of formally evaluating teachers as if popularity were quality.
Their learning evidence still matters.
If a school gathers student voice, the questions should be specific.
“Can you explain what you are learning and why?”
“Do you know what to do when you are stuck?”
“Does the teacher check whether the whole class understands?”
“Can you see how feedback helps you improve?”
These are more useful than “Do you like the teacher?”
Learners can also understand that an observer in the room does not mean the teacher is “in trouble.” Professional observation is normal in many education systems.
A practical route for parents
Parents usually see only fragments of teaching through homework, reports, communication and their child’s account.
That makes strong claims about teacher quality difficult.
If concerned, bring specific evidence:
“My child can complete the routine questions but cannot explain the concept.”
“Instructions in the online system are repeatedly unclear.”
“Feedback identifies errors but does not show what the next step is.”
Ask how the school monitors teaching quality and supports improvement rather than demanding one observation or one score.
Remember that professional appraisal should protect both sides: teachers deserve fair evidence and useful development; students deserve reliable standards.
A practical route for teachers
Before an appraisal, know the standard and the evidence.
Do not design a fantasy lesson that cannot be sustained. A useful observation should show real teaching.
Afterward, ask for specificity.
“What did you see that led to that judgement?”
“Which student evidence supports it?”
“Which change would you prioritise first?”
“What would improvement look like?”
“What support is available?”
Where you disagree, respond with evidence rather than defensiveness.
“This was silent independent practice because the previous lesson used paired comparison. Here is the sequence and the student work.”
That is professional dialogue.
If the feedback is valid, convert it into a small practice target and schedule the retry.
A practical route for appraisers and school leaders
Calibrate observers.
Two leaders should not reach radically different judgements because one prefers discussion and another prefers direct instruction.
Use standards that describe educational functions, not fashions.
Sample normal practice across time when stakes are high.
Separate descriptive notes from interpretation.
Use student work and curriculum context.
Prioritise feedback.
Connect weak areas to concrete support.
State the stakes clearly.
Return later.
And examine the appraisal system itself. Does it improve teaching, or mainly generate documentation? Do teachers understand the standards? Are observers trained? Are ratings consistent? Does professional learning follow identified needs? Do strong teachers receive useful challenge, or is appraisal only a deficit process?
The quality of appraisal should itself be appraised.
Evidence and caveats
The main contemporary evidence source used here is OECD’s Results from TALIS 2024: The State of Teaching, published 7 October 2025, together with OECD’s current teacher-appraisal topic guidance and the Singapore country note.
TALIS provides internationally comparable survey evidence about appraisal prevalence, methods, consequences and teacher experiences. It shows that observation is widely used, and that developmental follow-through such as improvement discussions, training plans and mentoring varies substantially.
The evidence does not establish a universal causal effect of appraisal on student attainment. Appraisal systems are complex packages embedded in different labour, curriculum and governance arrangements. Teachers and principals self-report many measures. Frequency is not quality. More observation can improve professional learning or increase performative compliance depending on design.
The safest conclusion is mechanism-based:
Appraisal has educational value when trustworthy evidence is interpreted against clear standards, converted into a proportionate judgement, linked to feasible professional action and checked later.
Everything else is administration around that loop.
The ordinary-weekday test
It is Wednesday afternoon.
A teacher has just finished an ordinary lesson.
An appraiser noticed that many students began independent practice with the same misconception.
The weak appraisal response is:
“Questioning needs improvement. Rating: 3.”
The stronger response is:
“Before independent practice, you took two volunteer answers and both were correct. In the first five books, four students then made the same conceptual error. Next time, let’s use an all-pupil check before release. I’ll help you choose the question. We’ll look at the first five independent responses next week.”
The difference is not politeness.
It is whether appraisal creates a professional learning mechanism.
FAQs
Is teacher appraisal the same as teacher observation?
No. Observation is one evidence source. Appraisal is the wider process of judging professional practice and deciding what follows.
Should appraisal be formative or summative?
Education systems use both functions. Developmental appraisal supports improvement; summative appraisal supports standards and accountability. Problems arise when purposes and stakes are unclear.
Can one lesson fairly determine teacher quality?
Usually not for a high-stakes broad judgement. One lesson can provide useful evidence, especially about a targeted practice, but teaching quality is a pattern across time and contexts.
Should student results be used?
They can contribute information, but results are influenced by many factors and should not be treated as a pure measure of teacher causal impact. The assessment’s validity and curriculum alignment also matter.
Are student surveys useful?
They can reveal aspects of students’ experience, but should use carefully designed questions and should not be treated as popularity ratings.
Does frequent appraisal guarantee improvement?
No. Frequency without clear standards, good evidence, useful feedback, support and follow-up can produce workload and performance theatre.
How does appraisal relate to autonomy?
Good appraisal can support professional autonomy by providing standards and feedback that calibrate judgement. Poor appraisal can suppress autonomy by rewarding compliance with narrow preferred methods.
What is the shortest test of an appraisal system?
Ask: “What changes in teaching after the judgement?” If the answer is “mostly the form gets filed,” the loop is incomplete.
The final idea
The easiest part of teacher appraisal is arranging for someone to sit at the back of a room.
The difficult part begins after the lesson.
What did the evidence really show?
What did it not show?
Which standard matters?
What interpretation is justified?
Which weakness matters enough to act on?
What support will make improvement possible?
When will anyone look again?
A profession needs accountability because children cannot wait indefinitely for weak practice to improve.
A profession also needs development because quality does not grow from judgement alone.
Teacher appraisal earns its place when it can hold both truths at once: standards matter, and people improve through specific evidence, supported practice and another chance to do the work better.
That is the mechanism.
Not observation for observation’s sake.
Not a rating for the database.
A loop from evidence to professional action and back to evidence again.
Surgical internal links
- How Professional Learning Cycles Work | Why One-Off Training Rarely Changes Classroom Practice
- How Instructional Rounds Work | See the Pattern Across Classrooms Before You Try to Fix the System
- How Teacher Performance Pay Works | Why Incentives Can Change What Gets Measured Before They Change Teaching
- How Teacher Inquiry Works | Turn Classroom Questions Into Evidence, Action and Professional Learning
- How Feedback Uptake Works | Why Advice Changes Learning Only When Students Can Interpret, Judge and Use It
Sources
- OECD. Results from TALIS 2024: The State of Teaching — The Demands of Teaching. 7 October 2025. https://www.oecd.org/en/publications/results-from-talis-2024_90df6235-en/full-report/the-demands-of-teaching_0e941e2f.html
- OECD. Teacher Appraisal. Current policy topic page. https://www.oecd.org/en/topics/teacher-appraisal.html
- OECD. Results from TALIS 2024 — Singapore Country Note. 2025. https://www.oecd.org/en/publications/results-from-talis-2024-country-notes_e127f9e2-en/singapore_8e88e29a-en.html
- OECD. Results from TALIS 2024 — Teacher Leadership and Autonomy. 2025. https://www.oecd.org/en/publications/results-from-talis-2024_90df6235-en/full-report/teacher-leadership-and-autonomy_24a4b36d.html