A single digit can at once summarise knowledge, pace, behaviour and the local strictness of the school. The same grade can therefore be awarded fairly and still not be compared automatically across classes. An anonymous test measures something other than a report card; an average can hide a specific gap; and the school’s strictness may not be visible from the digit when a pupil applies further.
1. A report card at which three meanings miss one another
Imagine a parents’ evening in January. A parent sees a three in mathematics and reads mediocrity from it. The teacher sees the same three and knows that the pupil can handle the algorithm but falls on the word problem. The headteacher defends the comparability of the report card against neighbouring schools, because grades enter admission. The guidance counsellor reads behaviour, homework and whether the child brings a notebook. The pupil reads a label: either I can, or I cannot.
The debate about the fairness of grades breaks into several different needs. The parent compares school success, the teacher looks for a gap in skill, the headteacher watches the consequences for the move to secondary school, the counsellor attendance and discipline, and the pupil easily turns the digit into a verdict on the self.
The three can be relevant for each of them, but it is not enough by itself for choosing a field of study. Without a description of the task and the criteria it is not clear how much of the result is knowledge, pace, support at home, the strictness of the teaching staff or the composition of the class. A “good grade” is therefore not a stable property of the pupil, but a local sum of several measures.
Before we start comparing strictness, we have to decide whether we are reading mastery of the material, work habits, behaviour or standing relative to the class. A test, a notebook and a report card may weight each of these things differently.
A grade is not knowledge. Knowledge is not effort. Effort is not readiness for secondary school. And a report card is not an admission procedure.
— Jiný Kontext
What studies of grades say — and what does not follow from them
We know: in the Czech data examined, grades are related to the local strictness of the school, to sex and to educational aspirations even after standardised test results are taken into account.[3][5][6][22] We do not know: observational data alone do not split this relationship into the influence of the pupil’s behaviour, the teacher’s expectations, family support, the composition of the school and other unobserved factors. They cannot determine the cause of a particular child’s grade, nor support the claim that changing one digit will by itself change that child’s future path.
2. What all fits into one grade
The digit on the report card is produced by a combination of the pupil’s performance, the tasks chosen, the teacher’s rules and the context of the class. Each component changes the information value of the result. Even two schools with the same framework curriculum may therefore not mark the same work in the same way.
Five layers of one classification
- Knowledge and skill: what an anonymous task could still capture.
- Performance in time: writing pace, stage fright, homework, a forgotten aid.
- Socio-emotional layer: conscientiousness, sitting quietly, the impression of the family.
- Local norm: the strictness of the teaching staff and the composition of the class.
- Use of the number: an application, a self-image, a reward, a punishment.
The Education Act requires that the assessment of a pupil be pedagogically justified and professionally sound. The decree on basic education leaves schools wide autonomy in the choice of numerical classification, verbal assessment and a combination of the two. That is an essential feature of the law. It is not automatic comparability of a one between two buildings.[1][2]
This is where the first error of most readings at home appears. The headline announces that the child “has a three”, but in reality it compares four different profiles. One understands the concept and bungles the arithmetic. The second commands the formula and does not know why it holds. The third writes excellently and loses points for the notebook. The fourth sits in a strict school, where the same work would elsewhere be marked a two. All four can leave with the field “3 — good”.
Breaking the grade down shows where to look for a correction. A penalty for forgotten sports kit says nothing about equations; a one that fails at the entrance exam may reflect the local norm; and a difference between identically written tests may come from further assessment criteria. The sentence “has a three”, without these facts, does not say what the pupil should practise next.
Illustrative diagram: The same digit can, in another regime of strictness, behaviour and use, produce a different path.
3. Useful assessment must make correction possible
Teachers and parents usually look for a grade that ranks the class “fairly”. That is reasonable, but incomplete. Equally important is what the error looks like, whether the pupil recognises it, how much it costs to correct, and whether the verdict is reversible. A slip in a dictation is a cheap error if the comment arrives in time. A three on the report card in May, from which an application is filed in June, sometimes is not.
Black and Wiliam, in the review Assessment and Classroom Learning (1998), showed that assessment which tells the pupil during learning where they stand and what to do next is associated with improved results; this is not one test in one school, but a synthesis of the empirical literature on formative use. Hattie and Timperley, in The Power of Feedback (2007), distinguish feedback on the task, on the process and on the person; feedback of the type “you are a pupil who gets ones” is a different object from “this is where the unit is lost”. These are peer-reviewed reviews, not proof that the Czech report card is pointless. They show what a number is enough for, and what it is not.[16][17]
| Type of failure | What it looks like | How to test it | What limits the harm |
|---|---|---|---|
| Compression | Four different gaps receive the same digit | Ask which skill the grade is punishing | A separate field for knowledge and for behaviour |
| Local norm | The same test, a different grade between schools | Compare the grade with an anonymous test, not with a feeling | A comment on the strictness of the teaching staff |
| Non-cognitive admixture | A notebook and silence improve the grade on the same test | Anonymous work beside a named report card | Assess behaviour separately |
| Wrong use | The digit becomes the choice of a field of study | Whether the pupil can say what to correct | Entrance exams and a second figure beside the grade |
An average makes sense only when the school can show what behind it can be corrected. When applying for a Maturita programme it is especially dangerous to mistake local strictness for inability. A report card does not predict a future professional error; it can, however, offer the reader a concrete trace for learning — or only a stamp.
Beside the pupil’s performance it is necessary to watch whether the gap is visible, whether the assessment can be corrected, and what effect it has on the next path. A school with a smaller share of ones, but a clear description of skills mastered and not mastered, can give a family more useful information than a mild one without support.
4. A grade summarises performance, but does not explain it
Grades are needed because they replace part of impression with a brief record. The problem begins when one field starts to stand for the whole of ability. In its publication on grades and educational expectations (Grade Expectations, 2012) the OECD notes that grades serve several functions at once — they inform, rank, motivate, punish — and that reading them as a single yardstick is therefore risky. This is an international synthesis on PISA data, not an audit of one Czech class.[18]
Every report card contains hidden decisions. Who chose the tasks? Does homework count? Is late submission punished? Is a one for the few, or for most? Is the exact procedure assessed, or the result? The answer to any of these questions can change the order of pupils who would stand next to one another in an anonymous test.
In a study by Münich and Protivínský (2022), the IDEA think-tank at CERGE-EI measured the strictness of grading on the Czech PISA 2018 sample: that is, among fifteen-year-old pupils, in mathematics, reading and science, on a representative school sample. They defined strictness as the relationship between the test score and the grades at a given school. In their model, the classification of pupils with comparable test results differed between schools by about one grade on average. This is one observational study on one PISA cycle, not proof of teacher dishonesty and not an experiment with the same written paper in two schools. It shows that a grade also contains a local marking bar.[5][6]
What must accompany the digit
- Subject, period (term, year) and whether it is coursework or the report card.
- What enters the average: papers, homework, activity, behaviour.
- Frame of reference: this class, this school, or an external test.
- Population: one pupil, or a comparison between schools.
- Known limits: anonymity, sex, socioeconomic status.
An analysis of specific tasks helps a family more than someone else’s ranking of schools. A list of what the child can and cannot do offers a plan for correction; an average of 2.4 without that information does not. A public test can complete the comparison, but further learning has to start from the gap that was found.
5. Compare schools, not digits
The market in grades can be described without a false podium of winners. The map below does not say which school is generally first. It shows the work around which each regime builds assessment, what kind of information it brings, and which failure the reader should look for on purpose.
| Regime | Main working position | What to check on the report card | Typical reason for the choice |
|---|---|---|---|
| Classification 1–5 | Quick ranking and an administrative close | What exactly entered the average | Legibility for parents and secondary schools |
| Verbal assessment | A description of progress and the next step | Whether the text says what to do, or only praises | Diagnosis at a younger age |
| Criteria table | Met / partly / not, against a specific aim | Whether the pupil sees the criteria in advance | Comparability without a single digit |
| Anonymous test | A comparison of knowledge without a name and a notebook | Whether the test measures the same material as the grade | A check on local strictness |
| Entrance examination | An external filter for secondary school | Whether the grade and the test speak of the same pupil | A decision the report card alone cannot carry |
The table does not judge teachers as a group. Assessment can differ between year-groups and in the course of a year. Its purpose is to expose a digit that does not show a specific gap and, instead of helping with learning, turns into a label on the pupil.
How to read the diagram: A one here is not a ranking of quality. The diagram joins performance, the rules of assessment and the purpose for which the grade is used.
6. Behaviour is not knowledge. It still enters the same field
In the study What do the grades on the report card hide? (Münich, Protivínský, 2018), IDEA compared report-card grades with anonymously marked tests at the transition to secondary school. In mathematics, girls with comparable test results had, in the model used, a grade better by 0.6 of a classification step on average than boys. In Czech the authors did not find the same difference at comparable results. This is an observational comparison of non-anonymous classification and an anonymous test in one educational transition, not the same paper once with a name and once without, and not proof of a universal mechanism.[3][4][22]
As a possible explanation the authors discuss socio-emotional skills and other features of performance that an anonymous test does not capture but a teacher can see in class. The study itself does not determine how much of the difference is explained by conscientiousness, behaviour, teacher expectations or another factor. A school may also want to assess work habits; it should then say that it includes them in the grade. Otherwise the parent reads the three only as a message about equations.
Federičová (2019), using data on applications to multi-year grammar schools, found that filing an application was strongly related to grades, especially to ones, more than to the expected result on the test, and that differences in applications by boys and girls persisted even at comparable chances of admission. This is an observational study of one type of transition; it does not by itself prove that the grade caused the decision. It does show that the grade enters the setting in which families make the choice.[7]
7. A strict school is not the same as a demanding school
In conversation, strictness is confused with quality. It is not automatically the same thing. On PISA 2018, Münich and Protivínský (2022) divided schools by how strictly they grade relative to the test. In strictly grading schools, 7 percent of the fifteen-year-old pupils at those schools received grade 1 in mathematics; in mildly grading schools, 25 percent. The most common grades in strict schools were steps 3 and 4, which together were taken by about two thirds of the pupils. This is the distribution of grades inside these groups of schools, not the claim that 7 and 25 percent are the share of pupils with the same PISA score. A separate conclusion of the study says that at the same academic results the grade differs between schools by a step on average. The two sentences must be kept apart.[6]
At comparable test results, pupils from strictly grading schools reported lower study ambitions than pupils from mildly grading schools. Of pupils with a one in mathematics, 87 percent wanted to complete a university degree; of pupils with a three, 39 percent. These are aspirations in the PISA 2018 questionnaire, not later actual diplomas. The result documents a link between grade, local strictness and aspirations; without further research it does not determine how large a part of the difference the grade itself caused.[5][6]
An illustration, not a report. Adam, at a strict village school, understands geometry, brings home a three from written tests, and the family reads an average. Tomáš, at a mild city school, collects ones for notebooks and, at the unified entrance exam, meets a task that the algorithm from the worksheet does not solve. The report card told both of them something. It need not have told either of them the same thing as what they can do.
8. The test and the grade capture different parts of performance
In the sample examined, girls in the ninth year had better grades on average, while on anonymously marked entrance tests in mathematics they achieved lower results on average than boys. That is how the 2018 IDEA study opens its problem. It is not, however, the same work marked in two ways. The grade summarises a longer period in class; the entrance test captures performance on a different task and at a different moment.[3][22]
In the 2022 study, 20 percent of girls and 13 percent of boys have a one in mathematics. Again this is a description of classification, not of anonymous literacy. Anyone who turns these shares into the sentence “girls are better at mathematics” joins two layers. Anyone who turns the test into the sentence “grades are useless” erases information about work, attendance and support at home that the teacher sees and the test does not.[6]
An honest reading therefore keeps two numbers side by side. It does not pick the one that better fits the wish.
A pleasant average can hide the same old problem
A good-looking 1.8 acts as a grade of understanding. It is not that automatically. When reading a report card it is necessary to check separately whether the average has not merged two different gaps into a smooth but non-existent picture of “it is going well”. The quality of the digit lowers the cost of talking to a secondary school. At the same time it can raise the cost of finding a substantive gap, because the result does not look suspicious.
9. PISA is not a report card. A report card is not PISA
The Czech debate often reads grades as proof that “children cannot do mathematics”. The international survey measures a different object. In PISA 2022, where mathematics was the main domain, Czech fifteen-year-old pupils scored 487 points. The Czech School Inspectorate ranks them among countries with an above-average result; nine countries achieved a statistically significantly better result, twelve had a comparable result. In the country note on Czechia the OECD states that 74 percent of Czech pupils reached at least level 2 in mathematics, more than the OECD average of 69 percent. This is the literacy of fifteen-year-olds in one cycle, not grades on a report card and not the Maturita. The OECD average in PISA 2022 mathematics is 472 points in the summary publication; older cycles had a different average, so 487 should not be read against a number from another wave.[11][12][13]
TIMSS 2023 measured pupils in year 8. Czech pupils scored 518 points in mathematics, above the average of all participating countries (478) and above the average of the thirteen EU member states in that survey (503). Czechia took part in the eighth-year survey in 1995, 1999, 2007 and 2023; against 1995 this was a fall, against 2007 an improvement. This is a different age, a different framework of tasks and a different year from PISA 2022. Neither survey says what grade a pupil has in geometry in March.[14][15]
The tension between decent literacy on a test and a feeling of failure at school is therefore not necessarily a dispute about “which number is true”. It is a dispute about what each number sees. PISA does not punish forgotten sports kit. A grade need not see a task with a graph from an unfamiliar situation.
10. What has to be checked behind the digit
Before the report card the parent guesses. With the report card the parent checks something harder: whether the digit supports the sentence they want to make from it. A one is navigation toward an impression of success. It is not automatic proof that the pupil knows the same material as a pupil who gets ones a district away.
Slow writing, an error in units and standing in a weaker class can all contribute to the same three, but they describe different parts of performance. The summary verdict “is average” therefore has no support without an analysis of the specific work.
A four-step check of the grade
- Existence: the number is from this term and this subject.
- Components: which papers, which weighted average, which behaviour.
- Support: a specific gap really supports a specific sentence about ability.
- Scope: school, class and sex fit, and whether an external test exists.
The most reliable working procedure therefore separates reading from deciding. First comes a table: what the pupil can do, what not, what is behaviour, what is the school’s norm. Only then the application. If the teacher cannot say which task the child should repeat, the claim is not raised to talent or inability merely because it is on paper.
11. A summative verdict is not a map of learning
The difference is not whether the school “assesses”. It is when, and for what. A summative grade closes a period. It says where the pupil ended, at a moment when in that term they can often do nothing more about it. A formative description runs along the way. It says which step holds, which does not, and what to try. In Strategy 2030+ and in the methodological recommendation for years 1 and 2, the Ministry of Education separates the formative approach from mere verbal style: a text without a next step does not fulfil a formative function.[8][9]
From the 2027/2028 school year it will not be possible to assess year 1 on the report card with a grade; verbal assessment, including criterion-referenced assessment, will be allowed. From 2028/2029 the same for year 2. This is a change in the lowest years, not the abolition of classification in the ninth year, and not a promise that a new form will by itself improve learning. Schools may choose among classification, words and a combination earlier; the duty shifts gradually.[9][10]
The aim of the ministry text is not a resignation of standards. It is a reduction of outside pressure where a relationship to learning is only just being formed. That is a policy choice with empirical support in the literature on feedback. It is not proof that from the sixth year the digit is harmless.
12. The price of a one is the price of the raw material, not the price of the path
A school uses several languages at once. The teacher pays with time spent writing in the class register. The parent pays with calm, or with anxiety. The system pays with the grade as a filter. The pupil pays with a self-image. A direct comparison “has a one”, without knowing what the one will be used for, therefore misleads.
A cheap one at a mild school can be economically excellent for peace at home. The same one can be expensive at the entrance exam if it does not cover the material of the test. A strict three can be needlessly expensive if the family turns it into an apprenticeship track, even though an anonymous test says something else. It can, however, be cheaper if the pupil sees the gap from the comment and corrects it before the application is assembled.
The real price of an accepted decision
Grade + teacher’s time on a comment + practice at home + an external test + the price of a wrong application.
A decision is cheap only when it passes a check and can still be changed.
13. What to read from a report card
A reasonable reading does not start with the digit, but with the kind of use. The recommendations below are not a ranking of schools. They are a signpost to a first analysis. For each subject it makes sense to ask the same question twice: what the number measures, and what it may be used for.
| Need | First figure | Decisive test | Unacceptable shortcut |
|---|---|---|---|
| Find the gap | A comment on one task | The pupil can say what to try tomorrow | Read only the average |
| Compare with classmates | The grade in this class | Whether the class is strong or mild | Take a three as a national percentile |
| Compare schools | PISA/TIMSS, entrance exams, IDEA | The same population and the same year | Mix 487 PISA points with a grade from a notebook |
| Choice of secondary school | Test + grades + aspirations | Whether the school’s strictness has lowered the aim | File an application on the strength of one three |
| Behaviour | A separate verbal field | Whether the notebook did not enter mathematics | Punish sports kit with an equation |
| Younger years | A criterion-referenced description | Whether the text passes the formative test | Replace the digit with empty praise |
Neither a foreign review of formative assessment nor the Czech PISA average will show whether a particular teacher counts aids, or what stands behind a three in algebraic fractions. The share of ones at a school also does not describe the cause of one pupil’s grade.[13][6] The rules of the specific class therefore have to be checked.
14. Fourteen days that say more than a class average
A short analysis of the report card can replace both panic and a wave of the hand. It is not meant to confirm the label “a pupil who gets ones”, but to find the specific skill hidden behind the digit and to check whether it can be corrected.
Fourteen days from impression to decision
- Days 1–2: Write down the subjects in which the number surprised you, and the price of a possible wrong choice of field.
- Days 3–4: Ask for a breakdown: which papers, what weights, whether behaviour entered.
- Days 5–7: Give the pupil three tasks from the gap and keep a register: can / can with a hint / cannot.
- Days 8–10: Compare the grade with one external figure, if it exists: entrance exams, a comparative paper, another teacher’s comment.
- Days 11–12: Separate what is knowledge, what is pace, what is the notebook.
- Day 13: A stress test: a mid-range task, a task from a graph, a task without a formula.
- Day 14: Decide what the grade may be used for, and set a date for the next review.
What needs to be tested
The first test concerns the components of the average. The second concerns the local bar: is a one rare, or common? The third concerns sex and behaviour: would an anonymous hand receive the same digit for the same work? The fourth concerns use: is an application to be made from the number, or only a plan at home?
How to test what a grade hides
Put several familiar tasks into the week and watch whether the pupil fails where the grade punished, or elsewhere. Ask about the contradiction between notebook and paper. Require one sentence “what tomorrow”. Repeat the task in a different order. And do not assess only the correct result: record whether the pupil admitted not knowing, or masked the gap with an average.
The pilot should also have exclusion conditions. For example: the teacher cannot say what the number measures; the grade will be used as the only criterion for an application; behaviour and knowledge cannot be separated. A high average must not override a gap whose impact the family cannot accept.
15. The entrance exam sees something other than the class register
The unified entrance examination is another layer, not a correction of the report card. Cermat sets it as a unified test in mathematics and in Czech for applicants to Maturita programmes. Report-card grades may join it as a further criterion, according to the school’s setting. This is a different population from PISA 2018, a different age from TIMSS year 8, and a different purpose from a term average.[19]
That is both an advantage and a warning. The family can at last see an anonymous task. At the same time it must not mistake a percentile from the entrance exam for proof that the teacher “measured nothing” all year. The entrance exam does not see homework or pace in the sixth year. The report card does not see a non-standard task that the workbook did not practise.
Federičová (2019) found that filing an application to a multi-year grammar school was strongly related to grades, especially to ones, more than to the expected result on the test. The selection of applicants therefore begins before the pupil sits down to the unified paper. The study itself, however, does not allow us to say how many children did not apply because of the grade and how many for other reasons.[7]
An honest comparison therefore keeps three objects apart: the school grade, the unified test and international literacy. They can be joined in one sentence about talent. That cannot be passed off as a measurement.
The unified exam gives the parent a second point of control beside the teacher’s assessment. It makes it possible to compare whether school preparation worked with similar material and whether the grade agrees with a more anonymous performance. Even a percentile, however, is not a complete measure of talent and cannot be mistaken for a PISA result.
16. A formative stamp is not a formative approach
From the 2027/2028 school year the ministry takes the report-card grade away from the first year, and from 2028/2029 from the second year as well. Schools may use verbal and criterion-referenced assessment earlier. The methodology warns expressly that a verbal text without a next step does not fulfil a formative function. A change of stamp is therefore not a change of teaching.[9][10]
That matters more than the dispute “for or against grades”. The family is not buying a form. It is buying whether the child will hear what to try tomorrow. A teacher who writes a paragraph of praise without a gap instead of a three has only translated compression into prose. A teacher who, with a classification, can say “this is where the unit is lost” already has the formative layer — the digit merely accompanies it.
Strategy 2030+ emphasises a formative approach and at the same time admits that a summative close of a period will remain part of the system. This is a strategy document, not a promise that from the sixth year the average will disappear. The lowest years are the first test. They are not the whole reform of reading a report card.[8]
A parent who in 2026 reads a three in the eighth year is not yet protected by this change. The decree’s autonomy of the school still applies, and it is still true that a one from a strict building and from a mild building are not the same object. The reform of years 1 and 2 is a laboratory field. It is not an automatic transfer into the ninth year, where a digit is turned into an application.
17. The result is decided by the use of the grade
The first generation of school assessment needed to rank the class quickly. The next generation added an average and entrance exams. Today’s report card sells something larger: a place from which the family reads the future.
Schools therefore do not compete only over who “teaches better”. They compete — often unintentionally — over whose digit will be readable outside the building. The more comfortable a single five-step scale is, the higher the price of leaving an error can be. Aspirations adjust. The application is not filed. The pupil stops distinguishing what they can do from the digit they hold.
The brevity of classification is useful when a result is handed on, but it requires conscious reading. The family needs a supplementary figure on skill, behaviour kept separate, information about local strictness, and a later check on whether the original reading still holds.
A useful grade is not automatically a one from a mild school or a three from a strict one. It must make it possible to find what the pupil has mastered and what should be corrected; its gaps should be visible, reversible and still usable for learning.
The moment the digit only closes a notebook, we choose a shortcut. The moment an application, a self-image and an estimate of talent are made from it, we choose the structure of a path. And then the question is no longer: “What grade does the child have?”
Before the next decision it is therefore necessary to find out: What exactly did this digit measure, what did it erase — and which decision may we still make from it?
— Jiný Kontext