A CBSE Class 9 or 10 subject is marked out of 100: an 80-mark annual or board paper, and 20 marks of internal assessment. The 80 is the part everyone worries about. The 20 is the part that actually goes wrong, because it is the only part the school computes itself - four separate components, each collected on a different scale, each compressed down to 5 marks.
This article follows one student's Science term and shows the same four components producing an internal assessment of either 14 or 16 out of 20, depending on nothing more than when you round. Two marks - and in this case, a different grade band. Then it gives you the mark sheet format that prevents it.
What the 20 is made of
The standard split for Classes IX and X is four components of five marks each.
| Component | Marks | What it records |
|---|---|---|
| Periodic Test | 5 | Pen-and-paper tests run through the term, usually three of them |
| Multiple Assessment | 5 | Quizzes, oral tests, worksheets, class response - several small things |
| Portfolio | 5 | Classwork, notebook upkeep, reflections collected over the term |
| Subject Enrichment | 5 | Practical work, project work, or the speaking-and-listening component |
| Internal assessment | 20 | Added to the 80-mark paper for a subject total of 100 |
CBSE revises this periodically and some subjects differ - check the current circular for your class and subject before the term starts, not after.
Notice what this scheme has already decided for you. In most weighting problems the hard question is how much each assessment should count. Here the board has answered it: each component is worth exactly 5% of the subject. The work left over is the compression - taking three periodic tests marked out of 20 and turning them into a single number out of 5.
One student, four different scales
Ishaan, Class 9, Science, one term. Every component was collected honestly, and no two were collected on the same denominator - which is entirely normal.
| Component | Raw | Score | Exact, out of 5 | Rounded |
|---|---|---|---|---|
| Periodic Test (11, 18, 13 of 20 each) | 42/60 | 70.0% | 3.50 | 4 |
| Multiple Assessment (7/10, 8/10, 3/5) | 18/25 | 72.0% | 3.60 | 4 |
| Portfolio | 9/10 | 90.0% | 4.50 | 5 |
| Subject Enrichment (practical) | 10/20 | 50.0% | 2.50 | 3 |
Two marks, decided by when you round
There are two obvious ways to finish this, and they are one keystroke apart in a spreadsheet.
Round each component, then add
Add the exact values, then round once
Every one of Ishaan's four components landed on or above a half mark, so rounding each one individually pushed all four upward and the errors stacked instead of cancelling. That is not bad luck. Five-mark components produce a .5 far more often than a twenty-mark one does, because there are only ten possible half-mark positions to land on.
Now add the 80-mark paper. Ishaan scored 55.
| Method | Internal | Paper | Subject total | Band |
|---|---|---|---|---|
| Round once, at the end | 14 | 55 | 69 | 61-70 |
| Round each component | 16 | 55 | 71 | 71-80 |
Same student, same evidence, same term. The only difference is the order of two operations.
Round once, at the very end. Keep every component as a decimal until the 20 is complete, and let the single rounding happen there. It costs one extra column and it removes the only step in this calculation that can move a student across a boundary without anyone deciding anything.
Three periodic tests, one figure
The other fork is how you get from three periodic tests to the Periodic Test component. Schools do this at least three different ways, and Ishaan's 11, 18 and 13 out of 20 show what each is worth.
| Method | Score | Out of 5 |
|---|---|---|
| Average all three | 70.0% | 3.50 |
| Best of three | 90.0% | 4.50 |
| Average of the best two | 77.5% | 3.88 |
A whole mark separates the extremes, which is less than most people expect - a student who scored 11 and a student who scored 18 on the same paper end up 1.00 apart on a component worth 5. That compression is the honest reason to stop arguing about the method and write one down: the choice matters less than the inconsistency does. What you cannot have is two teachers in the same school using two of these, because then the component means something different per classroom and the 20 is no longer comparable.
Pick one, put it in writing before the first periodic test, and use it for every student in the cohort. Best-of-three and best-two are both defensible if a test was disrupted; averaging all three is the most straightforward to explain to a parent.
The mark sheet
This is the format worth copying. One row per student, exact decimals in the working columns, one rounding at the far right.
| Student | PT1 /20 | PT2 /20 | PT3 /20 | PT /5 | MA /5 | Port /5 | SEA /5 | IA exact | IA /20 |
|---|---|---|---|---|---|---|---|---|---|
| Ishaan Verma | 11 | 18 | 13 | 3.50 | 3.60 | 4.50 | 2.50 | 14.10 | 14 |
The four /5 columns hold two decimal places. Only the last column is a whole number.
Two rules make it work. Every component column is marks obtained divided by that component's own maximum, times five - never a mark somebody estimated straight onto a five-point scale. And the raw denominators live in the sheet too, so that when a colleague asks in March why Ishaan's Portfolio is 4.50, the answer is 9 out of 10 rather than a recollection.
Every figure in this article was checked by running the components through the same calculation our term reports use, which divides one total by another rather than adding four rounded pieces - it returns 14.1 of 20, or 71%, for Ishaan.
Three things that will come up
A student misses a periodic test
Average the tests they actually sat rather than entering a zero. If Ishaan had missed the second test, his Periodic Test component is (11 + 13) ÷ 40 = 60.0%, or 3.00 of 5 - not (11 + 0 + 13) ÷ 60 = 40.0%, which describes a student who sat three papers and failed one. Absent and scored-nothing are two different facts and only one of them is true.
A component never happened
If Subject Enrichment was not run for a class, the subject is not out of 100 for that class. Do not silently award the full 5, and do not award 0. Either run the component late or record the subject out of 95 and say so on the report, because a mark sheet that quietly invents 5 marks for everyone is one that nobody can reconcile against the evidence.
A component's raw total changes mid-term
This is exactly why the sheet stores fractions rather than five-point estimates. If the third worksheet is marked out of 10 instead of 5, only the Multiple Assessment denominator changes and every other column is untouched. If your sheet holds hand-assigned scores out of 5, you have to revisit every student.
Write down the method, then the marks
None of this is difficult arithmetic. The reason it produces two different answers for Ishaan is that four small decisions - which periodic-test rule, what to do about the absence, whether the component ran, when to round - are usually made silently and separately by whoever is filling the sheet that week.
Decide them once, before the term, in a paragraph anybody can read. Then the 20 marks become what they were meant to be: the part of the grade your school can actually explain, as opposed to the part that arrives from the board.
The same logic applies once you have several assessments with different maximums to combine, which is covered in calculating a term percentage when tests have different total marks. And because internal assessment is collected on paper while much of the rest of the term is not, the record has to hold both - our worked example of keeping one student record across paper and online tests shows what happens when it does not.
Teemly stores each component against its own maximum and derives the internal assessment from those fractions, so the rounding happens once and the raw marks stay visible behind the total. You can see how that fits with the rest of a term on our assessment software for schools page.