Here's the worry, stated plainly: "I can't grade what I don't understand. If my student writes that the molarity is 0.103 M, I have no idea whether that's right."
That worry rests on a misunderstanding of what marking a lab actually requires. A science teacher grading thirty lab reports on a Sunday evening is not re-deriving anyone's chemistry either. They are checking the report against criteria: Was the procedure followed and documented? Was data actually recorded? Was the calculation attempted, and does it use the student's own measurements? Does the conclusion cite the data? Every one of those is a reading question, not a chemistry question.
The chemistry knowledge lives in two places you already have access to: the answer key (every serious curriculum provides one — checking a calculation means comparing against it, not re-deriving it) and the criteria themselves, which are written once, by someone who knows the subject, and then applied by you. Your job is the honest part no answer key can do: reading what your student actually wrote and ticking what's actually there. (Who writes those criteria is the real question, and Part 3 takes it head-on.)
Mark every lab out of 100, split into three buckets. The split below is the one we use on our own platform, and it is deliberately simple enough to read off without explanation:
| Bucket | Points | How it's earned |
|---|---|---|
| Doing the lab | 50 | 40 for completing the lab honestly, up to 10 for technique — careful, systematic work rather than rushing. If you have no way to judge technique, all 50 is completion credit. |
| Calculation questions | 20 | 20 correct (checked against the answer key), 15 attempted with the student's own data, 0 not attempted. |
| Analysis & conclusion | 30 | 30 strong (hits 3–4 criteria), 22 partial (1–2 criteria), 15 attempted, 0 blank. |
Notice the floor this creates: a student who genuinely did the lab, attempted the calculation, and wrote something in the analysis lands around 80 — a B-. That is not grade inflation; it is the same logic a classroom lab course uses. A student cannot fake their way through actually running an experiment, recording data, and writing a report — completing the work is evidence of learning, and the marks above the floor are earned by getting things right. The spread between "did it carelessly" and "did it well" is where the grade differentiates.
The analysis bucket is where parents worry most, and it's where criteria do the work. For each question, write (or use) 3–4 observable criteria — statements you can verify by reading, not by knowing chemistry. For a titration analysis question, the criteria look like: "identifies the equivalence point from their own data", "explains why the indicator changed color", "uses the term neutralization correctly", "connects the result to the balanced equation." You read the student's answer, you tick what's present. The chemistry is in the criteria; the honesty is in your reading.
At the end of the course, average the lab marks (excluding any labs you didn't mark — and note the basis, e.g. "based on 12 of 31 labs marked"), then read the letter off the standard band table:
| Percent | Letter | Grade points |
|---|---|---|
| 93–100 | A | 4.0 |
| 90–92 | A- | 3.7 |
| 87–89 | B+ | 3.3 |
| 83–86 | B | 3.0 |
| 80–82 | B- | 2.7 |
| 77–79 | C+ | 2.3 |
| 73–76 | C | 2.0 |
| 70–72 | C- | 1.7 |
| 67–69 / 63–66 / 60–62 | D+ / D / D- | 1.3 / 1.0 / 0.7 |
| Below 60 | F | 0.0 |
The total is a suggestion you make to yourself — you can override it. The buckets and criteria exist to make your judgment consistent and defensible, not to replace it. You are the administrator of your homeschool: if the arithmetic says 87 but you know the student coasted on a technicality, or earned more than the ticks can show, set the final mark you can stand behind — and note why, so the record explains itself later. (Override the final mark, not the point values — keep the buckets fixed across the year, or the marks stop being comparable between labs.) A mark you can override is a tool; a mark you can't is just someone else grading your child.
Put the percentage on the transcript alongside the letter. Admissions offices routinely recalculate GPA from the numbers under their own scale, so a transcript that shows "Chemistry with Lab — 91 (A-)" is doing the reader's work for them. Whatever scale you use, state it on the transcript (our Lab Science Guide covers the transcript format itself) and apply it consistently — consistency, not the particular boundaries, is what makes the grade credible.
First, to be clear about what you're getting: the sheet below is our own marking model — the same buckets, point values and floors our platform applies to every lab — published here for you to copy and use with any lab, from any curriculum. As far as we know, nobody else hands you one; grading advice for homeschool labs mostly stops at "find a curriculum with assessment guidelines."
And here is the truth about it: the rubric is the easy part. Buckets, point values, tick boxes — that structure took Part 2 a few paragraphs, and any parent can apply it in minutes. Everything difficult about marking a science lab lives in one place: the criteria lines. "Explains the indicator change in terms of the acid being consumed" is where the chemistry actually sits — and criteria like that have to be written per lab, per question, by someone who knows what a strong answer to that specific question contains. That is why even our sheet has blank lines in the analysis bucket: criteria belong to each specific lab, and no one-page printout can contain them. The blank lines are the real problem this page has to solve.
So if you're marking labs by hand — a kit curriculum, a co-op class, a textbook's experiments — you have one honest technique available:
Your curriculum's answer key gives model answers, not criteria — but a model answer can be decomposed. Take the model answer, split it into its 3–4 distinct claims, and rewrite each claim as a statement you can check the student's answer against. A model answer reading "The pressure doubled because halving the volume at constant temperature doubles the collision frequency of gas particles on the container walls" decomposes into: states pressure doubled when volume halved · attributes the change to particle collisions · notes temperature was held constant. That's a marking guide for that question — written by you, from material you already own.
Be honest with yourself about the cost. This works, and it does not require re-learning chemistry — but it takes 15–20 minutes per lab, every lab, all year, and its quality depends on how good the answer key's model answers are. Where the model answer is thin ("Answers will vary"), you are back to judging chemistry you don't remember — the exact situation you were trying to engineer away. This authoring burden, not the ticking, is the real reason marking lab science feels impossible; no printable sheet removes it, including ours below.
For the labs where you do author your own criteria, print this page (button top right) — the sheet below fits on one page and gives the method somewhere to live. The criteria lines are blank because, as you now know, they belong to each specific lab.
| Bucket | Check what's present | Points | Award |
|---|---|---|---|
| Doing the lab | Lab completed start to finish Data/observations recorded during the lab, not reconstructed after Procedure followed and documented Careful technique (steady increments, repeated trials, units recorded) — if unjudgeable, award with completion | 50 40 completion 10 technique |
____ |
| Calculations | Attempted, using the student's own measured data (15) Correct per the answer key (20) | 20 | ____ |
| Analysis & conclusion |
Criteria for this lab (write 3–4 per question, tick what the answer shows): ______________________________________ ______________________________________ ______________________________________ ______________________________________ 30 = strong (3–4 ticked) · 22 = partial (1–2) · 15 = attempted · 0 = blank |
30 | ____ |
Here is the method applied to one analysis question from an acid-base titration lab. The student's answer is reproduced above the criteria — that ordering matters, because the criteria are only meaningful when read against the actual answer:
Three of four criteria ticked — a strong answer: 30 points for this bucket. Notice what just happened: you assessed a chemistry answer accurately without needing to know any chemistry yourself. The fourth criterion (a genuinely subtle distinction) is exactly the kind of thing the criteria carry for you — you didn't need to know the distinction existed; you only needed to read that the answer doesn't make it.
To see what a full, well-documented report looks like before you mark one, our annotated sample lab report walks through every section and what each one is evidence of.
Then don't judge rightness — that's what the answer key and the criteria are for. Your irreplaceable contribution is verifying the work is real and reading what's actually on the page. If a specific answer stumps both you and the answer key, that's a fine thing to mark "attempted" and ask the student to explain aloud — having them teach it back is both a check and the best review session they'll get.
The method already does, in a bounded way: completion carries real points because completing a lab honestly cannot be faked, and "attempted" tiers exist in both written buckets. What effort should never do is buy back points for wrong answers marked right — tick what's there, and let the floor do the kindness.
Consistency beats coverage. Marking a representative set carefully is more defensible than ticking all of them from memory — just say so with the number ("based on 12 of 31 labs marked") so the grade's basis travels with it. What you should never do is average in labs you didn't actually read.
For a homeschool transcript, you are the school administrator — issuing course grades is precisely your role, the same way you grade their math. What makes a homeschool grade credible to an admissions office is not who issued it but what stands behind it: dated lab reports, a marking method applied consistently, and hours that were logged rather than estimated. (For what colleges actually look for in lab science documentation, see the Lab Science Guide — and for whether virtual labs count at all, its Part 6.)