Blog
Consistent grading across class sections means every section of the same course is marked against one answer key and one rubric. The first section and the third section take the same test, and you give every section the same key and the same rubric. Graider compares each answer with that key, and you check the marks at review in every section. That shared starting point is what consistent grading means here.
The standard slips when it lives only in memory. A wording you accepted in the first section can lose the point later, and a partial you allowed early can tighten by the last section. The written key and the written rubric give every section the same starting point. You review each section, and your override is the score that goes back.
Graider is the AI red pen for paper grading, a product of DM Liquidity LLC. You upload the test and your answer key, photograph the stack, and Graider marks each question against that key or the rubric you provide. Each question gets one score and one comment, and a whole essay gets one score and one comment. You review and override any mark, toggle feedback, keep the pen, and hand the work back as a PDF or an email. Each section follows that path in AI test grading.
Your key. Their papers. You review every mark. You keep the pen.
Write the answer key and the rubric once, before any section is marked. Upload that same key and that same rubric when you set up the test in each section's class. Photograph each section's stack on its own, and confirm the roster match for that section.
On the key, list the full-credit response, every equivalent you accept, and the points. "Photosynthesis converts light energy into chemical energy" might be 2 points. Write the number on every line you accept.
Open items live on the rubric you upload with the key: an essay, a multi-step explanation, or a response with many right wordings. The rubric feeds the one score and the one comment. You review that score and edit that comment. The same standard is the one in grade against your own rubric.
Number each question the same way on the test students write on and on the key. Question 4 in the first section is question 4 in the third section, and question 4 on the key. Print those numbers on every section's copy, and keep the question order identical, so the key you wrote once still lines up with every stack.
If a question has parts, number each part on the test and on the key when you score the parts as separate questions. Each numbered question gets one score and one comment. If you score the parts together, keep one number and list both the full-credit combination and the partial-credit combination there.
Decide partial credit before you mark the first paper in the first section. Put the example on the rubric, or on the key when the item has a fixed answer. Write the partial response and the points it earns. "Names the cause and leaves out the document" is 1 of 2. "Shows the method and drops the units" is 2 of 3. Give every section the same partial credit examples, and check each partial at review. How to write a rubric AI can follow walks through criteria, partial credit, and examples you can upload with the key.
Graider does not read or score diagrams, graphs, or drawings. If a question asks for a sketch, a graph, or a labeled drawing, set those points yourself inside the one score for that question.
After a section is marked, you review before you hand back. Keep a score when it matches your key or your rubric. Override it when you disagree, and edit the comment.
The only flag is the Low confidence badge on an answer Graider found hard to read. Open every badge in every section. Read the student's answer as text, compare it with the paper and your key, and override when you disagree.
Before you finish a section, check:
An override is your judgment on that question for that paper. When you change a score, name the reason in the terms on the key: a full-credit line, a partial-credit line, or points you set from the rubric. Put that reason in the comment.
Use that same reason when the same response shows up in another section. If the first section shows you that a wording deserves full credit, override that paper at review, and list that wording on your key for the next test. Look for the same wording in the other sections and override those papers the same way. A response unique to one student stays an override on that paper. Your judgment stays the final word.
Set the feedback toggle the same way on each section's test. Turn feedback on when comments should travel with the scores, and off when you will go over the test with the class. Make that choice once and use it for every section.
Edit the comment so it sounds like a note in pen. On a short answer, name the question and one next step: "Question 4: the key awards 1 of 2 when the cause is named and the document is missing. Add which source supports it." On an essay, the one comment covers the whole piece.
Hand each paper back as a PDF or an email after you finish review for that section. The file shows the points you kept, and the comment when feedback is on. When you finish review and hand back, Graider emails the student account a "Your grade is ready" notice. Graider does not send the paper to families for you. You hand the marked paper back.
Student papers and account data aren't used to train models or for advertising.
Consistent grading across class sections means every section of the same test is given the same answer key and the same rubric, and you check the marks at review. Graider returns one score and one comment per question, and a whole essay gets one score and one comment. You review, override, toggle feedback, and hand work back as a PDF or an email.
Write the key and the rubric once, then upload that same key and that same rubric for each section. Graider compares each section's papers with that key and that rubric, and you check every mark at review. The full credit and partial credit points are the ones you wrote. You review, override any score you disagree with, and keep the pen.
Number every question the same way on the test and on the key. Write full-credit answers, accepted equivalents, and partial credit examples with their points before the first section is marked. Essays and other open responses go on the rubric. Those examples feed one score and one comment per question.
Open every Low confidence badge. It is the only flag, and it sits on a hard-to-read answer. Compare the read with the paper and your key, then override any score you disagree with. Make the same override call for the same wording in every section, and set the feedback toggle the same way on each section's test. Set the points yourself on a diagram, a graph, or a drawing. When you hand back, Graider emails the student account a "Your grade is ready" notice. Graider does not send the paper to families for you.
Check every Low confidence badge, apply the same override thinking across sections, and toggle feedback the same way. You keep the pen.