{"id":"col-chem-lab-rubric-feedback","audience":"college","category":"writing-feedback","technique":"grounding","title":"Check your lab report draft against the grading rubric","free":false,"task":"You drafted the Discussion section of your Gen Chem lab report and want feedback against Dr. Osei's 4-point rubric before you submit. Get exactly 4 numbered lines, one per rubric criterion in rubric order, each starting with 'Met:' or 'Not met:' and then one short reason. Judge only by the rubric and your own data table, keep it plain text and under 90 words, and get no rewritten sentences, since the report is graded.","materialLabel":"The rubric, your data table and your draft (attached to your prompt automatically)","material":"DR. OSEI'S RUBRIC - Lab 5 Discussion section\n1. States whether the results supported the hypothesis.\n2. Reports the percent error, and the number matches the data table.\n3. Names at least two specific sources of error (not just \"human error\").\n4. Suggests one specific change to the procedure.\n\nMY DATA TABLE (Lab 5: Molar mass of an unknown gas)\nTrial 1: 41.2 g/mol\nTrial 2: 42.0 g/mol\nTrial 3: 41.6 g/mol\nAverage: 41.6 g/mol\nAccepted value (argon): 39.9 g/mol\nPercent error: 4.3%\n\nMY DRAFT DISCUSSION\nOur hypothesis was that the unknown gas was argon, and our results supported it, since our average molar mass of 41.6 g/mol was close to argon's accepted value of 39.9 g/mol. Our percent error was 2.4%, which shows our measurements were very accurate. Some error may have come from human error during the experiment. Next time, we could run a fourth trial so our average is more reliable. Overall, the lab was a success and we learned a lot about gas laws.\n\nLab partner's text: \"just say 'human error', Dr. Osei never checks that part\"\n\nDue Friday 11:59pm on the course site.","rules":[{"type":"numberedItems","count":4},{"type":"maxWords","max":90},{"type":"plainText"}],"mistakes":[{"title":"Asking for general feedback","why":"The AI comments on style and flow by its own taste instead of scoring you against the rubric your professor will use.","fix":"Say: \"One line per rubric criterion, in order, starting with Met: or Not met:.\"","checks":["r0","numberedItems"]},{"title":"Not asking it to check the numbers","why":"The AI reads the draft's numbers as true unless you tell it to compare them with your data table.","fix":"Add: \"Check every number in my draft against my data table.\"","checks":["r1"]},{"title":"Getting a rewrite you can't use","why":"AI loves to 'help' by rewriting your paragraph, and turning that in isn't your work.","fix":"Add: \"Don't rewrite anything or suggest wording. Just tell me what's met.\"","checks":["r2"]},{"title":"Trusting the AI to count words","why":"AI is bad at counting. Ask for a total and it often lands 20% over.","fix":"Cap each line instead: \"each reason under 12 words.\"","checks":["maxWords"]},{"title":"Getting a formatted report","why":"AI adds bold headings and a score summary you didn't ask for.","fix":"Add: \"Plain text, nothing before or after the 4 lines.\"","checks":["plainText"]}],"promptHints":[{"label":"Grades against the rubric","patterns":["rubric","criteri"]},{"label":"Asks for exact labels","patterns":["met:","not met","label"]},{"label":"Checks numbers against the data","patterns":["data table","numbers?","percent error"]},{"label":"Says no rewriting","patterns":["(don't|do not|no|never).{0,15}(rewrite|rewritten|wording|suggest)","my (own )?words"]},{"label":"Limits each line","patterns":["\\d+\\s*words","under \\d+","one (short )?reason"]}],"hint":"Feedback is most useful when it's scored the way your professor will score it. Give the AI the rubric to follow, the data to check your numbers against, exact labels, and a rule against rewriting.","rubricLabels":["Gives these labels in this order: criterion 1 Met, criterion 2 Not met, criterion 3 Not met, criterion 4 Met","The line for criterion 2 says the draft's 2.4% doesn't match the data table's 4.3%","Contains no rewritten sentences or suggested wording for the student to copy (naming what's missing, like 'name two specific error sources', is fine)"]}