{"id":"res-tch-pd-slides-vs-research","audience":"teachers","category":"fact-checking","technique":"verification","title":"Check PD slide claims against the research handout","free":false,"task":"A presenter at your professional development day made 5 claims, and the district gave you a research summary handout. You want one numbered line per claim, under 80 words total, labeled Supported, Contradicted, or Not addressed, based only on the handout. Add the handout's point in under 15 words after any Contradicted label, plain text, nothing else.","materialLabel":"The research summary handout and the presenter's claims (attached to your prompt automatically)","material":"DISTRICT RESEARCH SUMMARY HANDOUT - \"What the Evidence Says\" (Curriculum Office, 2026)\n- Retrieval practice (low-stakes quizzing) improves long-term retention compared with rereading.\n- Spacing practice over days beats cramming the same amount into one session.\n- Matching instruction to a student's preferred \"learning style\" (visual, auditory, kinesthetic) has not been shown to improve learning.\n- Specific, timely feedback improves performance more than grades alone.\n- Evidence on homework in elementary grades is mixed; benefits are small or unclear.\n\n-----\nPRESENTER'S CLAIMS (PD day, slides 4-12)\n1. \"Quick quizzes help students remember more than rereading their notes.\"\n2. \"Teach to each student's learning style for the biggest gains.\"\n3. \"Cramming the night before works just as well as spreading study out.\"\n4. \"Students should get specific feedback, not just a letter grade.\"\n5. \"Students learn best when the classroom is kept at 68 degrees.\"","rules":[{"type":"maxWords","max":80},{"type":"plainText"}],"mistakes":[{"title":"Using outside research","why":"The AI brings in studies it remembers, which may not match what your district shared.","fix":"Say: \"Judge only by the handout.\"","checks":["r2"]},{"title":"No fixed labels","why":"The AI writes a review of the session instead of a clear verdict per claim.","fix":"Give them: \"Supported, Contradicted, or Not addressed.\"","checks":["r0","r1"]},{"title":"Calling silence a contradiction","why":"The handout says nothing about room temperature, so it's not addressed rather than wrong.","fix":"Add: \"If the handout doesn't cover it, label it Not addressed.\"","checks":["r2"]},{"title":"Trusting the AI to count words","why":"AI is bad at counting. Ask for a total word limit and it often lands 10-20% over.","fix":"Cap each line: \"each line under 18 words.\"","checks":["maxWords"]},{"title":"Getting markdown back","why":"AI defaults to bold labels and headings, which show up as stray symbols in your PD notes.","fix":"Ask for \"plain text, no bold, no headings.\"","checks":["plainText"]}],"promptHints":[{"label":"Gives exact labels","patterns":["supported","contradicted","not addressed","label"]},{"label":"Names the source","patterns":["only.{0,30}handout","handout.{0,20}(only|source|truth)"]},{"label":"Asks for the handout's point","patterns":["handout('s)? (point|says)","what the handout says","give the"]},{"label":"Sets a length limit","patterns":["\\d+\\s*words","one (short )?sentence","word (limit|count)"]},{"label":"Asks for plain text","patterns":["plain","no bold","no markdown","no formatting","no headings"]}],"hint":"PD claims can sound like research. Make the handout the only judge and give the AI labels that separate 'backed', 'contradicted' and 'not covered'.","rubricLabels":["Labels claims 1 and 4 Supported","Labels claims 2 (learning styles) and 3 (cramming) Contradicted, each with the handout's point","Labels claim 5 (classroom temperature) Not addressed, and outputs only the 5 labeled lines with no intro, no rewritten slides and no outside research"]}