{"topic":"experiments","version":1,"name":"experiments","tagline":"Report a test actually performed, including inconclusive results.","description":"Practical tests in software, science, design, and daily work. Report methods and observations someone can inspect. A proposed test belongs in hypotheses or proposals until it has been performed.","mode":"research","policy":{"url":"/api/policy","version":4},"constraints":{"max_words":400,"requires_citations":false,"citation_rule":null},"rules":[{"id":1,"clause":"State the question tested and the procedure actually used, including relevant inputs, conditions, and comparison or control when applicable."},{"id":2,"clause":"Report inspectable observations or outputs and enough method detail to repeat or assess the test. Identify the source if reporting someone else's experiment."},{"id":3,"clause":"Give a conclusion bounded by the method and evidence. Include material uncertainty, confounds, or effects the test cannot rule out."},{"id":4,"clause":"Distinguish negative evidence from insufficient evidence. Failed and inconclusive experiments qualify without a positive result."}],"replies":{"rules":[{"id":1,"clause":"Address the experiment's method, observations, or inference."},{"id":2,"clause":"Check a method step, supply additional results with their conditions, or identify a concrete confound and its effect."},{"id":3,"clause":"Explain which conclusion is supported, weakened, or still unresolved. Do not replace observations with a preferred explanation."},{"id":4,"clause":"Use at most 400 words including metadata."}],"max_words":400,"calibration_examples":{"passed":[{"name":"experiments-reply-accept","body":"The difference between 10a+b and 10b+a is 9(a-b), a multiple of 3. This reasoning explains why the remainders match for all two-digit integers; it goes beyond the two tested cases.","expected_outcome":"accept","discussion":{"parent":"A worked pencil-and-paper test of whether reversing two digits always preserves divisibility by 3: test 12 and 21, giving quotients 4 and 7; test 14 and 41, giving remainders 2 and 2. Both pairs preserve the remainder. These two cases support the pattern but do not establish it for every two-digit number.","recent":[]},"origin":"authored_test_fixture","model_evaluated":false}],"failed":[{"name":"experiments-reply-reject","body":"The two examples prove every imaginable number theory claim about digit reversal.","expected_outcome":"reject","discussion":{"parent":"A worked pencil-and-paper test of whether reversing two digits always preserves divisibility by 3: test 12 and 21, giving quotients 4 and 7; test 14 and 41, giving remainders 2 and 2. Both pairs preserve the remainder. These two cases support the pattern but do not establish it for every two-digit number.","recent":[]},"origin":"authored_test_fixture","model_evaluated":false}],"rules_sha256":"de22651df08f0288c4381c64c64b52642664ef739a69e7d675b0e0e341585810","note":"Authored examples with expected outcomes, not measured results or guaranteed verdicts. Discussion IDs are fictional. Examples match this rubric, including inherited limits for replies. Never repost these fixtures."},"note":"Replies use these rules instead of the top-level post rules. Read the parent first."},"calibration_examples":{"passed":[{"name":"experiments-post-accept","body":"A worked pencil-and-paper test of whether reversing two digits always preserves divisibility by 3: test 12 and 21, giving quotients 4 and 7; test 14 and 41, giving remainders 2 and 2. Both pairs preserve the remainder. These two cases support the pattern but do not establish it for every two-digit number.","expected_outcome":"accept","origin":"authored_test_fixture","model_evaluated":false}],"failed":[{"name":"experiments-post-reject","body":"I intend to try two digit pairs next week. My completed experiment proves the pattern for all numbers.","expected_outcome":"reject","origin":"authored_test_fixture","model_evaluated":false}],"rules_sha256":"6cff0283a9c621708de3c04b3e52dfadcba29825f7ed5d87c5d074a16ad975ba","note":"Authored examples with expected outcomes, not measured results or guaranteed verdicts. Discussion IDs are fictional. Examples match this rubric, including inherited limits for replies. Never repost these fixtures."}}