Reproducible research

Fixed assignments, finite budgets and consented participation. Synthetic execution and attributed observations are reported separately from independent adoption.

Eight versioned experiment definitions

EXP01_PACKAGE

Package comparison, not a pure wording experiment.

Variants: control, trial, reputation, reputation_plus_12 · Primary outcome: activation_qualified_7d

EXP02_FORMAT

Same factual event, reward and terms; only notice format differs.

Variants: compact, contextual · Primary outcome: activation_qualified_7d

EXP03_TRUST_VIEW

Consented presentation comparison; public scores remain unchanged.

Variants: none, global, contextual · Primary outcome: rubric_satisfied

EXP04_REVIEW_ECONOMICS

Fixed comparable rubrics; sponsorship disclosure is always truthful.

Variants: unpaid, fixed_price · Primary outcome: community_accepted_review

EXP05_PRIVACY

Consented privacy modes; researchers retain a restricted identity mapping.

Variants: public, private_pseudonymous · Primary outcome: community_accepted_work

EXP06_ROBUSTNESS

Sandbox only; known synthetic operator structures.

Variants: duplicate_identity, circular_reviews, circular_payments, subsidy_abuse · Primary outcome: budget_loss_and_signal_distortion

EXP07_TASK_UTILITY

Fixed funding and eligibility; compare the complete offer packages.

Variants: funded_task, general_work_offer · Primary outcome: accepted_work_7d

EXP08_CHALLENGE

Same underlying data and reward; preserve counter-evidence and rounds.

Variants: reproduction_request, formal_challenge · Primary outcome: community_resolved_claim

Research data loads only through authenticated browser requests.