Can Reviewers Reconstruct an Employee Record Correction?

Research question: Does a four-event trace help reviewers detect incomplete or contradictory corrections more accurately than a current-value snapshot?

Methodology and scope: Construct sixty synthetic cases across directory, payroll-routing, start-date, and benefits-reference fields. Seed complete controls, missing approvals, effective-date conflicts, wrong-destination updates, and superseded requests. Randomize order and conceal the class from two reviewers.

Protocol and measures: Compare a snapshot reviewer with a trace reviewer. Both assign complete, hold, or cannot determine. Report correct holds, false clearance, review time, and agreement with the seeded class using rules locked before review.

Scope boundary: The unit is one synthetic correction version in one declared destination. The study tests reconstruction, not the truth or fairness of the underlying decision.

Inference boundaries: A lower false-clearance rate would support further testing on similar cases. It would not prove live completeness, legality, or authority. HR, payroll, benefits, identity, and system owners keep all decisions.

Limitations: Synthetic histories are smaller and cleaner than live records. The study omits restricted evidence, oral instructions, outages, fraud, and jurisdiction-specific rules. A live pilot needs privacy review and approved criteria.

References: National Institute of Standards and Technology, Cybersecurity Framework 2.0 (https://doi.org/10.6028/NIST.CSWP.29); National Institute of Standards and Technology, Privacy Framework 1.0 (https://doi.org/10.6028/NIST.CSWP.01162020); U.S. Government Accountability Office, Standards for Internal Control in the Federal Government, GAO-14-704G (https://www.gao.gov/products/gao-14-704g); Philippine Statistics Authority, Annual Survey of Philippine Business and Industry (https://psa.gov.ph/statistics/annual-survey-philippine-business-and-industry). These primary publications supply control, privacy, evidence, and sector context. None evaluates Outsourced Employment or the synthetic cases proposed here.