Türkiye's PISA Rise via Placebo Testing: Can Innovative Domains Audit Score Inflation?

PISA's innovative domain changes in every cycle. If preparation for the core transfers less to it than genuine learning does, it can audit the core. I measure the gap between each country's innovative-domain mean and the value its core mean predicts. A placebo-domain comparison is uninterpretable unless the innovative domain and the core are administered in the same mode. Türkiye falls in the lower tail in all three mode-matched cycles (2003, 2015 and 2025). The 2003 result predates the 2012–2015 institutional break by twelve years. From 2015 to 2025 the gap widened by about 20 points, among the largest in the OECD. This is ordinary given Türkiye's core gain, but that comparison has little power. The gap signals an anomaly but does not measure inflation. Because each cycle tests a different construct, the audit cannot separate post-break inflation from a weakness specific to the 2025 domain. Package contents: preprint: Full-text manuscript (PDF, DOCX, Markdown). code: Analysis and replication scripts data_derived: Derived country estimates, regression tables, and simulation matrices. FILE_LIST.md: Full manifest with file sizes and SHA-256 checksums. Note: Raw PISA microdata are not redistributed; they are publicly available from the OECD