Add published survey rows, written offers or safely redacted observations you can identify. The notebook rejects mixed evidence, ignores rows outside your freshness horizon and calculates a median and quartiles only after five comparable rows.
Lire en français.
Monthly rows are multiplied by 12. Annual rows are unchanged. Values are sorted, then Q1, median and Q3 use the NIST p(N+1) rank with linear interpolation and endpoint bounds. No outlier is removed and no weighting is invented.
NIST documents the percentile method and notes that interpolation conventions differ. ILOSTAT explains gross earnings concepts. Neither source supplies salary values here.
Rows remain in memory until refresh, clear or tab close. Nothing is sent to an API, AI, analytics or storage. The result describes only your sample; it does not establish a representative market rate, minimum wage, fair-pay conclusion, tax result or recommended offer.