I have several CSV datasets that need a careful sweep so every value left in the file genuinely reflects reality. My only objective is to boost overall data accuracy, and I want to follow a clear-cut, statistical approach: any record sitting outside a set number of standard deviations from the mean should be flagged and deleted outright—no replacement, no imputation. You’re free to work in Excel, Google Sheets, Python (pandas / NumPy), R, or any tool you trust, so long as the final files return in the same structure and encoding they arrived in. A concise log of how many rows were dropped per file will help me double-check the results. Deliverables • Cleaned CSV files, identical column order and headers • A brief summary (CSV or TXT) listing for each source file: total rows before, rows removed, and rows remaining I will supply the raw data immediately and can answer edge-case questions quickly so the process stays smooth.