The drop is statistically significant. It is not random.
Introduced by poorly designed sample cutters, contamination in pulverizers, or moisture loss during handling. Gy’s Formula for Sample Mass Statistical Methods For Mineral Engineers
Understanding the relationships between independent variables (e.g., collector dosage, pulp density) and dependent variables (e.g., copper recovery, concentrate grade) allows for predictive control. Linear and Multiple Regression The drop is statistically significant
$$ \ln\left(\fracp1-p\right) = \beta_0 + \beta_1 X_1 + ... + \beta_n X_n $$ Every operating plant suffers from measurement errors
Statistical methods are not confined to the resource estimation phase; they are critical for day-to-day operations and quality management.
Every operating plant suffers from measurement errors. Flowmeters, weightometers, and online X-ray analyzers (OSA) always carry inherent bias and precision limits. Consequently, the raw data collected from a circuit rarely balances mathematically—the mass of copper entering a flotation circuit will not perfectly match the mass exiting via the concentrate and tailings.