The methodology developed and used by SGS staff is unique. In determining calibration parameters, measurement errors of both the analyzer and the physical sampling and analysis are taken into account. This avoids the biases inherent in ordinary least squares regression, which is based on the premise of no error in measurements of the independent variable. Unbiased estimates of calibration parameters and measurement precision are calculated using a latent variable statistical model. No assumption about the status of the current analyzer calibration during the initial test period is required.

When a special test is needed for formal recalibration or precision testing of the analyzer, SGS will:

• Prepare detailed test specifications.
• Perform the necessary sampling and inspection work.
• Prepare and analyze the test samples in the laboratory.
• Do the detailed statistical analysis.  
• Provide a comprehensive report with results and calibration recommendations.