Comparison of the Target Values Established Using a Robust Statistical Method and Roche's Calibration Values
Related Experiment Video
Updated: Mar 7, 2026

Split Point Analysis and Uncertainty Quantification of Thermal-Optical Organic/Elemental Carbon Measurements
Published on: September 7, 2019
Background:
In EQA schemes, results are usually assessed against a target value.
Methods:
Outliers in each result were deleted using a robust statistical method. Mean was used as a target value if trimmed results were normally distributed; otherwise, median was used. Differences between target value and Roche's calibration value were calculated. Rates of acceptable performance evaluated by target value and Roche's calibration value were determined.
Results:
Target values of P, GLU, urea, UA, TP, ALB, TC, TG, ALT, AST, ALP, AMY, LDH, GGT, FE, and HBDH were 1.70 mmol/L, 11.0 mmol/L, 16.8 mmol/L, 307 μmol/L, 54.0 g/L, 38.1 g/L, 4.19 mmol/L, 1.46 mmol/L, 97.7 U/L, 98.2 U/L, 206 U/L, 180 U/L, 242 U/L, 115 U/L, 35.6 μmol/L, and 255 U/L, respectively. Differences between target values and Roche's calibration values were -3.45 to 1.48%. Rates of acceptable performance evaluated by target values and Roche's calibration values were 96.5 - 100.0% and 93.0 - 100.0%, respectively.
Conclusions:
Target values established using a robust statistical method and Roche's calibration values were comparable.
More Related Videos
Related Concept Videos
Calibration Curves: Correlation Coefficient
Instrument Calibration
Analytical Balance Calibration
An analytical balance measures mass and requires regular calibration to...
Data Validation
Key parameters for method validation include:
Calibration Curves: Linear Least Squares
For data that follow a straight line, the standard method for fitting is the linear...
Bioequivalence Data: Statistical Interpretation
Standard Deviation of Calculated Results
A broad Gaussian distribution curve has a wider standard deviation, representing a data set with...

