A statistical overview on univariate calibration, inverse regression, and detection limits: Application to gas chromatography/mass spectrometry technique.

The paper summarizes critically the current approaches for the calculation of the limits of detection and quantification. In the context of the description of the method based on the calibration line, the arguments concerning the underlying experimental design, the choice of the appropriate model in the univariate regression, the effects of the dispersion characteristics of the data are deeply discussed. The effects of the scedasticity of the experimental data are taken into account in the obtainment of the calibration curve and in its utilization. To gain transparency, adaptability, and tutorial effectiveness the explicit formulas relevant to the use of straight line and quadratic models are reported. An application of the described procedures to GC-MS data is reported as an illustrative example. Copyright 2006 Wiley Periodicals, Inc.