Evaluation of the Linearity of Quantitative Measurement Procedures: A Statistical Approach; Approved Guideline
|Publication Date:||1 April 2003|
This document presents a method to establish, verify and/or demonstrate the linear range of a quantitative measurement procedure. These methods do not identify the cases of significant nonlinearity. The method employs increasing numbers of samples for more definitive examinations of linearity. Therefore, if a failed demonstration is evaluated and it is determined that the experiment needs to be repeated, it can be done with more replicates or with fewer levels to cover a smaller range.
• This protocol is to assess linearity, isolated as much as possible from conditions of precision and trueness. It is understood that poor precision will hinder an effective assessment of linearity, so a check for poor repeatability is included.
• These experiments should use samples with a matrix appropriate to the specimens being analyzed (serum, plasma, urine, etc.)
• This protocol requires laboratories to set goals for nonlinear error. It provides basic concepts for setting such goals, but does not recommend any specific protocol.