3. What regression strategies are available?
In the literature, there are numerous strategies for estimating the parameters of a chosen model, each with its own advantages and disadvantages. The most commonly used technique is "least squares regression". This technique comes in a variety of more or less elaborate forms.
3.1 Ordinary least squares (OLS)
This method deals only with observed values of x and y, without taking into account the associated uncertainties. To be usable, this technique requires a number of conditions to be met:
no uncertainty in X (which is never the case in calibration, but can sometimes be assumed to be negligible);
identical uncertainty on y (which is rarely the case in calibration, as uncertainty...
The Ultimate Scientific and Technical Reference
This article is included in
Laboratory quality and safety procedures
This offer includes:
Knowledge Base
Updated and enriched with articles validated by our scientific committees
Services
A set of exclusive tools to complement the resources
Practical Path
Operational and didactic, to guarantee the acquisition of transversal skills
Doc & Quiz
Interactive articles with quizzes, for constructive reading
What regression strategies are available?
Which approach to choose?
Bibliography
P.H. Cornillon, E.M. Lober, Regression – Theory and application , Ed Springer
Collège Français de Métrologie (CFM): "Application of the new VIM 3 calibration concept" and the associated M-CARE processing software, downloadable from the CFM website
ToolNormative reference
Standard NF T 90-210 "Qualité de l'eau – Protocol for the initial evaluation of the performance of a method in a laboratory".
The Ultimate Scientific and Technical Reference
