Assessment of goodness-of-fit for the main analytical calibration models: Guidelines and case studies
TRAC-TRENDS IN ANALYTICAL CHEMISTRY(2021)
摘要
This critical review paper will discuss the main analytical calibration models as well as the guidelines for their practical use. The main models used to fit a multiple-point calibration dataset are: 1) linear unweighted or ordinary least squares regression (OLSR); 2) quadratic unweighted least squares regression (QLSR); 3) linear weighted least squares regression (WLSR). Unfortunately, there is no standard procedure in analytical chemistry for objectively testing the goodness-of-fit of calibration models. Different proposals were reported in the literature. However, none is more commonly used, and probably not more controversial than R-2. In this document, a three step simple calibration diagnosis has been proposed. It is based on a combination of different procedures such as graphical plots, statistical significance tests and numerical parameters. Experimental conditions and design of calibration procedures are very relevant for appropriate selection. Finally, some information on the choice of the different models will be reported in four case studies. (C) 2021 Elsevier B.V. All rights reserved.
更多查看译文
关键词
Back-calculated concentration, Calibration Determination coefficient, Goodness-of-fit, Least squares regression, Linear regression, Quadratic regression, Relative error, Weighted regression
AI 理解论文
溯源树
样例
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要