What is it about?
This publication provides an overview of the statistical methods used to evaluate and select an appropriate regression model for a given dataset. It covers the Lack of Fit (LoF), F-Snedecor, F-IUPAC and Mandel tests. This work can be used as a simple guide through which you can use the chosen tests to evaluate your own datasets, based on ready-made examples. The paper is made in such way, that there is no need for any paid software to be used. In the near future, it is planned to add a ready-to-use spreadsheet and/or an R-language program to determine the appropriate regression model.
Featured Image
Why is it important?
Nowadays the number of journals that can request complete statistics on the chosen regression curve (if applicable) is steadily increasing. This work explains these statistical tests from a very practical perspective whereas most works on this topic are more centred on the theoretical side of the problem. In this paper everything was explained using a certain dataset, that can be then used by other researchers as a test dataset to determine any problems with their calculations.
Perspectives
I hope that this work can be helpful for other researchers, who have to justify the regression model they have used in their paper. This paper was made so other people won't have to dig through literature or buy expensive programs just for the sake of a few statistical tests.
MSc Rafał Ireneusz Rawski
Politechnika Rzeszowska im Ignacego Lukasiewicza
Read the Original
This page is a summary of: Regression analysis in analytical chemistry. Determination and validation of linear and quadratic regression dependencies, January 2016, Academy of Science of South Africa,
DOI: 10.17159/0379-4350/2016/v69a20.
You can read the full text:
Contributors
The following have contributed to this page
