Uncertainty of Linear Calibration Curve & Detection Limit (Part 1)

A calibration curve or regression is always plotted with the instrument responses on the dependent vertical (y -) axis and the standard concentrations on the horizontal (x -) axis. As a linearity of calibration regression minimizes random errors of the instrument, we aim to find the best straight line to pass through these calibration points.

• What are the estimated errors and confidence limits we can estimate for the slope (or gradient) and the y-intercept of the line?

• What is the uncertainty of the analyte concentration found on the calibration curve?

• What are the least concentration of the measurand can be detected with, say 95% confidence?

• In other words, what is the limit of detection (LOD) of this method?

This one-day training program is designed to answer all the above questions and others related to the subjects of linear regression of instrument calibration and detection limits including IDL, MDL and LOQ.

Training Program (8:00 - 10:30 CEST)

• Introduction
• Different meanings for correlation and regression
• Assumptions of linear regression and correlation
• Residual analysis – what and how
• Uncertainty estimation of linear calibration regression
• Must your linear regression always pass through its origin (0,0)?
• The most controversial subject in analytical chemistry – Limit of Detection LOD
• Different definitions for LOD and its basic concepts
• Differences between IDL, MDL, LOQ
• Determining LOD by linear calibration model

Who should attend:

This course is designed for those who want to know more about the sampling procedures and how it affects the validity of their test results, including:

• Laboratory analysts
• Laboratory QA/QC personnel
• Lab directors / managers
• Personnel who carry out tasks such as method validation and uncertainty estimation

Pre-requisites

The participants are assumed to have had some basic statistical knowledge such as the calculation of mean, standard deviation, coefficient of variation, relative standard deviation, etc.

Participants are to have a laptop with Microsoft Excel installed and have some very basic knowledge of how to use the program.

Presenter: Yeoh Guan Huah

Mr Yeoh Guan Huah has B.Sc. (Hons) and M.Sc. degrees in Chemistry and was former Managing Director of ALS Technichem Singapore (1978-2006), former East Asia Regional Manager of ALS Group of Laboratories in Hong Kong and China (2006-2016). He is now the senior partner of GLP Consulting, Singapore.

Currently Mr Yeoh is an appointed External Lead Assessor and Technical Assessor of SAC-SINGLAS till 31 March 2025 and had represented DSM/SIRIM Malaysia as a active member of ISO Technical Committee TC 69 (Application of Statistical Methods) and Convenor of AHG1 from 2005-2016.