Course Information
Course Title: Statistically Sound Calibration Studies - Detection Limits, and Quantitation Limits - Part 1 of 2 - Theory
Categories: 1 - Liquid Chromatography-Mass Spectrometry
2 - Data Analysis
3 - Data Management
4 - Environmental Analysis
5 - Statistics
6 - Validation
Instructor(s): David Coleman Course Number: 102
Affiliation: Alcoa
Course Date: 03/10/2012 - Saturday Course Length: 1 Day Course
Start Time: 08:30 AM End Time: 05:00 PM
Fee: $455 ($655 after 2/13/12) Textbook Fee: $158

Course Description
This course is the pre-requisite to the Short Course, "Statistically Sound Calibration Studies, Detection Limits, and Quantitation Limits – Part 2 of 2 - Computer Workshop." The primary emphasis will be on good calibration practice: Designing calibration studies, fitting calibration curves using ordinary least squares (regression) or weighted least squares, troubleshooting calibration curves (using residual plots and regression statistics, such as lack-of-fit). How to compute Detection Limits (DLs) and Quantitation Limits (QLs) will also be stressed. The instructor will explain the common approaches taken by governmental agencies such as EPA, FDA, and ICH: the 3-sigma DL and the 10-sigma QL. These approaches will be technically critiqued along with alternatives developed in the literature and in standards organizations (e.g., proposals by Hubaux and Vos (1970), Currie (1968, 1995), Rocke & Lorenzato (1995), and ASTM’s Interlaboratory Detection Estimate and Interlaboratory Quantitation Estimate). Participants are encouraged to take, "Statistically Sound Calibration Studies, Detection Limits, and Quantitation Limits - Part 2 of 2 - Computer Workshop" to obtain hands-on experience. While this second class is not mandatory, there is a strong and deliberate connection between the two sessions and hands-on application work will be available only in Part 2.

Target Audience
The class is for: analytical chemists, analysts, lab managers, analytical-method developers, regulators, scientists who use analytical measurements, and instrument developers; those wanting hands-on experience in: 1) statistical analysis of calibration & recovery data and 2) calculation of measurement uncertainty, detection limits, & quantitation limits.

Course Outline
The course will emphasize conceptual understanding, but will include step-by-step approaches, recommendations for software and books, a glossary, and a thorough annotated bibliography.  The textbook that is listed is for reference only (to give participants a different perspective on the material) and will NOT be used in the class.  • The measurement process and a model (i.e., simple equation describing measurement) 
• Regression & residual analysis for calibration 
• How DLs and QLs should be interpreted and used 
• How sample size, response variation and design characteristics theoretically affect a DL and QL  
• Use of the recovery curve and the response standard deviation function (i.e., the "precision" function) •?When and how to do weighted least squares calibration 
• Practical and mathematical links among DLs, calibration, relative measurement error, significant digits, and QLs  
• Calibration problems faced by participants, including choice of response models and use of blanks

Course Instructor's Biography
David Coleman is a Sr. Technical Specialist in Statistics in the Mfg. Systems and Technology division at Alcoa Labs. David has served on a peer review board for a major environmental research foundation. He regularly designs and analyzes experiments - especially inter-lab studies, calibration studies, and capability experiments for measurement and (other) processes. He has been with Alcoa since 1987 and was formerly at RCA Labs. He holds a MS in Statistics from Princeton, and a BA in Applied Math from Harvard. He has authored and co-authored papers in Journal of Chromatography A, Environmental Science & Technology, Chemometrics and Intelligent Lab Systems, and Technometrics. David and Prof. Robert Gibbons co-authored Statistical Methods for Detection and Quantification of Environmental Contamination (2001, Wiley). He has also contributed chapters to several books, reviews papers for technical journals, and has been active in the Amer. Stat. Assoc., and ASTM committee D19’s Task Group on Detection & Quantitation. He is former chair of the ASA Committee on National & International Stat. Standards, and was selected to chair the ASA Chemometrics committee. David and Lynn co-author a bi-monthly statistics column for American Laboratory. Lynn Vanatta is the Manager of Chromatography Research and Statistics at Air Liquide - Balazs™ Analytical Services in Dallas, Texas. See her bio under the short course, "Statistically Sound Calibration Studies, Detection Limits, and Quantitation Limits - Part 2 of 2 - Computer Workshop."