Statistics in ASTM Standard Test Method Development, Application, and Quality Assurance

Meet the Instructor

About the Course

This course provides detailed methodology on how to set up Statistical Control Charts relevant to test method performance monitoring over time as per ASTM D6299. The course begins with an overview of the importance of data trustworthiness as a key business driver, followed by understanding common cause variation between test results using Statistical Thinking. Precision fundamentals, Industry standard test method precision performance metrics such as repeatability, reproducibility, site precision, and associated business applications will be presented in detail as the first topic, to serve as a foundation upon which the Statistical Control Charts methodology are based on. Bias fundamentals, along with use of Statistical Control Charts and other statistical techniques for assessing and monitoring laboratory bias over time versus industry, will be covered in detail, followed by an overview of relative bias between test methods as the final topic. The course will be conducted in a virtual setting, with web-based demonstrations for the set up and maintenance of statistical control charts, using example data from a variety of quality control testing results. Special techniques dealing with quality control material transition, accrual of control chart statistics over time, non-normal data, and inadequate measurement resolution will be covered.

Learning Outcomes

By the end of this course you should be able to:

• Articulate the core principles and concepts associated with statistical 'thinking'
• Explain the concepts and differences for test method repeatability, reproducibility, site precision
• Explain the concepts and differences between absolute and relative bias for test methods
• Assess single lab bias relative to CRM and ILCP Performance
• Apply D6708 principles to assess relative bias between two test methods
• Apply ASTM repeatability, reproducibility, site precision, and bias to make product quality compliance decisions
• List the implicit assumptions when using these quality metrics to judge the trustworthiness of test results
• Articulate the concepts and differences between false alarms and missed signals
• Quantify the statistical confidence and risks associated with the decisions based on measurement data
• Explain the primary purpose of a control chart
• Articulate the basic premise, statistical bases and limitations of a control chart
• Outline the steps associated with setting up and regular use of control charts
• Construct I / MR / EWMA control charts as per ASTM D6299
• Set decision limits based on concepts of false alarms and missed signals
• Examine control charts and decide on the in-control status of a test method
• Statistically combine the cumulated control chart data to obtain long term test method performance baselines
• Design a quality control program for test method performance surveillance for precision and bias monitoring
• Avoid the common errors and misconceptions associated with statistical control chart set up & operation

Who Should Attend

• Lab technicians
• Lab QA/QC personnel
• Lab managers
• Personnel involved with product quality assurance and regulatory compliance testing
• Process Analyzer and Control engineers involved with production planning and optimization

Virtual Course Description

Class: 10:00 a.m.- 3:000 p.m. ET on all 5 days

Day 1

• Understanding Variation in Measurements
• Statistical Thinking Principles and Concepts
• Precision Fundamentals
• Overview of the Normal distribution and standard deviation concepts
• Statistical bases and interpretation of test method repeatability (r), reproducibility (R), and site precision (R')
• How ASTM test method r & R are estimated using ILS
• Simple applications of r, R, R'

Day 2-3

In Statistical Control & Control Chart Fundamentals

• Data quality & site precision (R') basics
• The primary purpose of a control chart
• The basic premises & statistical bases of a control chart
• Concepts of false alarm versus missed signal, and power of detection
• Limitations of a control chart
• Set up and regular use of control charts
  • Demonstration on control chart set up
  • Demonstration on control chart maintenance: what to do with the cumulated control chart data
  • Group participation on control chart interpretation
• QC material transition from one batch to another
• How to handle non-normal data

Day 4-5

Bias Fundamentals

• Various forms of Bias
• Site Expected Value (SEV) & Accepted Reference Value (ARV)
• Bias assessment for a laboratory
• How to interpret ASTM ILCP outputs
• Relative Bias between two different test methods that claim to measure the same property
• Principles of D6708: regression with errors X and Y; lack of fit, insufficient variation, residual analysis
• Basic bias-correction schemes in ASTM D6708 & information required to perform D6708 assessment
• Application Examples of D6708
• Understanding Sample-specific Bias

Finale wrap up

In-Person Course Description

Class: 8:30 a.m.- 4:30 p.m. ET on all 3 days

Since trustworthy measurement data is a fundamental enabler to all aspects of business improvement, having a fundamental understanding of testing precision, bias (absolute and relative), data quality, how to design and execute control chart work processes to ensure data quality, and the implicit uncertainties associated with measurement data can significantly enhance business process improvement initiatives. Understanding test method performance behavior from control charts can lead to correct decisions as to when to adjust and when not to adjust test methods or process control strategies. Use of control charts to monitor testing performance and apply pre-emptive just-in-time failure-prevention actions can improve laboratory and field operations efficiency, reduce waste, and assure the test data produced are fit-for-use and defensible. Understanding of test method precision and bias enables release of product with confidence at optimal production cost.

Day 1

• Understanding Variation in Measurements
• Statistical Thinking Principles and Concepts
• Precision Fundamentals
  • Overview of the Normal distribution and standard deviation concepts
  • Statistical bases and interpretation of test method repeatability (r), reproducibility (R), and site precision (R')
  • How ASTM test method r & R are estimated using ILS
  • Simple applications of r, R, R'

Day 2

In Statistical Control & Control Chart Fundamentals

• Data quality & site precision (R') basics
• The primary purpose of a control chart
• The basic premises & statistical bases of a control chart
• Concepts of false alarm versus missed signal, and power of detection
• Limitations of a control chart
• Set up and regular use of control charts
  • Demonstration on control chart set up
  • Demonstration on control chart maintenance: what to do with the cumulated control chart data
  • Group participation on control chart interpretation
• QC material transition from one batch to another
• How to handle non-normal data
• General Q & A Wrap Up for Day 1and 2

Day 3

Bias Fundamentals

• Various forms of Bias
• Site Expected Value (SEV) & Accepted Reference Value (ARV)
• Bias assessment for a laboratory
• How to interpret ASTM ILCP outputs
• Relative Bias between two different test methods that claim to measure the same property
• Principles of D6708: regression with errors X and Y; lack of fit, insufficient variation, residual analysis
• Basic bias-correction schemes in ASTM D6708 & information required to perform D6708 assessment
• Understanding Sample-specific Bias

Finale wrap up

Referenced Documents

• D3244 Standard Practice for Utilization of Test Data to Determine Conformance with Specifications
• D6299 Standard Practice for Applying Statistical Quality Assurance and Control Charting Techniques to Evaluate Analytical Measurement System Performance
• D6300 Standard Practice for Determination of Precision and Bias Data for Use in Test Methods for Petroleum Products, Liquid Fuels, and Lubricants
• D6708 Standard Practice for Statistical Assessment and Improvement of Expected Agreement Between Two Test Methods that Purport to Measure the Same Property of a Material
• D6792 Standard Practice for Quality Management Systems in Petroleum Products, Liquid Fuels, and Lubricants Testing Laboratories
• D7372 Standard Guide for Analysis and Interpretation of Proficiency Test Program Results
• D7915 Standard Practice for Application of Generalized Extreme Studentized Deviate (GESD) Technique to Simultaneously Identify Multiple Outliers in a Data Set
• D8146 Standard Guide for Evaluating Test Method Capability and Fitness for Use

Fee Includes

• Referenced documents (listed above)
• Course notes
• Certificate of Attendance
• 2.1 CEUs (Continuing Education Units)

This course also includes access to two free ASTM course:

• ASTM 101: Introduction to Standards
• ASTM 102: Selecting and Utilizing a Standard

About the Instructor

Alex T. Lau is the Chair of ASTM D02.94 Subcommittee on Quality Assurance and Statistics, D02.01.0B on Precision, and Vice Chair of Subcommittee D02.25 on Validation of Process Stream Analyzer Systems. He has 40 years experience in the Petroleum Refining Industry. An ASTM International member since 1990, he is also an active member of Committees E11 on Quality and Statistics. He retired from Imperial Oil, Canada (an ExxonMobil affiliate), and formed his own consulting company (TCL-Consulting). His career focus specialized in gasoline and diesel fuel blending, direct blend to pipeline and ships, on-line process analyzer applications, statistical techniques for quality assurance and process improvement, development and implementation of industry standards.

He is the recipient of the ASTM International Award of Merit, ASTM D02 Scroll of Achievement, and several Awards of Excellence. Lau was cited "for outstanding contributions in Committee D02 on Petroleum Products and Lubricants and Coordinating Subcommittee D02.94 on Quality Assurance and Statistics toward standards development, both as an individual contributor and leader."

He graduated from the University of Toronto with a Bachelor's degree in Applied Science and Engineering Physics. Outside of ASTM International, Lau is a member of the Professional Engineers of Ontario (PEO) and the American Society for Quality (ASQ). He is a registered Professional Engineer in Ontario, Canada, and an ASQ Certified Quality Engineer. He is also the Convener of ISO TC28 / WG2 on Statistics.

About Sponsoring Committee

Organized in 1898, ASTM is one of the world's largest voluntary standards development organizations. ASTM standards have grown to be among the world's most widely used and accepted documents. The 82-volume Annual Book of ASTM Standards contain 11,000 standards written by 34,000 members on our 140 technical committees. Committee D02 on Petroleum Products, Liquid Fuels, and Lubricants developed the standards used in this course. For more information, contact Alyson Fick at (610) 832-9710 or go to our D02 technical committee page.

Questions About This Training Course

For information on this training course, please contact Training

Attention: Professional Engineers

Professional Engineers, if your state has a continuing education requirement for license renewal, ASTM training courses should count towards your Continuing Professional Competency (CPC) Professional Development Hour (PDH) requirement.

Training Available

ASTM can provide training customized to your organization. For more information, please contact sales here or call 1-877-909-ASTM.

How Learning will be assessed

Learning will be assessed through a series of question and answer sessions.

Stock Code