Normality Testing and Transformations (1/2-day)
Instructor
Course Objective
This half day course is designed to help you handle variables sampling plans. Variables sampling plans assume the data fits the normal distribution and are relatively sensitive to departures from normality. This course teaches how to test for normality and how to handle situations when the data is not normal including fitting other distributions and transforming the data. This course is task orientated in that it provides all the knowledge and tools required to handle variables sampling plans. No prior knowledge or special background is required. It covers STAT-18, Statistical Techniques for Normality Testing and Transformations, of the book Statistical Procedures for the Medical Device Industry. It utilizes the Minitab or Distribution Analyzer software packages.
What You Will Learn
- Answer the question: “Why is normality important?”
- How to test for normality.
- How to control false rejections due to large data sets and poor measurement resolution
- How to handle departures from normality including:
- Nonnormal distributions (fitting other distribution and transformations)
- Outliers
- Subgroups
- Shifts
- Truncated data
- Step-by-Step procedure for executing variable sampling plans
Primary Audience
This course is designed for those individuals who have variables data and are responsible for either manufacturing or validation sampling plans. This includes engineers, manufacturing and quality supervisors, and management. No special statistical or mathematical background is required.
Course Outline
- Overview
- Testing for Normality
- Reasons for Failing a Normality Test
- Preventing Failures
- Handling Failures
- Handling Variables Sampling plans During Validations (Writing Protocol)
- Handling Variables Sampling Plans During Manufacturing (Writing Procedure)
Software
Prerequisites
None.