Many of the commonly used statistical tests and calculations of chart limits (or other measurements) require that the data be “normally distributed”. This webinar will show you how to check for normality in your data and apply transformations to non-normal data.
Learn the theory and concepts of determining when a normal distribution is needed, how to transform data that is not normal, and what to do when transformation does not work.
Why Should You Attend
The FDA requires that company to have “valid statistical techniques” are “suitable for their intended use”. Many statistical tests require that the distribution of the data used is normal. Assuming that the distribution of data is normal, without checking to see if indeed it is normal, will cause errors in test results.
Errors in results will cause bias of interpretation, rejection of lots that should be passed (or vice versa, passing lots that should be rejected), failing processes that are in specification (or vice versa…), and other problems. In essence, performing many statistical tests and other measurements without the basis of a normal distribution is garbage in, garbage out (GIGO).
In this webinar, you will also learn tools and concepts to understand what makes a distribution normal and when a transformation of data is, or is not, necessary.
- Regulatory Requirements
- History of the Normal Distribution
- The Normal Distribution in Mathematical and Visual Form.
- To Transform or Not to Transform?
- Tests to assess the degree of non-normality.
- Evaluating normality visually.
- When transformations can do more harm than good.
- Other Options when Data is Not Normally Distributed
Key Learning Objectives
- Understand when a test requires normal data
- Test and visually inspect data for normality.
- Learn when a transformation of data is needed, and how to transform data.
- When the “Normality Assumption” can be “relaxed”.
- Alternative tests and/or adjustments to use for non-normal data.
Who Will Benefit
- QA/QC Supervisor
- Process Engineer
- Manufacturing Engineer
- QA/QC Technician
- Manufacturing Technician
- R&D Engineer
- Data Scientist