The right way of choosing sample sizes based on valid statistical rationale

[Adam Fleming]

Performing at least some verification testings or validation studies of design-outputs and/or manufacturing processes is essential for almost all manufacturing and development companies. Yet, it is sometimes difficult to explain the rationale that goes into the selection of the sample sizes used in such efforts.

Removing this doubt and showing participants a way out of this quandary will be the content of a webinar that Compliance4All, a leading provider of professional trainings for all areas of regulatory compliance, will be organising. To understand the rationale for the selection of these sample sizes; just log on to http://bit.ly/2cObvdM

This webinar explains the logic behind sample-size choice for several statistical methods that are commonly used in verification or validation efforts, and will describe how to express a valid statistical justification for a chosen sample size.

The speaker at this webinar, John N. Zorich, who has spent 35 years in the medical device manufacturing industry, will offer the all-important guidance on how to justify such sample sizes. This will lead indirectly to guidance on how to choose sample sizes.

The justification for choosing a suitable sample:

He will also offer an explanation of the justifications for the section of such sample sizes and give an understanding of the ways by which to document them in Protocols or regulatory submissions. This can also be shared to regulatory auditors who may ask for them during a company’s onsite audits. Overall, this learning session will help participants understand ways by which to avoid regulatory delays in product approvals and to prevent an auditor from issuing them a nonconformity notice.

Point to note:

Participants of this webinar, however, are requested to note that this webinar does not address rationales for sample sizes used in clinical trials.

This webinar will discuss the following statistical methods:

o  Confidence intervals

o  Process Control Charts

o  Process Capability Indices

o  Confidence / Reliability Calculations

o  MTBF Studies (“Mean Time Between Failures” of electronic equipment)

o  QC Sampling Plans

 John Zorich will cover the following areas at this session:

• Introduction
• Examples of regulatory requirements related to sample size rationale
• Sample versus Population
• Statistic versus Parameter
• Rationales for sample size choices when using...
• Confidence Intervals
• Attribute data
• Variables data
• Statistical Process Control C harts (e.g., XbarR)
• Process Capability Indices (e.g., Cpk )
• Confidence/Reliability Calculation
• Attribute data
• Variables data (e.g., K-tables)
• Significance Tests ( using t-Tests as an example )
• When the "significance" is the desired outcome
• When "non-significance" is the desired outcome (i.e., "Power" analysis)
• AQL sampling plans

Examples of statistically valid "Sample-Size Rationale" statements