SPC Sample Size

[hide_from visible_to=’public’]Please login with your site registration to view the recording.

[login_form redirect=”http://www.fmsreliability.com/accendo/webinars/spc-sample-size/”]

If you haven’t registered, it’s free and takes only a moment.
[/hide_from]
[show_to accesslevel=’free’]The recorded webinar streams from Adobe Connect servers. Click the play button to open a new browser window and view the recording. It may automatically start once loaded. Once playing you have control and can click ahead, replay a section, download files, and follow active links.

You may view an MP4 movie version of the event here (for non Flash playing devices)

Or, simply listen to event here

[/show_to]

Abstract

The purpose of control charting is to detect significant process changes when they occur. One important consideration in the selection and design of control charts (that is often overlooked) is the determination of an appropriate sample size.

A sample size should be selected that will align the statistical properties of the control chart with what are practically significant changes. That is, effective control charts will detect process changes that have a practical significance while minimizing signals resulting from smaller changes that are of minimal concern. In this webinar, we explore the impact that chart selection and sample size have on chart sensitivity (the ability to detect changes).

The factors to be considered when determining the sample size are explored and examples of determining an appropriate sample size for a typical application are presented. The relationship between sample size and sampling frequency is also discussed.

Finally, the use of OC curves to assess the risks of undetected process changes for a given sample size is presented.

Webinar Topics:

• Purpose of Control Charts

• Variation: Individuals vs. Averages

• Detecting Process Shifts

• Type II Errors

• Factors affecting sample size

• Calculating sample size for traditional (xbar) charts

• Sample size and sampling frequency tradeoffs

• OC Curves for assessing risk