|
Please plan to attend the ASQ Section 1206 spring conference on Tuesday, April 8, 2008 at Liberty Hall in Kimberly. National speaker and author, Davis Balestracci, will present his Data “Sanity” seminar. Dealing with the Data "Beast" - Simply... Sanely... Statistically
Despite what we've all been taught, "statistics" is not an arcane set of techniques used to "massage" data. In the current atmosphere of "accountability," there seems to be a new tendency for performance goals to be imposed from external sources, resulting in:
- Mass-produced tables complete with variances and rankings,
- Perceived trends being acted upon to reward and punish,
- Neanderthal budgeting processes,
- Labels such as "above average" and "below average" getting attached to individuals/institutions,
- People (especially politicians) being "outraged" by certain results and imposing even "tougher" standards.
These are very well-meaning strategies that are simple, obvious...and wrong! It is my hope that this seminar will:
- Help expose the severe limitations of "traditional" statistics in real world settings,
- Create awareness of the unforeseen problems caused by the exclusive use of arbitrary numerical goals, "stretch" goals, and "tougher" standards for driving improvement,
- Demonstrate the futility of using heavily aggregated tables of numbers, variances from budgets, or bar graph formats as vehicles for taking meaningful management action,
• Create proper awareness of the meaning of the words "trend," "above average," and "below average," and
- Enable people to ask better questions in response to everyday data issues.
|
Free Excel macro
There will also be a two-hour computer lab in the afternoon as part of the seminar to teach an Excel macro for run charts and control charts. Bring your laptop!
Meeting room temperatures may seem "cool" or inconsistent to some individuals. If you are more sensitive to cool temperatures, we recommend dressing in layers or bringing a sweater or jacket. We hope you can join us for this fun and informative conference!
Davis Balestracci
Davis Balestracci, M.S. statistics, has uniquely synthesized left-brain (analytical) and right-brain (psychological) approaches to quality with “the craftsmanship and passion of Beethoven composing symphonies.” While working for 3M, he received several corporate awards for his innovative teaching and uses of statistical methods.
His interests then evolved to utilizing the Deming philosophy in management and service contexts. For seven years, he functioned as a Deming statistical consultant for a major multi-specialty health care clinic with 500 physicians and 20 locations. He has since developed an international reputation for adapting statistical methods to healthcare improvement as well as innovative approaches to dealing with organizational cultural psychology and corresponding leadership issues through a philosophy of transformation.
Davis is a regular presenter at the prestigious Institute for Healthcare Improvement United States and European annual forums. He is known worldwide for a provocative, challenging, yet humorous and down-to-earth public speaking style, which one participant characterized as “the humor of Seinfeld with the intensity of Richard Dreyfuss” (Must have something to do with his Myers-Briggs “INFP” profile). People appreciate his awareness of the daily realities faced by quality “change agents”—including the inherent frustrations of dealing with “those darn humans!” This practical approach is described in his web site, book, Quality Improvement: Practical Applications for Medical Group Practice, 2nd edition (3rd edition in preparation), and monthly column in Quality Digest (www.qualitydigest.com).
Davis has a B.S. degree in chemical engineering, an M.S. degree in statistics, yet describes himself as a “right-brained” statistician (He is a pipe organist and has also done graduate work in orchestral and choral conducting!).
In 1995, at the invitation of Don Berwick, he was a member of a faculty team sponsored by the Harvard Institute for International Development that taught health care quality improvement methods in the Middle East to 80 health care leaders. He has subsequently consulted and given seminars in Canada, Palestine, Israel, Norway, Australia, New Zealand, Scotland, Wales, and, over 100 hospitals in England through the sponsorship of the National Health Service’s Modernisation Agency.
Davis is a senior member of the American Society for Quality (ASQ) and was the 2003-2004 chair of its Statistics Division. He is also a past president of the Twin Cities Deming Forum. He can be reached by phone at (207) 899-0962 or via e-mail: davis@dbharmony.com.
Directions to Liberty Hall
From Hwy. 41, merge onto WI-441. Take the CR-CE/College Avenue exit. Go east on CE/College Ave. Turn right onto Eisenhower Drive. Liberty Hall is immediately on the left at 800 Eisenhower Drive.
Special pricing/discounts
The conference fee is $200, however: current ASQ Section 1206 members will get a $150 discount, so their fee will only be $50. New members that join ASQ and our Section between Feb 1 and March 31 can attend for the cost of joining ASQ (the conference will be free!)
Spring Conference Registration
Register early as this conference is sure to fill up fast at these prices. Registration information will be available in future newsletters and on the website. |
|
Do you have a quality question or a problem at work you can't solve? Ask an expert! Please email Woody Santy at IQCWoody@instructor.net. Your question and Woody's response will be printed in the next newsletter.
Free advice! You can't beat it!
Sampling question (for “Ask Woody”)
We do incoming sampling inspection and, occasionally, other sampling. The samples are not pulled in a truly random fashion. The parts typically come in cartons on pallets. Samples are pulled from where the auditor can easily reach. That usually means a sample selected from the top layer. The amount of sampling/ inspecting time would increase greatly if a truly random sample was pulled.
What are the implications of pulling samples using this method?
Woody says:
The simple answer is an increase in consumer risk. How much of an increase cannot be quantified without delineating the assumptions.
I researched several books looking for more information on this subject. I found that there is a lot of information on Acceptance Sampling, but a very high percentage of it focuses on plan selection, AQL, OC curves and other such technical detail – all assuming valid sampling plans and good execution of random sampling, be it completely randomized, stratified, sequential, or systematic. Of the dozen or more books and web sites reviewed, I only found one discussion of this subject. The book is titled Industrial Quality Control by Charbonneau and Webster. On page 138 they list several myths about acceptance sampling:
“Random sampling is not necessary. Random sampling is as important in selecting a sample for acceptance sampling as it is in selecting pieces for control charting. An effort must be made to select a sample which will be representative of the lot. Random selection implies that each item in the lot has an equal probability of being selected for the sample. Individual laziness or indifference does more to hinder random sampling than anything else. It is easy to assume that pieces from the top of a heavy container or pieces from one of several boxes, represents the entire shipment. The random sample should contain parts from each box or each layer in the container. Failure to sample randomly can increase the risk of making wrong decisions.”
I have not found any literature that makes an effort to determine the amount of increased risk, which is at the heart of your question.
This is a response from Bob Einwalter. Bob earned his CQE in 1983 so has over 30 years of experience in the field of Quality.
Let me respond frankly with you. Implications of not using random sampling are apparent – you are probably wasting your time and fooling yourself. I suspect that the reason you are asking this question is because you are still getting ‘bad’ parts into manufacturing or assembly. This is the result of both the fact you may not be choosing the correct sampling plan AND aren’t random sampling. You can bet your supplier knows how you are sampling and on some occasional may even ‘seed’ (strategically placing bad/questionable product in locations where it wont get caught in inspection) a lot knowing how you sample. Some suppliers will even use the ‘seed’ to develop unofficial specifications because you used them once and never complained, so you will probably use them again.
The assumption with incoming inspection is to accept and reject loads based on risk. The underlying assumption is that you don’t believe your supplier will send you good stuff and you don’t trust them. Even with the typical sampling plans you are playing a game by choosing a plan that allows you to take more or less risk of accepting bad parts or rejecting good ones, neither is an acceptable outcome for your business.
Ronald Reagan one time said “trust, but verify”. Auditing material to validate is an ok reason for using incoming inspection, but it will never assure delivery of 100% good product. The trust part of the equation is working the supplier up-front to ensure they have good manufacturing and control techniques before you get the product. PPAP, SQC, proper gages, prints, and manufacturing processes at the supplier need to be verified. Once you can trust the suppler is making good parts, then you only need to verify.
In summary,
• Random incoming inspection will not provide 100% good parts, but will only reduce the risk.
• 100% inspection will also not provide 100% good parts, only using SQC techniques when parts are manufactured will do this.
• No incoming inspection plan will guarantee 100% good incoming parts.
• Build up trust with your supplier by working with them until you feel comfortable that they can produce good parts.
• Set up a completely random basis inspection to audit incoming product and use control charts and a Ppk capability analysis to ensure acceptable incoming product.
• Trust, but verify.
Thanks for the question! If any of the readers have insight or would like to weigh in on this, please send an email to myself or Jonna Gray (the newsletter editor.) --- Woody |
