theoretical Six Sigma background

Six Sigma – iSixSigma Forums Old Forums General theoretical Six Sigma background

Viewing 58 posts - 1 through 58 (of 58 total)
  • Author
  • #33190

    Markus Gahleitner

    Dear Sir, Madame,
    as a certified Six Sigma Black Belt I´m looking forward to find a theoretical background (theory) behind the Six Sigma concept because I´m acutally working at my doctor thesis at the university.
    Who could give me some advice (e.g. sources…..)
    Thanks in advance
    Markus Gahleitner


    Reigle Stewart

    You will find the theoretical basis of Six Sigma in a book
    (soon to be published) by Dr. Harry. This book answers
    questions such a “why Six in the phase Six Sigma” and
    “where does the 1.5 sigma shift come from.” As a
    technologist, you will appreciate this book. The title of this
    book is: “Resolving the Mysteries of Six Sigma: Statistical
    Constructs and Engineering Rationale.” You can also
    find this kind of stuff in a book called “Producibility
    Analysis and Process Characterization” authored by Dr.
    Harry and Dr. Lawson.



    Are you getting paid by Dr. Harry to promote him?
    I believe this is blatent promotion of goods or services which is not kosher on this forum. Please keep it to information and get away from the Dr. Harry promotion gig. Last time I checked he really did not need any help anyway.



    I agree with Stan, and I don’t find myself saying that very often :). This promotion is getting a little out of hand.


    Mel Alexander

    Davis R. Bothe wrote an article entitled “Statistical Reason for the 1.5 sigma Shift,” in ASQ’s journal Quality Engineering (Vol. 14, No. 3, MARCH 2002, pp. 479-487). There he addressed the question why six sigma followers add a 1.5 sigma shift to the average before estimating process capability.
    He also provided a statistically-based reason for using the shift with the process average, based on the control chart’s sensitivity to detect changes of various magnitude and devised a new capability index for incorporating this shift.


    Reigle Stewart

    I do not agree with your position on “the promotion of Dr.
    Harry.” I find your accusations most offensive and
    unprofessional. Please enlighten me as to the line-of-
    separation between trying to “answer a person’s question
    to the best of my knowledge” and “promoting the source
    of that knowledge.” I do not believe my post in this thread
    can be considered as “promotional” in accordance to the
    website’s posted standards and expectations. Should I
    delete Dr. Harry’s name? Should I not cite or reference
    his book that contains information that has been
    requested by a collegue? Should I not provide my fact-
    based opinions about the quality and validity of his work?
    Should I just say nothing because it might involve “his
    name?” Does his voluntary work and countless hours of
    contribution providing Q&A on this website constitute a
    “promotion.” Some have even went so far as to say he is
    getting paid to provide the Q&A (how absurd and
    offensive). How would you feel making such
    contributions and hearing such falsity? Understand that it
    was my privilege and honor to work with this fine man
    over many years. I cannot begin to tell you all the things I
    learned from him about engineering, statistics, quality,
    and most of all, Six Sigma. You know he has been
    referenced over and over as “the Godfather” of Six Sigma
    (and one of its original architects). If somebody talked
    about the work of Dr. Deming or Dr. Juran would you call
    this “promotional.” The merits-of-their-work is just that …
    the merits of their work. Because of this it is only natural
    that I cite the sources I know and “promote” the ideas that
    I have watched change the course of corporations. How
    many others have done this in the field of quality? Should
    everyone avoid referencing their names or books so they
    do not stand accused of promotion? Does my mere
    association with him automatically spell “promotion.”
    What kind of predjudice is this? This website has specific
    standards related to the “promotion” of products and
    services on the discussion forum, not about the
    “promotion” of ideas, questions, and reference material.
    Please enlighten me as to my transgressions with respect
    to this policy. You are so quick to accuse, are you just as
    quick to substantiate? I look so forward to examining your
    set of self-referencing standards on the next post. With
    such ethnocentric filters, you can feel much more
    confident about upholding the First Amendment for all of
    us. After all, we do need good “information policemen.”
    No, I do not get a dime from him for saying anything. He
    has no PR firm because knowledge of his work speaks
    louder. Perhaps the roots of such objection are found in
    soil of professional jealousy.



    Fair enough, Reigle. Here’s my data:
    In four separate posts within the last two days you have specifically referred to Dr. Mikel Harry’s “Resolving the Mysteries of Six Sigma: Statistical Constructs and Engineering Rationale”.
    iSixSigma states: “The iSixSigma discussion forum is not a medium for user or business advertising.” defines advertising as: “The activity of attracting public attention to a product or business, as by paid announcements in the print, broadcast, or electronic media.” defines product as: “commodities offered for sale.”
    Now the ball is back in your court. Is Dr. Mikel Harry providing this document free of charge, or offering it for sale? If it is free of charge, where can I obtain a copy?



    Dear RS,
    I hope you are able to take Terry & Stan’s feedback on a positive note – I am sure your intentions were very positive and there was no ulterior motive in recommending the author / book.
    However, one thing I would like to point out here – you do seem to be recommending the book a little too often and anybody who believes in data (myself included) would tend to agree with Tery / Stan.
    I hope you understand / appreciate the fact that I may be a good Christian and believe in Him and His ways – but do I tom-tom about it?
    No hard feelings…
    P.S. I ran a test and found that your 4 posts in the last couple of days do not add up enough proof so as to be blatantly accused of advertising ( p value is above 0.05). Since I cannot reject the Null I am giving you the benefit of doubt and hence this post.
    However if a fifth recommendation does come in again, chances are that the p value will dip below 0.05 and…………
    (just kidding……….;)……..)



    Your “lovefest” on here with Dr. Harry answers no questions. You are promoting a book that does not yet exist in the marketplace that is reported to prove the 1.5 shift mathematically and statistically. Truth is there is no proof. I can prove some companies experience less than a half sigma shift with data. Are they defying the laws of physics? I doubt it. The only proof out there is that no process can stay perfectly centered. But prove 1.5 sigma shift – just more hand waving by the PT Barnum of Quality.
    Stop promoting under the gise of answering. I almost bought into your “Guess Who” post, but you have made it clear you are just enamoured with the guy. Maybe your attack on Mr. Carnell was just jealousy.


    Reigle Stewart

    Bee:Thank you for your kind comments. I guess it is all a
    matter of perspective. Your type of demeanor is greatly
    appreciated and is very much taken to heart. The
    following is from me (personally), not someone else who
    is “referenced.” Others in this thread who still remain
    cloaked by code names and without posted credentials
    boldly talk about how they “know” the shift factor is not 1.5
    sigma. In an earlier post I laid out a step-by-step
    example. All necessary statistical details are present. It is
    so interesting that the problem I laid out will reveal a 1.46
    sigma shift (about 1.5 sigma), but yet none of you (Stan
    and Terry specifically) have worked through the problem,
    and then made conclusions. Stan and Terry, I await your
    analytical position. Right now your position is without any
    form of validity. How can you have the public audacity to
    make so many unfounded statements and not make your
    “proofs.” You talk about P.T. Barnum? I have made the
    shift information available and known to you in a scientific
    manner, the way of Six Sigma but you have no math to
    show. Now who is a part of the circus and who is
    “delivering the data.”? At least Mike Carnell uses his God
    given name. It is apparent that a handful of you don’t
    want to practice science and engineering. All you want to
    do is quibble about trivial things like “promotion” and
    quote the dictionary. How about quoting yourself through
    some good solid science. Please work through the
    problem parameters I given you and respond in a
    professional way … through the language of statistics,
    mathematics, and engineering. I ask that you don’t
    dissappoint me Stan and Terry. Even if you have some
    trouble with the concepts, I will help you through it.


    Reigle Stewart

    Terry:Perhaps you speak and make quotes out of context. Here
    is the full text clearly posted on iSixSigma .com . It is
    apparent that you know this text.Question: “Some posts are made by people pushing their
    own businesses on the forum, but no added comment is
    made following their posts. What is the guideline for
    posting products or services on the forum?”1) I am not “pushing my own business on the forum.”
    2) I have provided comments (of a scientific nature)
    following my posts.iSixSigma’s Response: “The iSixSigma discussion forum
    is not a medium for user or business advertising. Solid
    and factual posts by industry experts (who often work at
    companies selling Six Sigma related products and
    services) tend to generate sales leads because of
    intrinsic value, not promotional text.”1) I am the “user,” not Dr. Harry.
    2) I am not advertising a business (his, mine, or anyones).
    3) I am presenting “solid factual posts.”
    4) I am a recognized “industry expert”
    5) I do “often work at companies selling Six Sigma
    products and services.”
    6) I am promoting things of “intrinsic value,” not
    promotional text.Now that I have responded to your comments in a
    professional way, do me the same courtesy and respond
    to my “demonstration of the 1.5 sigma shift” in an earlier
    post. You and Stan should do this in a scientific way … a
    statistical way … the way of science … not through the
    dictionary. Please, step up to the plate and make your
    way of science known to all of us. This is not rhetoric.
    This is a plain old challenge. This is what good science is
    all about. Don’t disappoint us.


    Reigle Stewart

    This is wonderful that you have scientific knowledge
    about this subject. Lets see your “proof.” By the way, the
    1.5 sigma shift is an equivalent shift that is used by design
    engineers to better analyze the producibility of a design.
    So emperical “proofs” are not applicable. Do you even
    know what you are talking about? I does not sound like it.
    Please show us your mathematics and engineering.
    Stan, its time for you to “put up or shut up.” Place your



    I have asked you this before, but got no answer: You said:
    “…the 1.5 sigma shift is an equivalent shift that is used by design engineers to better analyze the producibility of a design”
    “the 1.5 sigma shift is very closely tied to the mathematics that underlies a design- process qualification (DPQ) procedure”
    I asked:
    “How on Earth do we get from 1.5 sigma shift in DPQ, to count defects, count opportunities, find how many sigmas away has an unilateral specification limit to be from the average of a normal distribution to produce as many defectives per units as defects per opportunity you found, call that “long term sigma level”, and add 1.5 to that number to get the “short term sigma level” regardless the sample size, of the process distribution, and of how it will be controled ?!?!?!?” (for an existing process, I meant. Not for DPQ)


    Reigle Stewart

    Gabriel:I have tried to share my knowledge in good faith, but you
    can see from previous posts how the act of “sharing and
    caring” is greeted. It seems that a lot of people really
    don’t want to discuss Six Sigma on a technical level in a
    scientific manner. This is so disappointing. Yes, I
    did address your question. Just work a little with the
    problem parameters I previously posted and you will see
    the answer. You probably know that is not possible to
    give every detail on a discussion forum. At the risk of
    being “promotional,” you really need to “drink from the
    source.” Just because answers dont jump up and bite
    you in the butt do not mean they don’t exist.



    Empirical data that supports a theory does not prove a theory, but empirical data that does not support a theory does indeed disprove a theory. If anyone wants to see evidence of musch less that a 1.5 sigma shift, they don’t need to go any further than the TPS.
    Do yourself a favor and step out of theory for a moment and go look at data. Your answer about Lean tells me you have not taken the time to learn and understand Lean.
    Theoretical pontification about better analyzing the producibility of a design is meaningless. If I am designing a product for a process I know I can control better than a 1.5 shift, why would I use 1.5? Seems to me your approach works only when I don’t really have process knowledge (and why would I do that?)


    Reigle Stewart

    Stan:I will no longer participate in further discussion with you.
    My position is clear and scientifically verifiable. Your
    math has not materialized. You speak of proof but
    produce none. Seems I keep getting pulled into debates
    where the rules of engagement are opinion based … not
    fact based. Therefore I must withdraw. It is amazing how
    some people (much like yourself) have apparent difficulty
    with technical things. Look likes they always “poo poo”
    the importance of things like math … much like some
    people “poo poo” the importance of higher education
    (when they have none). Remember what some people
    say about fancy cars and big diamonds … its always
    “gawdy” until you own it. Being “street smart” is of little
    consequence when someone needs to know how much
    thermal shock will occur under certain operating
    conditions (befor the product is ever produced) …
    especially for project that can impact human life. Like it or
    not, every phenomenon has a mathematical basis …
    some of which the world knows and has equations for
    and some of which the world does not know … this is the
    mission of science. Sound design engineering requires
    that we often work in the absence of emperical data.
    Again, you can “poo poo” this all you want, all day long,
    but in the final analysis, you are not practicing science. If
    was not for theory and math the practice of engineering
    could not exist … it is doubtfull mankind would have ever
    reached the moon … and beyond.



    Your math has not materialized, just your clear idol worship and promotion of Dr. Harry. Withdraw if you want, but please stop your blatent promotion of a non existent book.
    The best design execution organizations in the world do not use or need your 1.5 theory. It is not scientifically verifiable, but I do agree it is useable in the absense of real process data.



    Reigle, I do not view your posts as promoting Dr. Harry’s books, and thank you for choosing not to argue with these fools – you are better than they are.  Bee was polite and professional, while Stan is a legend in his own mind.  The saying “Never argue with a fool – people might not know the difference” holds true here; based on your work, we know you’re not the fool.  I normally don’t post here unless I feel I have something constructive or helpful to add to the forum, but this was getting a little one-sided and drawn-out.  I look forward to posts from knowledgeable, positive-minded people who are willing to share their knowledge and don’t have an agenda or axe to grind – and just hydroplane over the rest of the crap.  The silent masses are interested in learning the tools from those who know; please continue to share your knowledge and information, and we’ll continue to learn.
    Thanks again!



    Thanks for the vote of confidence. You may note that people have said much worse.
    The point is the assertion that Dr. Harry has a book coming, endorsed by MATHEMATIANS (wow) that proves the 1.5 sigma shift. Go read the book The Man who only Loved Numbers for some insight into proof of math theory. The truth is that there are thousands of processes that demonstrate less than a .5 sigma shift over a period of years. In math, empirical evidence of a theory not holding is proof the theory is not true. So why all of the noise of proof? Well Dr. Harry claimed to have proof with the famous Bender, Gilson, and Evans reference on page 9.8 of his famous cartoon book The Vision of Six Sigma: A Roadmap for Breakthrough. Turns out to be a bunch of hand waving, just like the book.
    Why care about this? One simple reason, the 1.5 sigma shift is useful when rational controls (not SPC) are in place, where detailed process knowledge is not available. Done, finished, thats it. Reigle’s own spreadsheet (circa 1996 AllidSignal) for predicting performance allows for putting in a number other than 1.5 where valid data is available. Why? Because when real data is available why assume anything.


    Reigle Stewart

    Mike:Thank you for your up-lifting comments. You tactfully
    reminded me of what matters most … passing on to
    others what we know. The “vital many” is what matters,
    not the “trivial few.” I guess this is the “Inverse Pareto
    Principle.” Did I just coin something new”? Because of
    people like you, I will again try to skillfully hydroplane
    over the pollution and continue to share my knowledge. It
    already feels better. I will share with you a quote “PT
    Barnum” sent me a couple of days ago …
    H. G. Wells once said “Moral indignation is jealousy with
    a halo.”
    Respectfully,Reigle Stewart,
    Old-Bald-Fat-Guy (OBFG)


    Reigle Stewart

    Stan:I am glad to hear you have read my book and used my
    software (spreadsheets). References are the sincerest
    form of complement that can be paid to someone. Could
    you gladly provide the titles of your books, papers, and
    software so everyone else can review your work and
    defending position. I wish you the best of luck with all
    your future literary works and research.The Old-Bald-Fat-Guy.



    Stan, I won’t lower myself to your level.  From your posts, you sound technically proficient, and some of your analysis may be sound, but you would do well to learn some humility and people skills.  Since we’re on the Book of the Month kick, I recommend that you read one titled How to Win Friends and Influence People, by Dale Carnegie.  Coupled with your technical bent, a compassionate, nurturing personality would be a great combination.



    I respect your work, but don’t buy off on the idea of a math proof of the 1.5 sigma shift. I’ve got way too many processes to run much better than that.
    As far as my published works – don’t have any, never felt the need. Certainly did not want to join the ranks of all the nonsense out there about SS. My opinion is your work and that of Ron Lawson are the best there are and that was 15 years or more ago.
    I have no disrespect for you only your assertions about proof of things that are not true. I know how to control a process with less than one third the shift you talk about. Most of the people in the TPS know how to as well. Go look.



    Thanks for the recommendation. I was the top of my class at Dale Carnegie (but I did think it was a joke on them).


    Kim Niles

    Dear Markus:
    Here are some related bits taken from my CSUDH MSQA thesis entitled “THE IMPORTANCE OF Metrics ON SIX Sigma Program IMPLEMENTATION”. 
    The 1.5 Sigma Shift Allowance is a way of approximating and accounting for long-term variation of the mean of a process metric.  It also conveniently allows Six Sigma targets to appear as more realistic goals (i.e. 3.4 dpm vs. 0.002 dpm).  Mikel J. Harry, the President of the Six Sigma Academy, is credited as being instrumental in the development of Six Sigma and the 1.5 sigma shift (Maguire, 1999, p.27-34).
    There are two main ways that Mr. Harry attempts to support the shift statistically.  Both ways are highly controversial.   Mikel Harry sites several old articles on statistical tolerancing and process shifting as a method to justify the shift.  Harry quotes Bender (1975) and Gilson (1951) to support this position (as cited in Harry, 1997).  An article by Tadikamalla points out that Harry is taking these articles out of context.  Tadikamalla states that shifts in the mean of a stack of disks would be expected to shift given the various disk sample combinations in the stack.  However, he says, “Harry seems to have misinterpreted the factor of 1.5 as the allowance for the shifts in the mean of a single component due to its being manufactured in different lots.  It does not make sense, to allow for a 1.5 sigma shift in the process mean of individual components” (Tadikamalla, 1994, p. 83-85). 
    The second slightly different and also misleading method used to support the shift is in regards to varying sample sizes and shifts in distribution normalcy.  Smaller sample sizes result in apparently larger variation and greater reported shifts even if the population universe does not change.  Harry (1992) refers to Evans (1975) to justify the 1.5 sigma shift in this manner.  He quotes Evans:  “…. shifts and drifts in the mean of the distribution of a component occur for a number of reasons…for example, tool wear in one source of a gradual (nonrandom) drift…which can cause (nonrandom) shifts in the distribution.”  Harry suggests that a generalization can be made and develops a mathematical equation for “the magnitude of inflation imposed on the instantaneous reproducibility with a compensatory constant used to correct the sustained reproducibility for the effect of nonrandom manufacturing errors, which perturbs the process center.”  He claims that the general range of the compensatory constant is between 1.4 and 1.8 and uses Z shift = SQRT {[c^2(ng-1) – g(n-1)]/ng} where c= his constant, n = subgroup size and g = number of subgroups; n is usually between 4 and 6, and g between 10 and 100.  Using c = 1.8, n = 5, and g=50 results in a Z shift of 1.49.  He calls this the standard mean shift correction.
    There are also motivational and practical reasons to support the shift.  Gary Wasserman (2000) is one of several that believe that in the real world special cause variation is the norm, data is often serial correlated and or non-normal.  Therefore, the shift tends to negate some of these real world data integrity problems (p.23).  
    Gregory Watson (2000), ASQ president in 2000 states in regards to the controversy over the statistics behind Six Sigma that there are three groups of thought, those in favor, those against, and now a new group of those taking the middle ground.  He says that those in the middle see the controversy but recognize that management is paying more attention to Quality.  He says, “It is acceptable to have a non-purist approach to statistics that gets results quickly, if that is management’s decision.” (p.16). 
    I hope that helps.


    Reigle Stewart

    Since you did not understand the “shift factor” from a chi-
    square definition, I am now presenting it from the Z
    distribution angle. Of course, the Z distribution is
    theoretically related to the chi-square distribution. So
    here is the lesson I made for you:1) Sample 4 measurements (N = 4).
    2) This is short-term sampling (like would be expected in
    a qualification test).
    3) Accept that the measurements are random and normal.
    4) Set the statistical confidence at .9973 (just like control
    limits in SPC).
    5) Transform the mean to 0 and the standard deviation to
    6) Compute the confidence limits of the mean using Z
    7) = mu + / – Z a/2 * [S / sqrt (N)],
    8) where a / 2 = .0027 / 2 = .00135 and Z is corresponding
    standard normal deviate.
    9) The Z value is shown to be 2.9999. This is close to 3.0.
    10) In this way we compute:
    11) = 0 +/- 3.0 / sqrt (4)
    12) = 0 + / – 1.50.
    13) This is the “shift.”
    14) This is how “sensitive” things are in the short-term (+ /
    – 1.5 sigma).
    15) This is the “real world,” like it or not.
    16) This means that 99.73% of the averages will be within
    the confidence interval .
    17) This means that 99.73% of the averages will be
    “shifted” from the true mean (mu) by 1.5 sigma or less.
    This is the long-term expectation.
    18) This means the “statistical worst-case” shift is 1.5
    sigma (in the short-term).
    19) This means that design engineers should “shift the
    nominal condition” of the design CTQ’s by 1.5 sigma so
    they can test the “design sensitivity” to process centering
    errors using Monte Carlo simulation. This allows
    engineers to examine the “worst case” producibility of
    their specifications and then make adjustments to ensure
    robustness to this level of process centering error.
    20) If you don’t get this, I don’t know what else to say …
    go to black belt training (or maybe get some retraining).Your constant servant and humble guide into the world of
    Six Sigma Engineering. Maybe you could give the rest of
    us such a lesson. We all want to learn from others with
    good knowledge.
    Reigle Stewart.


    Old Bald Fat Guy

    Kniles:Excellent discussion. I have read the work you refer to in
    your discussion thread. It was very well done but as I
    recall he did not provide any analytical support to the
    contrary. Having worked in mechanical engineering and
    product design for many years I can tell you he is in error
    about a lot of his statements. I just want to say what a
    GREAT job you did preparing your response. Could you
    read the one I just gave to Stan and comment on it
    (showing the shift using the Z distribution). You obviously
    know what you are doing and I would appreciate the
    sound interaction.Many thanks, Reigle Stewart (Old-Bald-Fat-Guy)



    By the way, I remember reading this book first time as a boy about 50 years ago and having gone thorugh it several times since then and not noticed any “mathematical” demonstration of its contents.  Experience has shown me that there are other types of demonstrations which have held true on other bases.
    Hope this helps



    Along with your “Your constant servant and humble guide into the world of Six Sigma Engineering.”, you forgot sarcastic and arrogant (both traits I too exhibit).
    Your example is interesting, but trivial. For example, what if I chose a sample size of 1 – my shift according to your logic would be 3. What if I chose a sample size of 9, my shift according to your logic would be 1.
    Does not make sense. Try to teach a lesson that does make sense next time.
    The lesson I give is don’t believe all the hype. Six Sigma is just a set of tools and the 1.5 is just a useful guidline in the absense of real process data. Any rational use of SPC or Lean tools will immediately detect and correct excursions of 1.5. The Six Sigma hype does not even consider them excursions, they are assumed to be permanent. If you don’t believe it, transform attribute data into Sigmas.
    Your constant arrogant and sarcastic critic of the world of Six Sigma Engineering.


    Reigle Stewart

    Stan:For the second time, please do not respond to
    my posts. I have no interest in conversing with
    you.Respectfully,Reigle Stewart



    I promise to only respond when you claim proof (math, statistical or otherwise) where there is only slight of hand like your normal distribution example or where you are clearly only on here to promote Dr. Harry.
    You seem to think you are the light and the way. I just feel a need to balance that.


    Dr. Scott

    Do you have a way I might contact you? I would like to run an approach to “process characterization” by you and get your thoughts.
    Dr. Scott


    Reigle Stewart

    Dr. Scott:Perhaps the best way to grasp, digest, analyze,
    and comment on your approach to process
    characterization is to clearly post it on this
    discussion forum. I would recommend that you
    provide the methodology in a “cookbook” format.
    This will help during the review process. We all
    look forward to your post.Reigle Stewart



    Reigle,When I access Six Sigma Management Institute I can see that Dr. Mikel Harry is the Founder, President and Chairman with you listed directly below him as a Vice President of the company. What did your previous post say????Looks to me like an intentional effort to promote your company against the rules & regs of this forum. This is factual information as well. Hope you fair better than you did as President of Six Sigma Supplier Institute.Cowboy Up…Yeeeha! “Hey little cobra don’t you know you want to shut’em down”HAUGHT



    I noticed Uniworld is doing your deployments for you. 
    Back in the TAG days Mikel used to refer to the management as “the Six Sigma Consultant Pimp” Does this make Antis your “SS pimp?”


    John Hickey


    Quoting Mikel is not science.
    Showing your ignorance about Lean is not science.
    Giving cute little slight of hand examples of chi-square shift proofs is not science.
    Take your own advice and stop promoting Dr. Barnum woops I mean Harry.



    I am so disappointed –
    “When I access Six Sigma Management Institute I can see that Dr. Mikel Harry is the Founder, President and Chairman with you listed directly below him as a Vice President of the company. 



    Since your credibility has seriously been damaged after finding that you are in fact an employee of Dr. Harry management institute after you claimed you were not promoting.  Why don’t you clarify on another point of yours……You claim Mikel has repeatedly been referenced as the “Godfather of SS” can you provide credible references on that?  We’re not asking for another Mikel puppet.
    You have chosen to play Renfield to Mikels Dracula. That doesn’t mean the rest of us are willing to sell our souls.
    Time to “cowboy up” and reference some of that mathematical, or scientific elitist garbage you have been handing out.
    Btw – What is Mikels PhD in?


    Reigle Stewart

    To my colleague Haught:I do believe you will find Dr. Harry clearly
    referenced as the “Godfather of Six Sigma” in a
    number of well-respected sources. For
    openers, consider the October 1999 issue of
    Quality Progress Magazine (a publication of the
    American Society of Quality) where he is cited
    as the “Godfather of Six Sigma.” You will also
    find Dr. Harry substantially elevated on page 2
    of Rath & Strong’s “Six Sigma Handbook.” In
    addition to this he has also been cited by
    Process Quality Associates: “To elaborate the
    evolution of Six Sigma, one Six Sigma authority
    has to be introduced: Mikel Harry, who is called
    the “godfather” of Six Sigma and is
    acknowledged as the leading authority on
    theory and practice. Even though he did not
    invent the concept, the way that it is currently
    practiced bears the unmistakable marks of
    Harry’s personality and personal history.
    Harry’s history path is followed here to reveal
    the evolution of Six Sigma.” From the website
    of Statware Products: “Please Note: The
    definitions of Six Sigma we used in the
    discussion below were taken from a recent
    interview with Mikel Harry, the “godfather” of Six
    Sigma … ” In the early 80’s Dr. John Ramberg
    was hired as an outside contractor by Motorola
    Inc. to help the Motorola Training and Education
    Center (Schaumberg, Ill) put together a design-
    of-experiments course (and some other quality
    related stuff). In the article “Six Sigma: Fad or
    Fundamental” printed in Quality Digest
    magazine, Dr. John S. Ramberg states “The
    late Bill Smith, a reliability engineer at Motorola,
    is widely credited with originating Six Sigma . . .
    Harry later teamed with Smith on the Six Sigma
    initiative and created Motorola’s Six Sigma
    Institute prior to forming his own firm. Smith and
    Harry’s initial Six Sigma umbrella included SPC,
    ADT and PE. Later, they added design for
    manufacture (product capability and product
    complexity) and, as quality was linked to
    business performance, accomplishing quality
    through projects … As Motorola set out on its
    quality journey, Harry noted that the company
    ran into a five sigma wall. Motorola found that it
    could attain a three-sigma level by installing
    process improvement and control in its own
    installations, and improve this to the four- or
    five-sigma level through the education of its
    suppliers … At this point Harry wrote a strategic
    vision for accelerating Six Sigma. This included
    a change in focus, anchoring quality by dollars
    and seeking a business transformation. It
    included a description of different competence
    levels in the Six Sigma methods, which, in the
    karate tradition, were designated by belts–
    Green Belt, Black Belt and Master Black Belt …
    It’s no wonder Six Sigma has gained industry’s
    Respectfully,Reigle Stewart


    Reigle Stewart

    To my esteemed colleague Haught:I thought you might enjoy some more quotes
    from various credible authorities (since I have
    so little credibility). The following is a quote
    from “General Electric’s Six Sigma Quality
    Program: A Statistical Methodology Requiring
    Training and Leadership” prepared by Dr.
    Shelton P. Rhodes, Sal Aurigemma, Debra
    Basilis, Robert Batie, Susan Anstead Bell, and
    Joseph Burroughs, University of Maryland
    University College, July 1998. They state
    “Theoretically, Six Sigma equates to a variation
    about the mean of no more than .002 dpmo
    (defects per million opportunities) outside
    allowable standards. In industry terms,
    achieving Six Sigma quality ensures that the
    defects in a given process or procedure do not
    exceed 3.4 dpmo. The difference between the
    two definitions stems from Dr. Mikel Harry who,
    while working as an engineer and statistician at
    Motorola Inc. in the mid-1980’s, noted that the
    mean of a normal Motorola process shifted up
    to 1.5 sigma in either direction of the target
    mean (see Figure 2). Applying this mean shift
    to the six sigma range variation, Dr. Harry
    designated the Six Sigma Quality concept to the
    world as 3.4 dpmo (3.4 dpmo is actually 4.5
    sigma with a stable mean). This change in
    definition of Six Sigma was not Dr. Harry’s major
    contribution. The most significant impact he had
    upon Motorola, and now GE, was the creation of
    a data-driven methodology that uses detailed
    statistical analysis to pinpoint process flaws.
    Additionally, by changing the quality discussion
    from percentage to dpmo, Motorola identified
    that modern technology had progressed far
    enough that traditional quality levels and
    definitions were no longer acceptable (Pyzdek,
    1997). Also note the quote from India Infoline
    (Business School): “The seed for the birth of Six
    Sigma was sowed when in 1979, Art Sundry, an
    executive at Motorola had the courage to stand
    up and question the quality of the Motorola
    products. This was followed by an extensive
    statistical study done by Bill Smith, an engineer
    at Motorola’s Communications Sector, who
    studied the correlation between a product’s field
    life (i.e.; product life in reality) and the number of
    times the product was repaired during it’s
    manufacturing process. In 1984, Dr. Mikel Harry,
    a senior staff at Motorola’s Government
    Electronics Group (GEG) created a detailed
    road map that took the approach of problem
    solving through statistical analysis to improve
    the design of the product, reduce the production
    time and reduce costs as well. Company saved
    $ 2.2 bn over a period of 4 years. Dr. Mikel Harry
    was joined by Richard Schroeder, VP and GM of
    customer services of Motorola’s Codex
    Subsidiary to further enhance the
    implementation of this concept in other
    companies. In 1994, they together founded the
    Six Sigma Academy Incorporate at Scottsdale,
    Arizona and their first clients were GE and Allied
    Signal. GE had Six Sigma related savings of $2
    bn in 1999 alone and Allied Signal had $1.4 bn
    (1992-96).”Regards,Reigle Stewart
    Old-Bald-Fat-Guy (OBFG)


    Tony Burns

    If we put aside the controversial origins of the +/- 1.5 sigma shift in the mean, any drift in the mean will cause points to fall outside control limits. A shift of 1.5 sigma will cause about 7% of points (depending on the shape of the distribution) to fall outside a control limit. Not something that would please customers! If every process has a +/- 1.5 sigma shift in the mean over time, can no process ever be “in-control” and therefore predictable ?
    Shifts in means do occur in processes such with tool wear, as described in the original six sigma papers by Bender, Taylor etc on tolerancing. The amount the mean changes depends on the amount the tool wears before adjustment. Shifts in means hopefully (or by good quality management) do not occur in the myriad of continuous process that are more common.
    Dr Tony Burns
    [email protected]



    Have you got anything to contribute to this forum other than Dr. Harry myths?
    Show of hands former Motorola employees – how many of you learned anything from Dr. Harry?
    I was busy working achieving >5.4 sigma in 1990 while the esteemed Doctor was hob knobing with a bunch of theoretical guys and promoting himself. We were exposed to the great Doctor in 1989 and everyone just laughed at his two fisted overhead flim flam show.
    No learning, no impact and we were amoung the first to get better than 5, 5.1, 5.2, … sigma. The great doctor was promoting his personal fame and not in the trenches where the real learning always takes place.
    Aren’t you a little old for hero worship? If not, try Pokemon. Everyone knows that is fantasy, so you will not confuse anyone.
    Or is it right that you really are VP of Mikel’s SSMI? Promoting goods and services is a no no. You are a slow learner.


    Reigle Stewart

    Stan:I am sorry you feel the need to degrade me with
    your personal attacks. It may very well be that
    your personality demands such venting. If so,
    its OK. Why is it that you seem to loath highly
    successful people? Without a doubt it was Dr.
    Harry that “made” the world of six sigma. You
    speak of being “in the trenches.” Seems to me
    that Dr. Harry spent a great deal of time “in the
    trenches.” Consider a few of the numerous
    case studies that Dr. Harry conceived, designed
    and executed while working as an engineer and
    scientist. He published a few of them in one of
    his books (1994): “The Vision of Six Sigma:
    Case Studies and Applications” Tri-Star
    Publications. You might want to read his work
    before you so easily make your condemnations.
    Here is a listing of those cases that are
    significantly detailed in the book: 1) A
    Quantitative Analysis of Solder Deposition
    Characteristics Resulting From A Wave Solder
    Process 2) An Engineering Proposal for the
    Design and Analysis of an Experiment to
    Surface Significant Manufacturing Variables of a
    Causal Nature Related to MIL-P-28809 3) The
    Results of an Experiment Conducted to Surface
    Significant Variables Related to MIL-P28809a 4)
    A Quantitative Analysis Related to the Effects of
    Flux Type on Corrosion 5) An Experimental
    Study of the Effects of Various Process and
    Design Characteristics on Solder Joint
    Cracking 6) A Statistical Analysis of
    Encapsulation Material Cracking 7) An
    Exploratory Study of the Effects of Component
    Tinning on Solder Joint Quality 8) Derivation of
    Second Order Polynomial Equations for
    Reverse Ballistic Air Gun Applications 9) The
    Evolution of a Statistically Designed Experiment.
    I have the book and keep it on my desk. I have
    studied it carefully over the years. I have
    learned a great deal about design of
    experiments from Dr. Harry and he has shown
    me many tricks-of-the-trade on how to make
    DOE’s even more powerful … especially during
    the design process and R&D. Maybe you could
    make your work available to us so that we can
    learn from your experiences.With kindness,Reigle Stewart
    Old-Bald-Fat-Guy (OBFG)


    Kim Niles

    Dear Dr. Burns:
    Regarding your first scenario (7% of points (depending on the shape of the distribution) to fall outside a control limit) … then I argue that you have your control limits set too close together … because the process is likely to shift if it contains more than just a few control variables and given enough time.  From my experience working with 30+ significant control variable manufacturing processes, they all shift given enough time.  All 30+ of those significant control variables are tugging away at the mean in their ever-present quest to reach entropy … [smile]. 
    Reviewing your second point (shifting due to tool wear) helps me explain my line of thought.  I agree with you as I understand your point that shifts due to tool wear are manageable and therefore are more of a quality control problem than a natural occurrence.  However, I go on to say that most likely all 30+ significant control variables in my stated typical manufacturing process are also manageable in the same way and therefore at some point we have to consider shifting a noise based natural occurrence relative to the effort it would take to understand and manage all the control variables tugging away at that mean.   Practicality begets noise, which begets the 1.5sigma shift. 
    KN –,



    The question to you was to supply references to substantiate your claim of “..referenced over and over as the Godfather of Six Sigma….”
    Your response has a lot of non-answer answers but it comes down to once. Once in Quality Progress. You have wrapped a lot of arm waving and smoke and mirrors in the typical Mike Harry show but if we read the response it is once.
    This is the new image. The Roy Rogers thing is gone and now we are into pin stripe suites? I suppose we will have to listen to stories about hanging out with Sammy the Bull in Tempe. Please spare us round two.
    You actually do not believe you have damaged if not completely destroyed credibility? Are you completely deluded? Your claims to not promoting Mikel. Your claims to no common interest. Your claims to multiple references. You are damaged goods. You try to wrap yourself in a cloak of science and mathematics and all you really are is Bunny Rabbit to Mikel’s Captain Kangaroo. How does that hand feel? You hit this site with arrogance and condescension. The most insulting part is that you are so arrogant and condescending that you believed you could do a website and not be found out. The second most insulting part is you think you are so revered in the industry it won’t matter. Most people had enough of the ego over the last decade.
    And the Phd is in what?
    As the saying goes “you saddle the horse, you ride it.”
    Cowboy up.


    Reigle Stewart

    To my emotionally suffering colleague Haught:I express my deep sorrow at your compulsive
    need for open hostility. I have spent my career
    in engineering trying to be constructive, learn,
    and create. It is apparent that you do not want to
    share in any of this. That’s OK … nobody twists
    your arm to read (or respond) to my posts. It
    seems you seek nothing more than to batter the
    accomplishments of others without: a) revealing
    your true name, b) providing constructive
    counter-arguments of a technical nature, or c)
    even acknowledging the engineering and
    statistical merits of that which has been
    provided. Again, what is the probability that
    everything I have presented or discussed is
    without merit and is all wrong? Because of your
    relentless pursuit to be personally and
    professionally defaming I will no longer
    respond to you. Again, I forgive you for your
    improprieties but will not be apart of reinforcing
    them. It is so easy for you to hide behind the
    name “Haught.” You seem to be like those that
    hide behind their automobiles on the freeway …
    darting in an out of traffic … making certain
    nasty gestures … ridding the bumpers of others
    … while screaming what bad drivers they are.With deepest pity,
    Reigle Stewart


    Tony Burns

    Dear Kniles,
    1. We do NOT “set” control limits as you suggest ! Control limits are calculated from process data. Control charts are “The Voice of the Process”. They tell us what is happening in our processes.
    The only setting of limits that can be done is with specification limits. Specification limits are “The Voice of the Customer”. They are what the customer has specified. Hopefully, they reflect the customer’s needs and not something that we want to impose on long suffering customers, in order to reduce our costs.
    2. It does not surprise me that you have processes with drifting means, that are “questing entropy”. This is a common situation. This is (hopefully) the reason that companies introduce quality programs in the first place !  Processes should be managed rather than being victims of chaos.
    3. You suggest “…the effort it would take to understand and manage all the control variables “. Control charts are not an academic exercise. The very purpose of control charting is to gain insight into the process ! It is for this reason that it does not matter that a process does not have to be brought into control before calculating limits, nor does it matter if data is autocorrelated. The aim is insight so that we can work on reducing variation. If we wish to sit back and let “noise based natural occurrence” take control, we might as well go back to the bad old days of Process Control and simply count defects at the end of the line and rework them or throw them away.
    It is worth emphasising the often forgotten basic:
    World Class Quality = On Target with Minimum Variation.
    I strongly suggest reading Wheeler “Understanding SPC”
    Dr Tony Burns
    [email protected]


    Kim Niles

    Thanks Dr. Burns:
    I believe we are thinking on parallel paths, just offset a bit from one another due to our vantage points (context). 
    With regard to your number one, I should have chosen a different word other than “set” as I meant what you stated.  My mistake. 
    With regard to your number two, again I agree but add that it all depends upon the context of the situation.  What does implementing a quality program really mean?  One person, a team, 50 QE’s, etc.?  How many significant control variables are we talking about?  How much drift is there?  How much does it cost to control that drift relative to profits made with the drift “as is” (including soft costs)?
    With regard to your number three, again I agree relative to using SPC charts to monitor a process.  However, after considering different context as outlined in the paragraph above, my statement of “…the effort it would take to understand and manage all the control variables” can have a lot of different connotations.
    Along these lines, regarding your statement of “World Class Quality = On Target with Minimum Variation”, that’s a great general statement that also depends upon the context surrounding that statement.  What does “minimum variation” really mean?  How much does it cost to reach that “minimum variation” or  “World Class Quality” relative to long term profits made (including soft costs)?  Of course obtaining “World Class Quality” doesn’t mean that all other aspects of business performance are compromised.  In fact, they all must work together in optimized harmony to obtain “World Class Quality”.
    Does this help?
    Sincerely,KN –,


    Tony Burns

    You asked: “What does “minimum variation” really mean? ” in relation to my statement that:
    World Class Quality = On Target with Minimum Variation.
    In relation to variable data, good quality means much more than zero defects or fewer defects. A defect occurs when a product is outside specification limits. The number of defects depends on where the specification limits have been set. I have read statements by consultants boasting how they reduced defects in companies. This is easy, just change the spec !
    Reducing variation is more difficult. It involves working on special then common causes. Only by focussing on variation and keeping the product on target, can the process be improved.
    Dr Tony Burns
    [email protected]


    John Hickey

    Just a slight modification to your equation:
    World Class Quality = On Target with Minimum Variation= On Target with Maximum Good Information about the Process.
    Example; An np chart with a favorable downward trend as a result of Good Quality Improvement Efforts is “Technically Out of Control” and contains more Good Information than the Chart that is in Statistical Control with no assignable causes. When the new spec limits are set and a new equilibrium is reached, the Long Term Shift value changes because it is tolerance dependent. Hence Quality Improvement Efforts require variable instead of constant(1.5 sigma) shifts with the effort being limited by a suitable loss function.
    -John Hickey


    Tony Burns

    Your name rings a bell … do I know you?
    In relation to trends, it’s worth reading Wheeler “Advanced Topics in SPC” page 136 to 137. Despite the usually taught stuff on trends, Wheeler suggests they add to false alarms and that it’s better to rely on the limits. Forget about constant corrections … fiddling with the controls is what operators commonly do … it doesn’t help. If your mean is shifting by 1.5 sigma or anything else, your process is out of control. Look for causes.
    Best regards,
    Dr Tony Burns
    [email protected]


    John Hickey

    With respect to my name, I hope it’s not from the Irish Limericks that I had published as a result of a bet. I seem to remember a Father Burns and several grammar school classmates with that name. I have your email address and in Mark Twain fashion can share in the telling of some funny stories and remembrances.
    Seriously, I am a fairly frequent participant in the discussion forum and I have contributed two Articles to the site*. Thanks for the tip on Wheeler. It is a habit of mine to take nothing for granted and to think things out for myself which comes from reading Richard Feynman. Trends or more specifically patterns of variation on charts could also be indicative of Chaotic non-linear influences and result in the onslaugth of sudden out of spec conditions. The variance improvement/ process adjustments that I am familiar with occur during the new/modified design validation phases for a process according to specific protocols. This usually leads to long stable processes when the process specs are followed. My interest in the 1.5 etc. Long term random shifts arose from my scientific curosity with regards to it’s origin and it resulted in my posted critique on this subject. I currently do not use this correction for process calculations.
    * A Simple Model of a Variance Stable Process                            *Sample Economy and Process Capability Decisions
    Best Regards
    John Hickey


    Tony Burns

    Trends are one of the 4 detection rules in the Western Electric Handbook. They are indicators that a process may be about to produce out of spec product (even if spec limits have been set at 6 sigma ! ).
    Shewhart only used Rule 1 : Lack of control is indicated when a point falls outside outside a control limit.
    Wheeler recommends Rule 1. He suggests Rule 4 may be added when personnel become more sophisticated. (Rule 4: A lack of control is indicated when 8 successive points fall on the same side of the center line.)
    I had a look for your critique of the claimed +/-1.5 sigma but could not locate it. Was it on the forum?
    Dr Tony Burns
    [email protected]


    John Hickey

    Re: 1.5 Sigma shift critique
    Same subject but an earlier message Click on the word attachment.
    John Hickey


    Tony Burns

    Would you mind posting your document as an RTF please. My virus checker detected a malicious script. DOC’s can have attached viruses. RTF’s can’t.
    Dr Tony Burns



    Now this thread become valueable for readers.
    I thinks all of us will  appreciate it


    John Hickey

    I am glad that you found useful technical information in this thread. In my opinion is one of the best Statistical Engineering Websites in the World.
    Best Regards,
    John Hickey

Viewing 58 posts - 1 through 58 (of 58 total)

The forum ‘General’ is closed to new topics and replies.