iSixSigma

Mike Archer

Forum Replies Created

Forum Replies Created

Viewing 25 posts - 1 through 25 (of 25 total)
  • Author
    Posts
  • #171945

    Mike Archer
    Participant

    Hello.  Thanks to everyone so far for the responses.
    Daves – For the dice, I am creating random integers from 1 to 6.  I create a column for each die, then add the columns together for the total roll.
    What I am only just now realizing though, is that the p-value for normality becomes smaller for all normal data as the sample size increases.  So data that is measured to even 100 thousandth decimal, will eventiall get a p-value less than .05 no matter how normal it is.  So as the sample size increases, the confidence in the mean and SD increases, but the confidence in p-value decreases.  Makes me wonder how much data is too much when testing for normality?  30 works but is 100 too much??
    Mike

    0
    #171938

    Mike Archer
    Participant

    Actually, the more data that is collected, the lower the p-value gets.  I found that to be true even for randomly generated numbers with lots of decimal places.  So I guess that more data is not always better for testing normality?
    Mike

    0
    #160456

    Mike Archer
    Participant

    Any help would be greatly appreciated… even opinions.
    Thanks,
    Mike

    0
    #155656

    Mike Archer
    Participant

    A statician’s wife told him that he was just average.  He replied to her that she was just mean.

    0
    #155651

    Mike Archer
    Participant

    Hello.  I’m new in training, but I did learn that a 99% confidence interval is needed for any hypothisis test related to medicine or health.  That equates to a p-value of 0.01 or less to reject the null.
    Mike

    0
    #155650

    Mike Archer
    Participant

    Thanks to everyone for all of the replies.  I found every one of them to be very helpful.
    There have been 2 GB projects closed on this very issue (both of them claim victory).  The last thing they need is a failed BB project to damage the credibility of our SS organization.  So I want to make sure that it is done exactly correctly from the very first step.
    I have 2 more questions:  Has anyone seen a Lean roadmap that has phases, tools, and deliverables?  Also, I suspect that scheduling and consumption logistics will turn out to be the main cause of the overproduction (not so much variation in the build process).  If that is the case, what is typically the metric and defectLead time, inventory level, something like that?
    Thanks again to everyone,
    Mike

    0
    #152027

    Mike Archer
    Participant

    Name of The Game – Thanks for the numbers.  The data cubes are a different Mike.
    Thanks again.
    Mike

    0
    #88901

    Mike Archer
    Participant

    DaveG,
    Thanks for those comments.  I needed to feel re-assurance from an outside party that we are using SS correctly AND for the right reasons.  Your comments have helped me feel more confident that we are using it correctly, I am still lacking confidence in the latter.
    You wanted me to explain my concern on $$.  It is hard for me to tell if our SS program is driven by a sincere desire to make positive changes, or if we are eager to justify our projects to management and to enhance our visibility.  Does the drive matter that much, and more importantly should it matter to me, or should I just worry about myself and my project?  By the way, my deep concern for this issue comes from reading posts on this forum (apparently from “true change agents”) who sometimes warn people to examine drives and so forth.  At what level should I concern myself with this, and at what level should I act?
    Mike

    0
    #88653

    Mike Archer
    Participant

    Gabriel,
    I think that I was pretty carefull to minimize the Hawthorn effect by instructing the participants to not change anything about the way they do their business.  Also, the Inspectors should not have been aware that data collection was taking place.  A Hawthorn effect should have indicated opposite results since most participants are biased in believing inspection wait time is a problem.
    What do you think?
    Mike

    0
    #88518

    Mike Archer
    Participant

    Dave,
    Could I get a faster result by using a benchmarking technique?  Maybe I could research a company that uses a similar system and knows a value figure to use.  How would I go about doing something like that?
    TIA
    Mike

    0
    #88404

    Mike Archer
    Participant

    Dave,
    I should state this more clearly:  The results of quality inspections are not being recorded.  My GB project is an initiative to begin doing so.  How valuable will this data be in dollars to the company?  This where I need assistance:  I cannot conceive a way to quantify the value of the data or the pain of not having the data.
    TIA
    Mike

    0
    #88399

    Mike Archer
    Participant

    Jeff,
    Okay.. maybe I am pretending to be a little bit of an expert.  I just wanted to add one extra thought.  If you assumed that your data set was a population and not subject to hypothis testing, then you would have to say that the difference between 119.0% and 119.01% is significant.  I say proceed with hypothisis testing.
    Mike

    0
    #88397

    Mike Archer
    Participant

    Jeff,
    I’m not going to pretend to be an expert on the subject, but in my mind, you are always working with a sample in inferential statistics no matter how inclusive it is.  If I shot the catapult 100 times and looked at that data, is that the population?  Not really – the inferences would represent an assumption that I continue to shoot the catapult without changing the process.  I’m only a GB, so you may have just got a stupid answer to your question :-0
    Mike

    0
    #88299

    Mike Archer
    Participant

    Stan,
    Would you say that a person’s drives should be the number 1 thing that they consider prior to committing to BB training (isn’t one’s abilities almost equally important)?  And how well should their personality line up with the profile you describe as a “true change agent” (partially, mostly, completely)?
    I tend to be a perfectionist, and my quality of work really suffers if I have to divide my attention too much.  You pointed out that having a need to be dominating and feeling impatient can be positive characteristics.  Good news for me, those have been character flaws of mine in most of life’s situations, but I do hate distraction.  I still feel like Six Sigma is a direction I want to go and that I could realize some success in it.  Your thoughts please.
    Mike

    0
    #88275

    Mike Archer
    Participant

    Vinod,
    I have asked similar questions in the past.  It seems that hoping for Six Sigma to enhance your career is a real sore spot with many of the BBs.  From some of the posts I have read, you would think that wanting Six Sigma certification is a terrible sin.  Why the strong emphasis on motive and what seems to be emotional sensitivity to the subject?  I wish I knew.  I predict that some of the replies you get to this post will scold you for having a strong personal interest in Six Sigma.
    Mike

    0
    #88266

    Mike Archer
    Participant

    Stan,
    In the short time that I have been reading the posts in this forum, I have noticed that there is a lot of attention paid to discussing people’s motives in SS training.  I have been through GB training and would like to explore the possibility of BB training in the future.  I don’t want to become a BB that falls flat on my face, so could you help me evaluate my motives?… or any seasoned BB is asked to reply to this message.
    I am interested in Six Sigma for several reasons:  I would like the certification for my resume (I will admit that I want to enjoy a nice career).  I also enjoy problem solving.  I want to see SS work for my company (I want the company to enjoy success).  I have a fascination with the way the tooling works (I think that it is just plain cool).  I believe that I can be valuable to the program (but I’m not sure).  I think that I am capable of learning and using the science (however, statistics is no where close to being a passion for me).
    In your experience, am I someone who will do well embarking on a career that SS is an important part of, or will it be a disastrous path for me.  I seriously want to know.  Feel free to ask more questions about me.
    TIA
    Mike

    0
    #88251

    Mike Archer
    Participant

    Robert,
    Thank you for that answer.  I have one question for you:  If you saw a histogram from a large sample that had a perfect bell shape, and had perfect symmetry (I mean, not even one data point is not perfectly mirrored by another data point.), could you unequivocally reject the sample for being biased or even fabricated?  Based on your evaluation of Burt’s case, I suspect your answer is no.
    P.S.  I know I just made myself into a liar by posting on this thread again.  Oh well.
    TIA
    Mike

    0
    #88250

    Mike Archer
    Participant

    binomi,
    I don’t mean to be so defensive (happens to be a character flaw of mine).  I MIS-interpreted your reply as abrasive.
    There is no real life example that caused me to ask this question, so I should probably leave it alone.  This will be my last attempt to explain my concern with what I perceive to be a flaw in statistical analysis (bear in mind, I am only a GB student with virtually no real life experience).
    Let’s say someone claims that a certain brand of shoe improves a certain track event time by some percent, and they have data to prove it.  Well, maybe he removed some data points that don’t favor his claim, then he has to remove other data points to regain normality.  I would want to prove or disprove that his sample is actually random by comparing it’s normality with the normality of samples from other processes known to be random.  Can I suspect a biased sample selection if a histogram is too close to a perfect bell shape, or too symmetrical, or a normality plot is too straight, or a P value is too high?
    My question:  Is there a way to look at the sample and say, “Although the sample has normality, it is not consistent with a typical random sample and therefore has a high probability of being biased.”?
    I am now done with this thread (whatever the answer is).
    TIA
    Mike
     

    0
    #88216

    Mike Archer
    Participant

    I was not aware that my messaging had a poor content/word ratio, so thank you for pointing that out.  I hope that you can excuse me for having a statistical aptitude that is inferior to yours.  I guess that there is such a thing as a dumb question after all.
    My appologies,
    Mike

    0
    #88172

    Mike Archer
    Participant

    So there is no standard rule.. only gut feeling?
    If I have a P value of .05 or greater, that tells me that I have a normal distribution, but is there a similiar gauge to tell me when a distribution is so normal that it violates the random nature of reality?  If you had a P value of 1.0 for a large sample of people’s age distribution, wouldn’t you HAVE to conclude that the sample was deliberately aranged by age?
    Mike

    0
    #87365

    Mike Archer
    Participant

    Ryan,
    Because I am new in SS, my ignorance in the methodology may come across as bad intent and may be misinterpreted as wanting to do things half-heartedly or taking hasty shortcuts.  I believe in SS and want to do things right, so I want to tell you how much I appreciate your candid comments.
    Can you give me some idea of a roadmap to follow so that I don’t un-implement the (incomplete) system that I and others have worked so hard on, and still come out with a Six Sigma improvement?
    The current call system is paper and optional.  Not all inspectors use a call sheet (other than the “yo” system).  I know that there is a lot of financial pain that the improvement will touch.  Not just the time of the person waiting on the inspector, but other people waiting on the unit to move and then there’s the inspector’s time and ability or inability to prioritize what is more important. Lastly, we will be utilizing technology that we already own by collecting the results of each inspection and storing them in a database.
    There is only one (newly trained) BB at our facility and 24 GBs.  I’m not sure he is equipped to support all the GB projects that have been started.
    I hope I didn’t throw too much at you, but I really need some expert input.  I know if I could see the light to start, I could run the race well.
    Thanks again,
    Mike

    0
    #87364

    Mike Archer
    Participant

    Ryan,
    Because I am new in SS, my ignorance in the methodology may come across as bad intent and may be misinterpreted as wanting to do things half-heartedly or taking hasty shortcuts.  I believe in SS and want to do things right, so I want to tell you how much I appreciate your candid comments.
    Can you give me some idea of a roadmap to follow so that I don’t un-implement the (incomplete) system that I and others have worked so hard on, and still come out with a Six Sigma improvement?
    The current call system is paper and optional.  Not all inspectors use a call sheet (other than the “yo” system).  I know that there is a lot of financial pain that the improvement will touch.  Not just the time of the person waiting on the inspector, but other people waiting on the unit to move and then there’s the inspector’s time and ability or inability to prioritize what is more important. Lastly, we will be utilizing technology that we already own by collecting the results of each inspection and storing them in a database.
    There is only one (newly trained) BB at our facility and 24 GBs.  I’m not sure he is equipped to support all the GB projects that have been started.
    I hope I didn’t throw too much at you, but I really need some expert input.  I know if I could see the light to start, I could run the race well.
    Thanks again,
    Mike

    0
    #87362

    Mike Archer
    Participant

    In my training we did a gage R&R by blowing up balloons and measuring them, and by inspecting the “m” on m&m’s (let the participants determine the characteristic they inspect).  We had to lean a lego race car assembly plant.  We also had to tell a joke if we were late.
    Mike

    0
    #87360

    Mike Archer
    Participant

    Thank you Ryan.
    I need to clarify my situation.  I elected to become a GB candidate primarly because I have been working on this electronic call system for some time and have partially implemented it already, and I wanted to apply SS methodology to the project.  So the ease of completion for me will be relatively nice.  I don’t think that CTQ identification is real crucial for my project, I see my challange as quantifying the pain of not having the system.  Any advise is appreciated.
    TIA
    Mike

    0
    #87317

    Mike Archer
    Participant

    Brian,Michael Carver advised that I work my project as DMAIC – post # 29296 (the thread disconnect is my fault).I understand that examples could close my mind, but because this is my first project, I would hope that an example could also spark some understanding.  Nothing against my BB, but I fear that if he is my only resource, I will be closing myself off from other potentially very valuable influences for the direction I head with my project.Thanks,Mike

    0
Viewing 25 posts - 1 through 25 (of 25 total)