iSixSigma

Lomax

Forum Replies Created

Forum Replies Created

Viewing 94 posts - 1 through 94 (of 94 total)
  • Author
    Posts
  • #186560

    Lomax
    Participant

    Lata,If you are looking at online training, then why look within India. Online = worldwideHaving said that, out of my practical experience, online training doesnt work for something like six sigma. Its like learning driving online….it just doesnt work.Thanks…neil
    [email protected]

    0
    #186537

    Lomax
    Participant

    GauravI m not sure if you have got yourself certified. Nevertheless, if you need to any guidance, please write to me.Thanks..neil
    [email protected]

    0
    #186536

    Lomax
    Participant

    SwathiIt is very difficult to say which is the best institute, as there are several parameters to consider.If you can share what you are looking for, then I can advise you.Thanks…neil
    [email protected]

    0
    #186535

    Lomax
    Participant

    ShivkumarIn short, projects exists in these areas and they have been successful too.There are projects in Training, Reward & Recognition including optimizing actual payouts which impacts the company’s budgets and several other areas.I will able share my experiences, please do mail me.Thanks…neil
    [email protected]
    Canopus Business Management Group

    0
    #182385

    Lomax
    Participant

    CT,
    You are correct in that the 6 sigma methodolgy reflects how much of the curve fits between the customer specifications.  Standard deviation, however is a measure of the spread of data as it pulls away from the mean and though it can be partially reflected in the shape of the curve does not directly relate to its shape.  It can still be bi-modal, favor one side or the other (poisson), etc without being an effect of the standard deviation.

    0
    #182384

    Lomax
    Participant

    Six Sigma is 3 standard deviations on either side of the mean (total of 6 sd spread) give or take the motorola shift.
    reflects the data inside the specifications, all data is under the curve.  The curve is the data

    0
    #182383

    Lomax
    Participant

    ASQ might disagree with your statement that there is no standard for certification since they are the most recognized body for certifying six sigma professionals.  Granted, there are other sources, but the reputable ones still work through the ASQ model.  Either way, saying that there is no basis to  establish who is and is not certified six sigma professional is like saying you dont know who is an MBA or not, or who has a bachelors degree or not.  There are tons of studies that show wage seperation based on these credentials, what makes it any different finding the same information related to people owning six sigma credentials?

    0
    #182133

    Lomax
    Participant

    This doesnt seem to fit right for me.  I would think the defintions are flip flopped.  From my understanding
    Z-score is a measurement of Standard Deviations from the Mean
    Standard Deviations are a measurement
    Sigma is a designation of the spread.
    If Sigma is the measure from the mean, then a 6 sigma process actually has 12 standard deviations from the lower limits to the upper limits?  (sigma as a measurement from the mean, so 6 on each side).  Sigma cant be a measurement from the mean or we have a general terminology problem.
    I though that sigma designated the entire spread, the Z-score does indeed designate the number of std deviations from the mean and measurement intervals are referred to in standard deviations.
    A Six Sigma process has a Z-score of 3 which is 3 standard deviations from the mean on either side.

    0
    #178594

    Lomax
    Participant

    Im also looking for any video clips on lean
    [email protected]

    0
    #165914

    Lomax
    Participant

    Thank you all.

    0
    #165841

    Lomax
    Participant

    Depending upon the project scope. It might take up to 6 months in some cases.

    0
    #165840

    Lomax
    Participant

    Thank you very much. Is this Minitab which is the statistical software? I haven’t used minitab before. Do we need any training for minitab? Any idea how much the software costs?

    0
    #160577

    Lomax
    Participant

    I use a Project Scoring for any Proposed Projects and Pipeline C&E scoring. Very basic but useful especially when the people proposig a Project are not Six Sigma trained – e.g. estimate savigns from Project, expected chance of success, timescale of Project etc.
    Give me your e-mail and I will send you my Charter which has the scoring within it. It sohuld hep you.

    0
    #151401

    Lomax
    Participant

    The software that will help is MINITAB. Best of luck.
     

    0
    #147241

    Lomax
    Participant

    Let me put it simply – the guy responsible for SS was not a statistician and he got the basics all screwed up. Much of what he claims is not only wrong, it is fraudulent – he did it simply to make heaps of money.
    If you are interested in quality read books by Shewhart and Deming.

    0
    #141036

    Lomax
    Participant

    Neal
    As a Black Belt for an air-conditioning company I have recently done a project where we proved via a DOE on a chiller how much our computer system was saving the company (and by changing certain settings how much more it could save).
    I can discuss more if needed.

    0
    #57016

    Lomax
    Participant

    Would you also post or send an example using Minitab?
    Thanks,Neil

    0
    #139652

    Lomax
    Participant
    #134515

    Lomax
    Participant

    The questions and replies on SPC and what seems to be managements desire to take the “short cheap route” is one that seems to be in ever small to mid size company (my experience). The responses were good although it seems that while trying to convince management or the powers-that-be to invest in MINITAB or similar products often fall on deaf ears and tight wallets. Is there a simple training presentation or an “easy to read and understand” article that explains SPC, Control Charts, etc. on the SixSigma website that anyone is aware of? Or one elsewhere? This may help answer a lot of questions to the beginning stats people and/or QE’s. Feel free to contact me at [email protected] if anyone has any info and good luck EZ.
    Neil

    0
    #133717

    Lomax
    Participant

    Johann,
    Would you mind sending me an email of your Excel file? I’ve never used the X-S chart but it might come in handy. At the moment I am trying to purchase MINITAB and I have used it at my last company with great success but it may not be as easily accepted here. My email is [email protected]
    Thanks,Neil

    0
    #60742

    Lomax
    Participant

    Karim,
    I’m not too concerned with the Healthcare aspect as much as just the QFD in general. My work is in manufacturing/electronics and I’ve gone through BB training at my previous company and I’m trying to inspire the folks at my new company so any assistance would be a benefit. My email is : [email protected]
    Thanks,Neil

    0
    #60738

    Lomax
    Participant

    Karim,
    If you don’t mind, please send me some of your examples of QFD in healthcare.
    Thanks in advance,Neil

    0
    #124477

    Lomax
    Participant

    Folks,
    Thanks for your input. The problem is:
    1. The instruments are “home made”
    2. One is in a Cleanroom while the second is on the manufacturing floor.
    3. I was thinking of performing a paired “t-test” as many suggested.
    4. The “measurement” isn’t exactly what most people are used too (spring preload aka spring droop)
    I really can’t get rid of one of the instruments (not practical) and I really can’t use it on the production floor and then bring it into the cleanroom (again not practical). Having samples is not a problem and I really think the paired t-test will be the best approach.
    Thanks to all for the input,
    Neil

    0
    #123723

    Lomax
    Participant

    John,
    Is DFMA “Design For Manufacture Ability?” I’ve seen to accronyms for this and just want to make sure.Thanks,Neil

    0
    #123722

    Lomax
    Participant

    What? Darth is leaving? Where is he going………….I ask a simple question………get a QUICK response and then he leaves……what is the world coming to? I’m taking an early lunch.

    0
    #122431

    Lomax
    Participant

    I have used the analogy brainstorming technique in the trainings and in real life six sigma :
    For example if one wants to hire BBs… brainstorm on the analogies of going out for fishing, coin collection, treasure hunt and correlate back to back to the original problem of hiring.
    Triz is another example of De Bono’s approach of lateral thinking used in six sigma
    Basically De Bono’s approach contradicts six sigma thought process. De Bono says dont go analytical on solving problems… dont analyze to find the root cause. He says fix the problem with creative solutions by experimentation and verify them using Shanin techniques (implicitly told by De Bono)
     

    0
    #120498

    Lomax
    Participant

    Another simple way to understand dof is :
    If one were to break a stick into 3 pieces of different lengths… he/she would have only 2 chances to do that…(in order words the degree of freedom to determine the outcome(length of 3 pieces) is governed only by 2 chances

    0
    #120459

    Lomax
    Participant

    1) There are no norms… generally companies report the benefits as of control phase and extrapolate it for the year (financial or calendar depending on the business cycle
    2) In multi-stage if the benefits are substantial in each stage it is reported for each stage…else for ease it is reported only at the end
    3) Usually not (considering ease) but can be done
    An organization i knew is using a quarterly benefits accured as a result of productivity and quality benefits across the organization and reports it to the group
     
    neil

    0
    #120458

    Lomax
    Participant

    Avinash
    DMAIC methodology is used to solve problems in a structured way. By using DMAIC root causes of the problem will be killed or counteracted.
    If there is an opportunity to solve any problem you can do a project…
    The term project is used just to indicate that there is start & end point to this activity… it does not mean a study, report, a survey, etc., which is the commonly used for projects during grad & pgs
    BS7799 certification is more meeting standards. By not doing so there will be a compliance violation leading to risk.
    There is nothing common between recertification of BS7799 and DMAIC project. So suggest you should drop the idea of doing a project with recertification

    0
    #120449

    Lomax
    Participant

    statistical significance refers to the probability that your conclusions are right. Hence if there two things are statitically different, then it means that at a certain level of probability one could conclude that there are different. So it just means that any conclusion is not absolutely to right.. there is a probability of its correctness or in other words there is a risk of concluding that way.
    Hence confidence levels are assumed upfront rather than calculated (usually). All hypothesis tests like chi-square, 2t test, 1t test, anova are examples of identifying statistically significant difference between samples.
    It cannot be achieved thro six sigma directly..
    targets are usually defined by customers (directly or indirectly) hence we are set a target ourselves
    For more on these refer to terms like confidence level, level of significance, alpha risk in the glossary.

    0
    #120447

    Lomax
    Participant

    Thanks  PMG for your comments
    Why dont you attempt answering the query.. ??
    Cheers…neil

    0
    #120446

    Lomax
    Participant

    Safa
    Questionaires are structured or unstructured, objective or subjective, etc,
    Best option will be to use structured questionaire with objective questions.
    Design your questionnaire yourself.. have not more than 5 ~10 questions. Make your questions objective..Make it multiple choice – even choices.
    Use top-box analysis to analyze the response
    Have a separate session with 1 ~ 2 questions that require subjective answers that will provide you some more insight on the subject.
     
    Cheers..neil

    0
    #120444

    Lomax
    Participant

    1) Suggest that you should undergo the green belt trg before starting the project…. that helps
    2) If you have already undergone the trg..suggest you refresh the basics
    Wondering as I havent used any jardons in my msg… so not sure what you want me to elaborate
    Cheers…neil

    0
    #120442

    Lomax
    Participant

    1) focus on mapping the process end to end.
    2) After mapping explore quick hits to improve (dont implement but park the improvements)
    3) Measure the TAT end to end with break up of different stages by bands
    4) Identify areas were there is a large delay/waiting
    5) Check what is different with data points (recruitments) in different quartiles of the data set…. why a particular recruitment takes lesser time than others
    6) Steps 3), 4), 5) will help in identify the probable causes
    Then it will the usual improve and control phase with pilots, hypothesis testing to prove improvements, etc.,, Ensure int he improve phase to implement quick hits
    Cheers..neil
     

    0
    #119314

    Lomax
    Participant

    I agree with Tronan.
    Monetary Rewards are good.. but they should be fewer and should be given for really worthy employees.
    Often projects dont take-off as it comes as a mandate from the management… challenge really is to convert the people on the floor to understand how this project will benefit them….
    One can circulate the list of benefits for employees of each level once the project is completed… for example.. increasing productivity means $$ for the top management… means efficient management & flexibility to utilize resource elsewhere for middle management & reduced unnecessary work which means a few days early at home for the lower management..
    Its all in how you SELL it… I would nt be ashamed to use this word for six sigma… after all it is good for the organisation.. so never mind … walk into your office tomorrow imagining that you are a going to Sell six sigma rather than implement six sigma… & you will see the fruits
    Thanks..neil

    0
    #116448

    Lomax
    Participant

    Hi,
    This is definitely a very normal practice but two things:
    1. Floor level metrics are displayed team wise and not at an individual associate level.
    2. Agent metrics are displayed at the team desk near the TL.
    Thanks

    0
    #116446

    Lomax
    Participant

    Hi Craig,
    1. If the big Y is data Quality which definitely is an output but is also a project deligh parameter, a defect will be any deviation from the customer expectation in terms of data quality. For ex: If accuracy of data is one CTQ, defect can be measured as financial aerrors, statistical error, or typo error.
    2. Craig,  I also feel that this is a DMADV project if we have the task at hand to set up the data capture mechanism / putting up the data quality reporting system. It will become a DMAIC project if the problem at hand is with the quality (accuracy, % right etc.) of data is not meeting requirements. And then setting up the right data capture mechanism can become a step in M phase (by strengthening the measurement sytem).
    Thanks
    Neil

    0
    #115779

    Lomax
    Participant

    To clarify.  In a value stream map, process boxes do include cycle time data that is gathered through performing timed random observations. 
    However, by constructing a higher level value stream map of the “flow” of material from one function to another, and by physically counting the amount of work in progress (WIP) between each function one can arrive at a Total cycle time and then cycle efficiency
    Total cycle time= (WIP/Exits) x Available hours per day less breaks
    Process cycle time & Total cycle time of a product value stream are different. Total cycle time evaluates the life of a product & identifies/quantifies bottlenecks.
    Evaluating where bottlenecks are between processes can allow you to either add or remove operators and balance work eliminating overburden and creating flow.  Generally, lean solutions result in greater productivity & efficiency.
    efficiency = Value added Criticalpath process cycle time in minutes / total cycle time in minutes  x 100
    Sorry for the lengthy post.
     

    0
    #115228

    Lomax
    Participant

    Saw some “key” words such as “Kaizen” and “Value added time” in the last post.  If you are looking at chronological process steps performed by different individuals you may consider a constructing a Value Stream Map.  This should display process steps (Cycle time for each & number of operators), Work in Progress as well as being a good high level characterization of the “flow” of work (and where it hangs up).
    Just something to consider.
     

    0
    #113809

    Lomax
    Participant

    Hi
    I would also like a copy of the Best Format for six sigma final report
    Thanks
    Neil

    0
    #106605

    Lomax
    Participant

    Hi everybody,
    Personally, I use the following for test between two  samples:
    Comparison of Means::: If sample size <30 – t test . else. z-test
    Comparison of variances :: F Test
    These are based on statistical equations and are quite robust. I have made them into easily usable excel file – But do not know how to attach it here. If anybody is interested, can write to me at [email protected] – provided of course, you promise to help me in future if I have a technical problem and you are in a position to help ;-)
    Neil
     

    0
    #106604

    Lomax
    Participant

    Hi Everybody. This is my first post. Correct me if I am wrong!
    The p value is merely a measure. In very simple terms – it would mean that, you are accepting or rejecting the hypothesis at (1-p)% confidence level.
    E.g. If the p value is between 0.01 and 0.05, it would mean that the decision has a confidence between 95% and 99%. This is the beauty of p value. Now, whether to go for p of 0.01,0.05,0.1,or any other value for that matter, will depend on the severity of the decision and robustness of the sample. A call on it has to be made by the project leader and project owner in cosultation with the affected parties. There are certain “thumb rules”, But, of course who knows your process better than yourself!!!

    0
    #97253

    Lomax
    Participant

    Melissa,
    The industry figure you are referring to comes from a study done by Dun & Bradstreet.  They are quoted in the following article. I’ve attached the link for your review.  D&B stated that customer data decays at an annual rate >35%, which would equate to ~3% per month.  You might be able to contact D&B directly to get a copy of the report for their study.
    Neil
    http://www.findarticles.com/cf_dls/m0FOX/4_4/54299235/p1/article.jhtml

    0
    #97051

    Lomax
    Participant

    Kelvin,
    I have been working with a direct marketing firm in Atlanta, GA for over a year now implementing Six Sigma.  I will be glad to give you more information.  Feel free to email me at [email protected].
    Neil

    0
    #96974

    Lomax
    Participant

    Ooops…meant Faceman instead of Statman…sorry about that Face!

    0
    #96972

    Lomax
    Participant

    Lisamarie,
    Don’t use the “randbetween” function in column B.  Like Statman and Gabriel said above, use just the “=rand()” function in column B and fill it in for all 1100 rows.  Then, the formula in Column C that you have should work.
    Neil

    0
    #96582

    Lomax
    Participant

    Anonymous,
    I am assuming you are tracking continuous data if you are considering use of a X-bar/R chart.  Given that assumption, if your sample size varies from one time increment to another, then use an X-bar/S chart.  “S” is the sample standard deviation.
    If you are tracking discrete, or attribute data, you will need to use a U-chart or a P-chart, depending on whether you are tracking “Defects” or “Defectives” respectively. 
    If you have any other questions, let me know.
    Neil

    0
    #95753

    Lomax
    Participant

    Jb
    I really don’t know which parameter you are talking about.
    However having a LSL of Zero and no USL is supraising. If the parameter is something like run-outs, concentricity, etc in manufacturing or something like Delays in service industry then in the ideal world one would expect Zero… but there should be allowable limit for variation…. So you need to find out the USL for the process.. check up with process owners or competitors database.
    Check exponentially of data by :
    #1 Is the Mean & Std Dev are equal or nearly equal (ensure sample size)…. This is a thumbrule for a exponential process
    #2 Also try fitting to a distribution using minitab options.
    Control charts do not assume for normality…. but it is not advisable to plot this data on Xbar-R or I mR chart.
    Convert the data into discrete… Defective & non-defective and then plot them in appropriate discrete control chart

    0
    #89904

    Lomax
    Participant

    Lidia,
    I have been working with a direct mail marketing firm for over a year now on projects similiar to yours.  If you have not considered it as of yet, a conjoint analysis may be what you’re looking for.
    The basis for a conjoint analysis is that consumers often make trade-offs in their buying decisions.  A conjoint analysis allows you to determine “utility values” for different attributes (color, size, font, gift, etc.).  The format for a conjoint analysis is identical to a DOE.  You can segment the data by market for age groups, demographics, etc.  For example, we found certain font colors had higher utility values in certain age groups. 
    A simple internet search will give you plenty of reading material on conjoint analysis.  If you have any other specific questions, let me know.
    Good luck!
    Neil

    0
    #89791

    Lomax
    Participant

    Gabriel,
    I agree with your comments completely.  My assessment of Stephane’s situation was that she is only getting, on average, 20 data points per month.  I was trying to articulate that she should use all 20 data points and not “sample” from them.
    I would think that reporting on a quarterly basis with, on average, 60 data points could reflect a short-term Cpk, but obviously she should focus on the PpK to ensure long-term sustainment of any improvements she makes.  As you point out, the more data points she collects will improve the confidence of her capability index.
    Thanks for the comments.
    Neil

    0
    #89784

    Lomax
    Participant

    Stephane,
    Unless, I’m missing something here, sample size is not a concern for you.  Sampling of a “population” is done because typically one does not have access to all the data.  Thus, we take a representative “sample” of the data to infer information on how the “population” would behave.
    In your case, it appears that you have access to every Priority 1 data point, whether it is 6 or 20 in a month. If this is true, you should be able to calculate a process capability utilizing the total number of data points accumulated.  I’m assuming you are measuring a continuous variable realted to these Priortiy 1’s with upper/lower specs.
    There are two things you need to keep in mind though. First, “No Stability = No Capability.”  Place your data on an appropriate SPC chart to evaluate whether the system is “stable”, that is, only under common cause variation.  Given the relatively small number of data points and the singularity of the data, I would use an I-MR, individuals moving range chart.  Once you have confirmed stability, then you can feel comfortable with the capability indices.
    Secondly, a lot of people want to throw capability indices on situations that have no “practical signficance.” If you are only getting six Priority 1’s per month, what is the practical significance of this? If you’re averaging 20 per month, is there a sufficient business case to improving this process by focusing on the capability?
    I hope I’ve understood your problem correctly.  Let me know if I can be of any further assistance.
    Neil

    0
    #89504

    Lomax
    Participant

    Spencer,
    Because the product is low volume, and you are trying to establish a baseline, my advice would be to chart each of the 8 measurements separately on an I-MR chart.  This assumes you are measuring the same 8 CTQs on each part. 
    Thus, you will have 8 separate charts to maintain, but the successive data points will be independent.  In this manner, you will be able to detect any special cause variation from process changes on the individual measurements.  You may find a change only affects one measurement and not others.  Calculating a mean value from the 8 data points on each part could “mask” some process change effects.
    I have seen this technique used in several companies successfully.  Good luck!
    Neil

    0
    #89024

    Lomax
    Participant

    Tyson,
    I have been working with a direct mail marketing form for over a year now with much success.  This firm’s product is a mailer that markets and sells insurance products through a consumer’s bank.  One of the key metrics is response rate, the number of people who respond to the mailer and purchase insurance.  As expected, there is quite a bit of “noise” in this process.  We have used conjoint analysis and DOE to improve response rates.  For example, factors in a recent DOE were: 
    1.  Introductory letter.  One letter was more “bold” in that created a sense of urgency to respond.  The other letter did not indicate a “must reply by” date.
    2.  Offer. Two different options of free gifts for responding.
    3.  Benefit.  Two different levels of insurance benefit were offered.
    We will continue to utilize Six Sigma tools to improve this company’s marketing processes.  They have a wide client base which offers unique opportunities, and challenges, to the use of statistical tools.
    Neil

    0
    #88611

    Lomax
    Participant

    JM,
    The mean is the measure of central tendencyVariation is a measure of the dispersion of the data from the mean.
    Neil

    0
    #87824

    Lomax
    Participant

    Alex,
    Cp, and Cpk calculations utilize the standard deviation to estimate the process capability with the assumption the data is normally distributed.  The Cp formula denominator is 6*sigma.  If the data is normally distributed from a stable (in statistical control) process, we would expect 99.73% of the data points to lie within +/- 3 sigmas(total of 6 sigmas) from the mean.  A process that is not stable (i.e., has special causes)has an unpredictable amount of variation and thus violates the 99.73% assumption.   Therefore, we can’t trust the capability indices from this data.

    0
    #86941

    Lomax
    Participant

    Diane,
    The first step your company should do is to have the Executives attend an Executive Overview training. This is typically 1-2 days.  Following this, your Executives should understand the importance of returning to your company and developing a Six Sigma strategy.  The strategy should outline the goals and objectives, cover the Six Sigma organization (Deployment Champion, MBB support or consultants, Initial Project Champions, etc.), Project Selection/tracking/closure process, Candidate selection process, reward/recognition process, and initial timelines.  I would advise getting an MBB consultant to assist your Executives in completing this strategic plan.  The next step is for the Project Champions (plant/dept. managers, process managers, etc.) to receive Champions training.  This is typically 3 days.  Following this, your Champions and Executives need to sit down (once again, with an experienced MBB) and identify the first round of projects based on the Six Sigma strategy and the training the Champions will receive.  Once you have identified the projects and the associated complexity, you can decide on how many BBs and GBs you will need to complete the projects.  Always pick the projects first and the people second.  Most likely, you will need to prioritize the projects and develop “waves” of belt candidates to attend training over the next year.  The belt candidates will be assigned a project that will act as their “training” project during BB/GB training.  Towards the end of the first year, you will want the MBB/consultant to provide train-the-trainer to your BBs to allow them to act as GB trainers in the future.
    The most important thing you can do at this point is build the Executive and Champion support/knowledge, along with the foundations mentioned above before starting belt training.  Hire an experienced MBB or consultant to get you through the first year.  
    Congratulations on the decision to proceed with Six Sigma…Good Luck!
    Neil 

    0
    #86842

    Lomax
    Participant

    Shelley,
    My first question to you would be, “How do you know that your customers are “satisfied” with the quality and length of current calls?”  Secondly, do your customer satisfaction metrics link back to process metrics?  For example, if shorter call length is a desirable customer metric, then a linked process metric may be call center agent availability measured by say, the average number of calls handled per shift.  This way, there is an incentive to reduce the call length time to simultaneously improve the number of calls an agent handles.  Obviously, quality measures should be in place to ensure the appropriate level of response to your customers.
    I’m sure you’ve heard the maxim, “You get what you measure!”  Most companies do a poor job of selecting metrics because they don’t understand the differences and linkages between “results” metrics, “process” metrics, and “input” metrics.  Thus, when I visit clients, I often hear, “I can’t improve this metric without hurting another one.” Depending on how it is measured, productivity is often the culprit of this dichotomy.  I’ve said all this because you may be in a similar situation with your internal measures.  Make sure you define “result” metrics first, then link them to “process” metrics which in turn, drive “input” metrics.  This “Balanced Scorecard” approach should clear up the external vs internal customer debate. 
    I apologize for the long-winded answer…
    Neil

    0
    #86774

    Lomax
    Participant

    Lewis,
    Is this your belt training project? No one in training should be assigned a project that does not have a baseline already established.  That doesn’t mean the measurement system is credible, but at least one is there.  Your first project is critical and should be focused on learning the tools of Six Sigma.  In addition, it sounds like this project is not very well scoped (“size and complexity of our company”).  You and your champion(?) should scope this project to one facility, or one department prior to starting the project.
    My thoughts..for what they’re worth…this project, due to complexity, scope, and lack of measurement system, is not suitable for someone embarking on their first six sigma project.  You would be wise to develop a better project charter and get this project down to manageable size or hand it off to an experienced Black Belt. 
    Neil

    0
    #86769

    Lomax
    Participant

    Shirls,
    First of all, I am an independant MBB contractor, but I am not giving this advice as self-promotion.  I have dealt with numerous companies in your exact situation.  It is very cost prohibitive for some companies to send multiple trainees to belt training.  It is much less expensive to pay per diem for a single trainer!  More importantly, an independant contractor (one who is knowledgeable and competent, obviously) should take the time to assess your company’s culture and modify his training materials to incorporate examples, case studies, etc. that are meaningful to your trainees.  If I’m teaching trainees from a metalworking facility, I don’t use injection molding examples.  Conversely, if I’m teaching marketing/financial trainees, I don’t use manufacturing material.  It is the trainer’s role to ensure that your trainees walk away with the ability to take the tools taught and apply them to your company’s processes.  Six Sigma is applicable to all kinds of processes, but it is absolutely critical that your trainees are able to bridge the gap between the training material and the processes they have been dealing with for 1,5,10, 20 years or more!
    Neil

    0
    #86735

    Lomax
    Participant

    John,You need to keep in mind that statistical significance depends on the sample size and sampling plan/methodology.  One can certainly “create” statistical significance by altering how the data is collected.  If your data is very noisy, then you obviously need a “large” sample size to feel relatively confident about any statistical tests you run on the data.  A p-value is the probabilty you would see a statistical result given the null hypothesis is true.  Thus, a small p-value sends you in the direction of rejecting the null.  In terms of alpha-values of 0.05, keep in mind what the real risk is:  Rejecting the null hypothesis when in fact it is true.  Typical  null hyptheses preclude “no difference.”  So, what is the risk (financial, customer, safety, etc.) to your process if you “take a special cause action” when you really shouldn’t have? 

    0
    #86634

    Lomax
    Participant

    Nadia,
    As the others above have stated, your instructor is oblivious to the true concept(s) of Six Sigma.  I am current helping a direct-mail marketing firm improve their processes with Six Sigma.  It is challenging and somewhat cutting edge in the direct mail world, but nonetheless achievable. 

    0
    #85962

    Lomax
    Participant

    Steve,
    You must fix the measurement system before proceeding forward.  Once you have less variability in the measurement system, you will most likely need to re-evaluate the baseline and possibly the entire project.  It never ceases to amaze me how many projects find their root causes in the measurement system, especially attribute.
    Good luck!

    0
    #85886

    Lomax
    Participant

    Fabiola,
    What is the nature of your Y?  Are you counting the number of DEFECTS or classifying as DEFECTIVE (0/1, Good/Bad, etc.)?  If you’re dealing with DEFECTS the poisson distribution must be used to transform the data.  If it is DEFECTIVEs, the binomial distribution must be used. 
    The point I’m concerned with is the same one raised by Michael in his previous post.  If your DPMO is “low,” the sample size must be quite large for the transformations to stabilize the variance.  Are you positive that there is enough “cost reduction” in further reducing the DPMO to warrant the amount of work to conduct this DOE and subsequent DOEs if necessary?? 

    0
    #85879

    Lomax
    Participant

    Fabiola,
    Are you referring to the Y (measured response) being discrete? If so, there are formulas developed to handle this situation.  However, I would have to agree with Michael…you should do everything in your power to use continuous or scalar data if possible. 
    Neil

    0
    #85877

    Lomax
    Participant

    Gavin,
    I agree with Zilgo…Why are you conducting normality tests in the first place? Typically, normality tests are performed to satisfy underlying assumptions of more powerful statistical tests.  What is the ultimate goal(s) you (or your Quality mgr.) is trying to achieve? You already mentioned the product is meeting specifications.  I would be concerned about your reference to control limits (SPC) compliance depending on the degree of abnormality in the data.
    Setting an alpha or beta value to compare the p-value too is nothing more than a line in the sand.  Where you draw that line depends on the risk level, financial impact, etc. of the decision towards some goal or objective. 
    Call me a purist…but every tool has an intended purpose and limitations. When you extend beyond the purpose or ignore the limitations, something or someone typically gets hurt… 

    0
    #85798

    Lomax
    Participant

    BG,
    Not every process improvement project is a Six Sigma project.  There are some things that are simply “Just Do it” projects.  A document control system, in my opinion, is a “Just Do it” project.  As a former Plant Manager and MBB, I would have heartburn directing a dedicated Six Sigma resource to this project.  It certainly should not take a Six Sigma GB or BB to develop and implement a document control system.  Obviously, it would be wise to include a GB/BB on the cross-functional team tasked with developing the document control system to ensure the system is streamlined and integrates well with other quality/six sigma initiatives.
    Just my two cents…

    0
    #85634

    Lomax
    Participant

    Jill,
    That is simply not possible…GB training should be at least 2 weeks, require a successful, completed project, examination, and MBB review.  BB training should be at least 4 weeks, 2-3 successful projects, examination, and review by MBB.  MBB training varies but typically includes an additional two weeks beyond BB, plus 6-8 months of training delivery, coaching, and business exposure.
    Neil

    0
    #85504

    Lomax
    Participant

    Craig,
    The first two weeks of training is the same for both Green and Black belt candidates.  Green Belts are done after two weeks, the Blacks go on for Weeks 3 and 4.  Week 3 is focused on DOE, while Week 4 is focused on advanced stat topics, culture change, etc.
     

    0
    #85394

    Lomax
    Participant

    Thank you Pam!

    0
    #85016

    Lomax
    Participant

    Vitto,
    I must concur with what “Helper” and others have said.  You seem to be on the right track.  Don’t feel pressured to use an Elephant Gun (DOE) to bring down a Rabbit.  Once you have baseline data for your process, use the VSM, Process Map, and Multi-vari studies to identify and quantify the mean/sigme in cycle times.  Eliminate or reduce as much of the non-value added time as possible and re-design the process accordingly.  Ensure everyone is well trained in the new process, monitor your cycle times (SPC?), calculate your statistically-valid sample size (I agree with Power >= .80) and run your t-tests to validate the change.  The key point that “Helper” made is that this change needs to be practically significant to the customer.
    Good luck…and keep us informed on the results!
    Neil

    0
    #84975

    Lomax
    Participant

    Vitto,
    Are you sure you’re ready to conduct a DOE? If you are measuring cycle time in a continuous data format, with discrete/attribute factors, you should be fine in terms of a DOE.  Have you identified these “critical” factors for your complaint process? If so, have you further defined them in terms of levels? 
    I’m sure you’re aware, but a DOE is a powerful tool that requires 75% planning and can be very expensive to conduct.  My advice would be to exhaust the process map tools, FMEA, value stream analysis tools, etc. before you launch into a DOE. 
    Just my two cents…
    Neil

    0
    #83210

    Lomax
    Participant

    Kathy,
    I was an IE for a tire manufacturing company and can certainly relate to your situation.  You are on the right track.  I’m assuming you have collected data from the OSHA 200 logs and injury reports?  I’m also assuming you have interviewed the injured associates?
    My only question/concern is can you accurately measure those “x’s”? Strains are typically an acute injury as opposed to a cumulative trauma.  Given the “instantaneous” nature, it may be difficult for you to measure some of the “x’s” in a passive data collection (multi-vari) study. If you remember from your training, Six Sigma is about taking a practical problem and turning it into a statistical problem. From here, you collect data and generate a statistical solution. At this point, you have to put your “engineering” hat back on and develop a practical solution.  My point for bringing this up is that the statistical tools of Six Sigma can only take you so far in solving your ergo problems.  Use stats to identify and statistically validate the “x’s”, provided you can measure them.  These key “x’s” may end up being inputs to other ergonomic analysis tools such as NIOSH action limit, energy expenditure models, etc.  These tools assess the RISK of an ergonomic injury.
    I hope this helps…
    Neil
     

    0
    #83197

    Lomax
    Participant

    Dear BPauquette,
    The key thing to keep in mind when selecting BB/GB candidates is that these people will be your CHANGE AGENTS! They must possess two things, Desire AND Readiness.  Desire is more qualitative…you want people who aren’t satisified with the status quo, who rely on facts to make decisions.  Their Readiness lies in their current skill sets (education, experience, etc.), the level of respect and cooperation they have in the workplace, etc. 
    If you view Six Sigma as both a Culture Change AND a Toolset, then select your candidates based on their ability to “Change the Culture” and apply the tools to the organization.
    Just my two cents…
    Neil

    0
    #83080

    Lomax
    Participant

    Pat,
    I don’t think there is, nor could be, an industry standard for a DMAIC project.  My “rule of thumb” is 3-4 months for a Green Belt project and 4-6 months for a Black Belt project.  This time includes project identification to control plan/training in place.
    Given the data intensity of a DMAIC project and the multi-month timelines, concurrent SS projects are necessary for a company to realize significant improvements as soon as possible.  The key here is that the projects and their potential “overlaps” with other processes are planned and coordinated to prevent wasted resources and sub-optimal improvements.
    Neil

    0
    #83037

    Lomax
    Participant

    Confused,
    I foresee several problems with the background scenario you outlined. 
    First of all, SS is a culture change, not just a new way to “identify/define/manage projects.”  Second, if your CEO is REALLY serious about SS, he needs help (via consultant, hired MBB, etc.) to do this right.  As others have mentioned, Executive and Champion’s training is a must.  Following this, I recommend an off-site meeting facilitated by an MBB or consultant to map out a SS strategy.  This strategy would identify process owners, project selection and candidate selection methodologies, financial justifications, etc.  A SS startup will require full-time MBB/BB support to be successful. 
    As to your specific question on projects:
    Project selection is a collaborative effort.  A “champion” may initiate the project to address a specific BUSINESS NEED, but he alone cannot and should not define the entire project.  A champion should provide the business case for the project…why is it important? the goal(s) and objectives?  THe champion should be able to allocate the resources necessary to complete the project.  The process owner can provide customer insight (CTQs) as well as input to the scope of the project.  The BB/GB should help put the performance measures in a form of Y=f(x).  The Belt should help the process owner and champion understand the criticality of how we measure the Y.  Also, the Belt should provide input as to the timeline for DMAIC.  The Belt can also work with the accountants to determine the financial gains of the project if the objectives are met. 
    Far too many companies do not spend enough time in project selection, definition, and chartering.  They want to see someone “doing SS” and not setting around in a conference discussing all these “trivial” details.  However, when the project scope creeps, the timeline slides, and you find out you’re measuring the wrong Y, you’ll wish you had taken the time upfront.  In my years of experience, most SS project failures can be traced back to a lack of effective project selection, definition, and chartering.  The tools work if you’ve done you’re “homework” first.
    Good Luck…and get some experienced help, quickly.
    Neil

    0
    #82811

    Lomax
    Participant

    Marvin,
    Most states have an Industrial Extension or Manufacturing extension program that is typically run through local universities, community colleges, etc.  You might be able to team up with one of their agents to conduct a SS project for one of their clients.  Good Luck.
    Neil

    0
    #82810

    Lomax
    Participant

    Rich,
    I helped start up and became an MBB in a $500MM division of a company.  The Senior Mgmt. mistakenly made the decision to include SS as “part of normal job duties.”  The fallacy here is that, for the first 3-5 years (could be much longer), you need 100% dedicated resources to integrate and institutionalize SS into the company culture.  There is way too much fire fighting in most companies today to launch a SS program with people who are partially dedicated.  It simply will never get off the ground. 
    Depending on your position in the company and other political factors, you may need to call a “time-out” and convince mgmt. to develop a long-term (5 year) SS strategy with dedicated resources (MBB and BB) to ensure the program’s success, or else it will die a slow, painful death.
    Neil

    0
    #82802

    Lomax
    Participant

    Rich,
    The issue is not the time, but whether your company really needs these individuals to be BBs.  If so, then their training should be geared towards that from the beginning.  I have seen too many companies “train” people to become BBs just to satisfy some pre-conceived quota and then not be able to utilize their new “agents” effectively due to a lack of projects, MBB support, and most importantly, a lack of management support/direction.
    If you are training BBs, am I to assume you are a MBB? If so, you should be more concerned about the mis-understanding of the Six Sigma philosophy on the part of your management than the time between belts.
    As you should well know, Six Sigma is predicated on data and facts.  So my advice would be to determine the critical business needs, identify associated Six Sigma projects to address these needs and let that determine how many BBs and GBs you need. 
    Just my two cents worth…
    Neil

    0
    #81980

    Lomax
    Participant

    Jackey,
    A few comments/questions for you:
    1.  Given your process, is 36 samples sufficient to characterize the capability?  How long did it take to gather 36 samples? In that amount of time, would the process have exhibited a characteristic amount of variation?
    2.  If the data is not normally distributed, you first have to ask why? Most processes do not exhibit normally distributed data.  You may simply have not collected enough samples for the process to exhibit normality.  Or, there may be some form of “special” cause variation occurring that is affecting your data.  If so, you just found an “x.”  It could be that your process will never exhibit normality.  But either way, you need to rule out sample size and special cause variation as potential factors to non-normality first.
    3.  In terms of “solving” the problem, I would collect more data first and see if it approaches normality.  If that doesn’t work, you may be able to subgroup the data and analyze the averages of the subgroups.  You may be able to transform the data with Box-Cox, et al.
    Good luck.
    Neil

    0
    #80650

    Lomax
    Participant

    Mike,
    Thank you for the link.
    Neil

    0
    #80449

    Lomax
    Participant

    Sue,
    In my humble opinion, the role of Six Sigma champion includes:
    1.  Foster the culture change to Six Sigma through their actions. Talk the talk, and walk the walk.
    2.  Project identification, scoping, and chartering.
    3.  Candidate selection
    4.  Ask the “right” questions through periodic project reviews.
    5.  Remove organizational barriers. 
    6.  Provide resources (time, people, money)
    7.  Drive project closure accountability.
    8.  Ensure execution of control plans.
    Have you provided Champion’s training to your organization? Do you have a Six Sigma strategy document that clearly outlines Sr. Mgmt., Champion, MBB, BB, and GB roles and responsibilities?
    Just a few of my thoughts…if you would like to discuss further, I will be glad to give you my email address.

    0
    #80242

    Lomax
    Participant

    Austin,
    ASQ has an office in downtown Milwaukee.  I’m sure you can find what you’re looking for there.  Also, you may want to check with GE Medical Systems in Milwaukee about any possible internships.

    0
    #75373

    Lomax
    Participant

    Brooke,
    If you create a dotplot of your data, the tallest column of dots will be your Mode.  You will find the dotplot under the “Graphs” menu.

    0
    #75240

    Lomax
    Participant

    Balpreet,
    First, the use of a questionnaire as a data gathering tool brings about several more questions.  Was the questionnaire designed to minimize bias? How were the 50 people chosen (Randomly? Position? etc)? I’m assuming the “performance measurement score” is your Y.  If so, what is the nature of this score? Is the score a performance rating that is subjectively given or is it quantitatively calculated from trustworthy data? 
    Assuming all these questions (and I’m sure there are others that could be asked) are validated, it appears to me that you are dealing with discrete or attribute X’s and Y.   Thus, the appropriate analysis tool would be the Chi-Square test for independence.  The null hypothesis for the Chi-Square test is Ho: X and Y are independent (no relationship).  The Chi-Square test is one of the least informative statistical tests.  If your Y is a continuous variable, with discrete X’s, then ANOVA would be the appropriate analysis tool.  The other concern I have is whether 50 is large enough sample size given that you have attribute data.  Again, I don’t know your process, but in general, attribute data require larger sample sizes to detect statistical significance. 
    As many have said before, these are my thoughts and I could be wrong!  Hope it helps in some way…

    0
    #75232

    Lomax
    Participant

    Balpreet,
    Before using ANOVA or Regression, have you utilized any data mining techniques to narrow down the 15 x’s into a more manageable list?  Are the X’s discrete or continuous? Is your Y discrete or continuous? Have you validated your measurement system and understand the amount of measurement error? Is your process in statistical control?

    0
    #74829

    Lomax
    Participant

    Hugo,
    If you have not already, you may consider using a Lean Enterprise tool called Value Stream Mapping to help you identify and target the sources of this inventory (Finished goods, WIP, Purchased Material).  I would imagine that you will use many of the “Lean” tools in conjunction with some statistical analysis on your project. For more information on VSM, take a look at http://www.lean.org
    Neil 

    0
    #74569

    Lomax
    Participant

    One of the things I try to instill in all my belts during training is that, like Six Sigma, they are not Silver Bullets.  They cannot solve all of the Corporation’s problems by themselves.  Their success will depend on how well they can lead/contribute to a TEAM.  They must be able and willing to share and receive ideas, tools, knowledge, and guidance from people at all levels of the organization when appropriate.  I view this discussion forum in much the same light.  If we, as Six Sigma practicioners, intend to proliferate Six Sigma as a profession and skill, we must set the example by sharing ideas, tools, knowledge, and guidance amongst each other.

    0
    #74475

    Lomax
    Participant

    I have sent the RTY Exercise to iSix Sigma submissions for posting and subsequent download.  It should be available soon.

    0
    #74423

    Lomax
    Participant

    Matt,
    I need your email address.

    0
    #74395

    Lomax
    Participant

    John,I have the RTY training exercise material you are referring to. Download RTY Exercise Templates (Microsoft PowerPoint format) [Download Help] Download RTY Exercise Templates (Adobe Acrobat format) [Download Help]Viewing Tip: Usually, you can click on a link to view the document — it may open within your browser using the application. If you are having difficulty, try right clicking the link and selecting ‘Save Target As…’ or ‘Save As…’ to save it to your computer harddrive.Virus Note: All files are scanned prior to uploading to iSixSigma. No prevention program is entirely safe. FOR YOUR OWN SAFETY, PLEASE:1) Re-scan downloaded files using your personal virus checker before using it.2) NEVER, EVER run compiled files (.exe’s, .ocx’s, .dll’s etc.). If you don’t have a virus scanner, you can get one at many places on the net including McAfee.com.

    0
    #74383

    Lomax
    Participant

    I didn’t mean to create “battle lines” by suggesting the use of EWMA charts.  I was simply responding to Gil’s request for advice on how to reduce the time lag of problem detection…this is the premise of the EWMA chart, to detect mean shifts quickly.  As Monte and others stated, the key is the correct application of the right tool for the right process.  On another note, there has been alot of discussion about whether SPC provides a proactive or reactive means of controlling a process.  I am in the court of belief that SPC is a proactive tool when you are monitoring the “Critical X’s” or KPIV’s.  The only caution here is that in order to be proactive and detect a potential defect before it happens, one must apply the other 7 “rules” (prior to a point outside control limits).  The risk is that the more “rules” you apply to a particular chart, the higher the probability that it will give you a false signal (detect “out-of-control” when in fact the process is still “in-control.”).

    0
    #74352

    Lomax
    Participant

    Gil,
    Is your process one that you could use an EWMA chart?  It will detect changes in the process mean fairly quickly depending on how frequent your data collection.  Obviously, you would benefit most from using an EWMA chart on your critical X’s to prevent the defect rather than post de facto.

    0
    #73491

    Lomax
    Participant

    Why do you bother if you’re that PO’d?

    0
    #65858

    Lomax
    Participant

    It sounds like what you want is a statistical tolerance interval, which makes a statement about a specific percentage of a population. The tolerance interval would be a function of the observed data upon which the assertion “meets six-sigma” was made. So there’s no simple answer to your question.

    0
Viewing 94 posts - 1 through 94 (of 94 total)