# luke skywalker

## Forum Replies Created

Viewing 100 posts - 1 through 100 (of 137 total)
• Author
Posts
• #186153

luke skywalker
Participant

Sweet. It’s been nearly a year since I’ve paid attention here, including the whole crash and recovery. What do I see? The same guys, cracking the same lines.

Nice to know that not everything changes. Rock on, crusty old dudes.

0
#182770

luke skywalker
Participant

First, keep in mind that there are no ‘standards’ for any belt definition, so just stick to what your company has decided on to avoid many unnecessary headaches (and posts here).
Cool idea, trying to make standard deviation meaningful to a group of newbies. This might be a helpful exercise: Briefly discuss with the class “driving to work”. For you this is the idea of establishing a process and you’ll quickly see that there is variation present. People will tend to have a “normal” way they get to work as well as a “back up” route in csase of accidents, etc. You shoud have them focus on their “normal” route for this one.
Pick 3-4 folks and have them give you a typical drive time, from home to work and write it down on the board. Then ask them if that is the exact same time every single day. Obviously, they’ll reply ‘no’, but will probably comment on how there is a typical window, or plus/minus span they work with, as in “about 25 minutes”. Next get them to offer up that typical window and show the range around each of their original point estimates for ‘drive time to work’.
From here, you can get to the ‘value’ of standard deviation. Certainly our little exercise will not provide a statistically correct SDEV (so back off math nerds), but the real point here is to get them thinking about how we manage variation.
Ask them is they plan their days to the point estimate, or if they plan more towards the varying times, or the “abouts…” Next, set up a scenario where they do plan to the point estimate – you can draw up a ‘day planner’ on the board showing their drive to work and several subsequent meetings – all planned perfectly to the exact times. Then simply ask what happens when the drive takes 3 minutes longer.
You might be able to get them to continue this idea through meetings that start and end later than scheduled, and perhaps even note a trend to increasing lateness as the day wears on. You can ask them where they have seen allowances for variation (please allow 6-8 weeks for delivery…)
Finally, you can wrap this discusssion up with transferring the idea of variability to the project timeline that each of them is part of. What happens if they take 2 weeks longer with Define, or if they realize in Analyze that it will now take 3 months to collect any sufficient amount of data to “prove” their improvement will really deliver the promised results?
Hopefully this helps you – what ever exercise you use, you’ll get more impact if you keep it simple and have it deal with their daily lives where they can internalize the concept of variation.
Cheers.

0
#181960

luke skywalker
Participant

Before you get completely insulted by the reponses you are sure to receive on this, please consider what you’ve actually asked here. Your post shows that you’ve put no effort whatsoever into answering your own question.
Have you ever purchased anything? What makes you happy as a customer, or what makes you decide to stop doing business with someone?

0
#181647

luke skywalker
Participant

Well, at least we made you come to class to boot modules a few times. Nice salad bowl though.

Cheers!

0
#181630

luke skywalker
Participant

Back before the rebellion was crushed, as Darth was rising to power, his mentor, the grand wizard Mikel, made a proclamation to the masses: “Behold. Green Belt wannabes, only for as long as you shall hold dearly to your control plans will your little processes operate at nearly th esigma level you have claimed. Without your intense micromanagement, the variation creeping into your process could degrade your results by….TWICE AS MUCH!”
Mikel held for an inordinately long dramatic pause and watched the eyes of the disbelievers. Sensing he may have stretched it wee bit too much, he shrugged his shoulders and muttered, “Well, maybe not THAT much, it might only be about 1.5 times.” And then in his very best Tony Soprano, he glared at the little Green Belt candidates and said, “You gotta problem wid dat?”
Those very GBs grew up to be consultants/prophets who have spread the wisdom of the 1.5 shift ever since.
And congratulations, you are now the 1,000,000th to have asked about the legend of the 1.5. You get a bright yellow picket protector.Please allow 4-6 weeks for delivery.

0
#181361

luke skywalker
Participant

I’m sure there is an easier way to do it and you have stated such. It’s the warm sene of compassion for newbies that really chokes me up. Must be great for drumming up new business.

0
#181356

luke skywalker
Participant

This sounds pretty cool.
Since you mention that both the parts and the measuring are leading edge, is it possible to effectively split this into a 2-part study, meaning you could potentially use a previous generation of part that might be more stable or at least better known to pin down the measurement variation? From there, part 2 would be to use the now mor efamiliar measuring tool on the variable parts. At least that way you would have a means to check stability within the measuring. Seems this needs to be addressed before plunging into crossed/nested designs, etc.

0
#181341

luke skywalker
Participant

Can you share what the parts are, or what they are made of? Also has anyone checked to ensure the measuring device is not broken?

0
#181333

luke skywalker
Participant

So basically, the lean guys are the Neidermeyers and we are the Blutos, Hoovers and Otters at Improvement U?

0
#181207

luke skywalker
Participant

Disagree
I believe Villanova is the best!

0
#181206

luke skywalker
Participant

0
#181173

luke skywalker
Participant

Delta began their program within the maintenance division in 1998. Never had top level executive sponsorship though, so despite some great work, the overall business has not capitalized on it despite several critical moments such as 9/11, the subsequent bankruptcies, and their current merge with Northwest.

0
#181164

luke skywalker
Participant

I can see it! I can see it! The Jedi thing is working!
And Darth said I could never do it! HA!

0
#181159

luke skywalker
Participant

Hi Ken!
Hope you had a nice day in DC yesterday. I’ll take the bait on your taunt as one of the guys still here in the trenches, though no longer in the SS group.
Some interesting data – I just referenced Google finance for these stock prices, which is what publicly owned companies use to inidicate performance. Since Q3 of 2006 (to steer clear of the ‘start’ of the eroding economy), these SS “power houses” have shown ‘stellar’ performance:
Motorola: down 75%
GE down 67.5%
Honeywell down 21% (but off nearly 50% from  a high in 07)
Ford down 75%
So is BAC in a rough patch. Yes. Does SS determine the fate of any company? No.

0
#181119

luke skywalker
Participant

As books on SS go, it’s fine. They all have pros and cons and are all dependent on your starting point as an individual. If you’re going through GB, prsumably it’s provided by some trainer type with a big, honkin’ binder of power point slides. That binder should get you through a lot of the static.
Books that are “good”? Anything by Peter Pande. He writes with a good, practical point of view and makes it approachable.
The “Dummies” book will be a good guide for stuff you might ask along the way, but keep in mind, that since there is no standardized BOK for SS, the books will probably not synch up with your specific training experience.
I’m happy to continue the dialog offline if you’d like. Let me know and I’ll provide a different e-mail addy.

Cheers.

0
#180577

luke skywalker
Participant

Mike, you are correct. However we have been conditioned to have less agressive expectations. Spelling is still several cycles of learning away.

0
#180517

luke skywalker
Participant

“using” will be the term to define here. What we taught and what I see got nuthin to do with each other here. Well, that’s not true. Many more people can say “DMAIC” now.

0
#180490

luke skywalker
Participant

Ouch. What a hack.

0
#180427

luke skywalker
Participant

It would seem that the Empire has located the rebellion’s current hide out, as the death star is powering it’s primary weapon for a demonstration of force.  Tick tock…Rough time to be a Jedi.

0
#180409

luke skywalker
Participant

I’m happy to help, as I have a particular passion for feather merchants. Let me know and I can provide offline contact info.

0
#180299

luke skywalker
Participant

You could add 1-2 slides in your presentation stating that change management is good and is an integral part of an evloving quality system within an organization. Use clip art and some animation.
Darth can help with that part.

0
#180154

luke skywalker
Participant

Loved the name of the event – “The Retreat to Go Forward”

0
#179672

luke skywalker
Participant

Arnel,
It seems from your post that the goal of a BB cert is to help you get a better job, and that the ‘pedigree’ of the ‘school’ providing the cert will add to that.
Given that, I would suggest that you do a few simple and very economical things. First – go to a book store and pick up 2-3 books on SS – doesn’t really matter by whom, as long as you can recall the authors’ names later for name-dropping purposes in interviews. Then go to a nic epaper supply or office supply store and buy a very professional looking pre-formatted certificate. Take it home, develop a Power Point slide stating your achievement in SS as a BB and provide a date, the name of some consulting firm, and then print & sign it. Also, get a nice holder or cover for it – so you don’t damage it during travel for interviews. All in, this should cost you less than \$75 – a far better solution than \$8000.
If you’re still reading, sarcasm aside now – If the real intent of this BB cert is to help you get a job, then follow my instructions above. Be honest and admit that is what you want. IF, however, you actually have interst, and perhaps, passion for improving a business and succeeding through those steps, then look fro a real program, with projects(multiple) required, classroom or instructor presence integral, and a heavy emphasis on “hands-on” or applied learning.

Good luck.

0
#178085

luke skywalker
Participant

Agree that TPM is but one of many tools in the box. However, during the brief period of the Allied – Honeywell merge (when Jack Welch almost bought A-S), a number of A-S MBBs jumped ship and began teaching TPM as the basis of the process methodology, making it more than a tool. Not saying it’s the way to go, just sharing my experience with it. I believe they were in alignment with the published authors of the TPM papers ( Sanders, Hild, Ross).

0
#178082

luke skywalker
Participant

The training your folks went through is a version that I believe originated in Allied-Signal and uses a premise of Thought Process Mapping as a blend of the socratic method and deductive reasoning to move from an intial problem statement to a conclusion by using the tools as needed to solve the problem. It can work well in an engineering based  corporate culture. As I wok for a company that is not engineering based, it would have failed immediately had we used that approach. DMAIC and it’s corresponding tollgate reviews, etc., provided enough structure for new GBCs and BBCs to succeed.
Having been trained on the TPM approach and now having worked with the DMAIC approach for a while, I would not recommend using TPM for GBs without a solidly established BB core that really understands both the methods and certainly the business – would agree with Stan’s conclusion in that case.
Cheers.

0
#176335

luke skywalker
Participant

Well, if we do as Darth requests and revisit his original post, he does not mention “less than or equal to”, or “greater than or equal to” as logical conditions, but simply “”. I guess this is one of those subtle Jedi mind things where we are all supposed to know what Darth really meant, instead of what he typed. :0)
As it is stated, any line items that are either 0.8 or 0.95 should return a N/A result.

0
#176331

luke skywalker
Participant

Chatted with him on Saturday adn the report was ‘very busy”, but have left 2 subsequent messages since Sunday without response.
I guess the battery in his light saber/cellphone needs to be charged.

0
#176326

luke skywalker
Participant

Well, it could be 1st shift, 2nd shift, 3rd shift, management, HR for the 5 big causes of variation in mfg…

or more likely you are looking for ” the 6 Ms”: man, method, machine, material, mother nature (environment),and measurement.
Draw a fishbone of those and ask yourself what variation exists within those categories and I think you’ll find plenty to work on.

Cheers.

0
#176325

luke skywalker
Participant

Sorry to step into the ‘Wheatie pissing contest’, but there are at least 2 “simple”ways to accomplish it (using Excel 2003 as a refernce version).
If your results values (A, B, C) can be represented via color management, as in a dashboard, then you can use conditional formatting, found under ‘Format’ in the main menu bar. Excel 2003 has a limit of 3 levels of conditional specification there.
If your results (A,B,C) lend themselves to something other than colors, a nested ‘If’ statement will provide the solution:
=If(C1<0.8,"A",If(0.80.95,”B”,”C”))

0
#176322

luke skywalker
Participant

Well, glad to see your obvious meter is calibrated today. Given the vague statement in the initial post that restarted this old thread, I think several viable options have been offered up free of charge, so that has to be a good thing. Hope you have a happy day.

0
#176319

luke skywalker
Participant

Try to affect a mock British accent when you say this over your company’s PA system: “But this one goes to 11”

0
#176081

luke skywalker
Participant

I know it’s in the course material here these days (CI for variances, etc.). I’ll e-mail the nuts and bolts to you – just let me know if it works for you.

0
#176078

luke skywalker
Participant

One thng to keep in mind with all of these Six Sigma flavored stats is the timeline of their development. I’m a bit surprized the wizened OLD Darth didn’t mention it. The CLT is a practical, though intensive sampling strategy to “get normal behaving data” – so you can use it with your limited box of analytic tools, historically based on the normal.
Sometime when you have a few minutes free (ha ha) take a look at when some of these tools were developed and contrast the traditional stuff against non-parametric tools and yes, even Shewhart’s charts.
We exist in a cool time – sort of like CGI, where the tools and technology actually look good, we have a good number of tools available to essentially work with data in its raw form without need of transformation (most of the time).
And Darth is right – In classrooms, the CLT is mostly useful as a way to explain the standard error of the mean, which has a bit more play within most SS curricula.

Happy charting.

0
#175594

luke skywalker
Participant

Sure – it’s simple and will provide an answer. I guess I was reading into the post that gaining more process knowledge would provide a more holistic answer. Silly me.

0
#175583

luke skywalker
Participant

There are probably a few things to check at this point. Honestly, you score points for having the wherewithall to check the variances prior to ANOVA.
Hopefully you understand where your data came from – how it was collected, and from which process/processes. Look at the data graphically – perhaps a simple dot plot or box plot will show you how the data behaves. Also pay attention to the smple size(s) – are they equal, similar, or lop sided? Were the samples collected within essentially the same timeframe, or at different times? These are all potential indicators of you having 2 actually different sets of data, where unequal variance may be part of reality.
If you’ve made it through all that, and everything checks out, you could leverage Welch’s test, which is a variant of ANOVA that can account for variance and sample size. Try googling it, it’s an approachable formula and can easily be put into a spreadsheet.

Good luck.
If

0
#175434

luke skywalker
Participant

Well, most talented, brilliant, higher thinking executive types require stop sign colored dashboards to “get it” – Seems only “Green means good, red means bad” sinks in for most. No stats, no control charts to cloud their pure and empty thoughts.
Stan’s suggestion of main effects plots is spot on, but honestly, you will still be counseled on that, as they will not be used to seeing it. Different is bad.
So see if you can get a look at what the braintrust uses now, tweak it ever so slightly and see if it’s sticky enough to get you invited back for more. If not, you don’t want to be there.

0
#174617

luke skywalker
Participant

Hey Darth,
Was in Indy yesterday and had a chance to see the “beautiful” Lilly building on the way back to the airport. Could have sworn I saw a big flag that showed “Lilly, a Vader Subsidiary”. You must have really done some significant work up that way.

0
#174096

luke skywalker
Participant

Oh, by the way, found out it’s now Dr. Yoda. He passed the last hurdle recently.

0
#174035

luke skywalker
Participant

Lots of fun here in exile. You wouldn’t even recognize the place these days.

0
#172740

luke skywalker
Participant

Yes, your sample size is several order of magnitude too high for a p-chart. It can adequately handle samples around 500 but fades as the sample sizes increase from there.
There is a set of test criteria to assess this: np>5 AND n(1-p)>5
If you satisfy both of those, the stats books suggets that your binomial data approximates the normal enough for you to use an I-mR chart. The observations plotted there would be th eindividual sample proportions.

Cheers!

0
#59357

luke skywalker
Participant

Makes sense – again, it poses an interesting scenario for charting. Thanks.

0
#59355

luke skywalker
Participant

Stan – Did you mean to post this in the financial forum?
Case 1 lends itself to and X-bar/R rather than an individuals – Assume there’s no issue there.
For case 2 – not sure how one would homogenize 5 adjacent FX trades or adjacent complaint calls. Seems more like a wet chem issue, hence the question.
Cool think think about though. Say Hi to little Stan from the folks at BofA.

0
#170085

luke skywalker
Participant

Well, since the control limits tell you the boundary of expected and not-expected variablity and the moving range is best calculated from 2 points, zero is not outside the expected range.

0
#170036

luke skywalker
Participant

There are already a few good comments to help you an dhopefully this will as well. I’m not sure where the magic 30 showed up for sample size, but given your discrete response (A or B) and the discrete factors (citizenship, etc.) you’ll want much more than 30 observations to make a conlcusion with any statistical power behind it.

BTDT’s suggetsion of using Minitab is great – they have a sample size calculator which should be useful for you.
Also be aware of lurking variables that may really be driving the response outside the short list of questions you’ve provided here.

Have fun with it.

0
#169696

luke skywalker
Participant

Well there’s a joke that prefaces this, but here it is in 5 words:

Run faster than the bear.

0
#169621

luke skywalker
Participant

20 different people will have 20 different opinions.

0
#161776

luke skywalker
Participant

A couple items – using Minnie and your info provided in your post, I get a 2-proportion sample size of 88, and you need to use the alternate hypothesis as ‘greater than’ since Minnie’s convention is to test proportion 1 against proportion 2.
I set up the 2 proportion test using the 88 sample size with corresponding defects and got a p-value of 0.001, which idicates a significant difference between your estimated 10.5% starting rate and the 0.5% after rate.
Happy to help with the step by step and interpretation if you want. You can reach me at everydaoandzen@yahoo

0
#161761

luke skywalker
Participant

Regardless of cited style, to me the most compelling refernce is a 60-90 page Power Point deck for each concept. It’s particularly helpful if each multi-terra byte file is attached to the post.

0
#160212

luke skywalker
Participant

Multiple regression migth not be your best bet since your output is a count, and therefore discrete. If the output is a large number and you have data showing a wide range of values, MLR could provide some decent guidance. If the counts are low and the range of vlaues is narrow, you might look to logistic regression.

0
#155757

luke skywalker
Participant

Omer,
Do yourself a favor and continue reading. Also, anticipate several ‘less than polite’ responses for asking this question, as the 1.5 sigma shift issue as been a common topic of discussion here for over 4 years.
Given that, there is a veritable plethora of information about the 1.5 shift on this website – simply use the search tool and you’ll be directed to a number of articles and discussion threads.
Outside of this forum, you might ‘google’ 1.5 sigma shift and Mikel Harry – that should get you a great reading list.
I hope this helps.

0
#155526

luke skywalker
Participant

One explanation that always made sens eto me is this:
Imagine you’ve just purchased a new truck and are about to park in your garage.
Cp answers ‘Does the truck fit in the garage?’ (Does our process variation fit within the specifications?)
Cpk answers ‘Where is the truck located within the garage?” (Is our process centered between the specification limits?)
Hope that helps!

0
#153773

luke skywalker
Participant

You so funny! You funny, funny man.
Seriously, based on what I see from a recruiting role, reverse your statement and you’ll be heading in a more correct direction.
Love ya,
Luke

0
#152003

luke skywalker
Participant

Well, once again we see the mash up between eastern and western thought. Oh, and cool trivial linkage between martial arts and SS…2 Kanos have contributed to this messy bundle of SS. Noriaki has given us his analysis to see what our cutomers want and like. Jigaro gave the world the modern belt ranking system from Judo a little over a hundred years ago. He started simply with white, green, and black.
Now if you visit any strip mall Take Yur Do dojang, you’ll see a veritable Crayola box of belt colors hanging on the wall. That’s so every 6 weeks they can charge you another \$45 for your kid to get the dark green belt with the blue stripe – all to match our sand flea level attention spans.
And so we follow suit…the corporate workspace demands reward and recognition – so we must have many certificates to show, and consultants must have something to differentiate their service from the next guys (reference the Gold Black Belt: 4 weeks of BB with 1 week of Lean).
All that really matters is having the desire to improve things and getting the skills to accomplish that. Having a Gold Sequined Spandex Belt might get you an interview with Daddy Darth, but once you’re there, only proving it will keep you there. So take any training your company wants to provide you for free and learn like crazy – it might help your career path.

0
#152000

luke skywalker
Participant

The brochure was not a bad read, but it was disappointing to find out at the end which industries they focused on. That does hint of a commissioned report for iGrafx. I doubt 99% of users actually use flowcharter to pmap (as the article states) – Visio seems to be more prolific from what I’ve seen.
Still the motivation and momentum issues are interesting. Thanks for the link.

0
#151926

luke skywalker
Participant

Marko,
Get ready to hear conflicting responses. EdG’s response was quite good – Kaizen can be thought of as a part of Lean. Lean is a long term philosophy and Kaizen is a day by day component of that drive to eliminate waste and rework in order to create and preserve a smooth, flowing operation.
Kaizen roughly translates to continuous thoughtful improvement and involves each employee taking 1 small task and finding a way to do it bette – then continuing each day with another small item. With all employees having this ‘continuous improvement’ mindset, the results are not only visible, but sustainable.
You will hear of “Kaizen Blitzes”. Are you familiar with the term ‘oxymoron’? Kaizen is steady, incremental, grass roots level improvement. A ‘blitz’ is anything but. Typically, there is much excitement, an influential executive shows up to stress the importance, a bunch of people swoop into an area and clean it up – called “The 5 S’s”, except they usually make it through the first 2 before losing monentum in their 2 day “blitz”. Then about 90 days later, that area is back on the schedule for another “kaizen blitz”. It’s rather like a sugar rush when you’re trying to run a marathon. It feel sgood for about 10 minutes and then you crash.
Pick up a copy of “Lean Thinking” by Womack and Jones – you’ll find it more approachable up front than a book on TPS.
Hope this helps…

0
#151798

luke skywalker
Participant

No, your invitation did not get lost  – I have to protect my bride form the dark side as long as possible. Should be a fun party, so watch your mailbox…
We’re exploring the deeper levels of chaos theory here these days – How’s the placebo business treating you?

0
#151786

luke skywalker
Participant

Glad to see you’re still yourself.

0
#150297

luke skywalker
Participant

It was actually the Evil Empire and some guy named Darth  ;-)

0
#148603

luke skywalker
Participant

Are you suggesting that having an improvement effort that incorporates both continuous and ‘breakthrough’ improvement will not work? That would see to fly in the face of Michael George’s company (despite employing Darth Vader and Darth Maul), which is rather successful at this point.
Perhaps it would be beneficial to remain flexible in our thoughts and draw in those concepts and methods that enhance effectiveness and efficiency of problem solving – much like the original Motorola team must have done.
Just a thought.

0
#148555

luke skywalker
Participant

Kind, supportive amd compassionate mentoring in action. Makes my heart feel all warm and fuzzy. Proud of you, Daddy-O.
Hope all is well with you.

0
#146769

luke skywalker
Participant

I’m not too good with p-values, but I think you could use a formula like:
1 – (cumulative probability of the correct F distribution at the location of the test statistic) to get p.
Of course looking it up in a table or having a computer do it for you like everyone else does would be way easier.

0
#145717

luke skywalker
Participant

Sure…
Your expecter proportion of 0.10 is the Hypothesized (expected) proportion in Minnie. Using your precision of +- 2%, you’s supply Alternative proportions like 0.08 and 0.12, etc.
You need to determine the power, though it typically ranges between 0.8 and 0.95 – you can put multiple values in the box.
Unfortunately, the problem with Minnie is your ‘population’ size of 332. Minnie does not employ the FPC that Daddy Darth has mentioned. For that, the Excel file he offered will get you to the 241 figure you have predetermined.
Keep in mind that both of these sample size formulas are for describing the characteristic of interest – be it a mean or proportion, with a confidence interval around it. If you are looking for sample sizes to conduct specific tests, you’ll need to look further and Minnie can help with that.

0
#145707

luke skywalker
Participant

same place: Stat/Power and Sample Size
select 1-proportion and that will get you the equivalent for discrete.

Darf Duz Statz… hee hee

0
#144305

luke skywalker
Participant

The Jedi are few in number and are battle weary (light saber batteries are really low!). Our planet is heavily populated with representatives of the Empire and exile seems to be the remaining path for survival.
Other than that, it’s just dandy.

0
#143858

luke skywalker
Participant

Same question as before they’re just asking it a different way. You’re given the actual means, the actual sample size (n=30), and you’re given the expectations (0.1 for answer time and 0.5 min for decisioning). You can set up your picture for each, then set up the t calculation to see if they are different. Keep in mind, you’ll be doing a t-test here and specifically a 1-sample t-test for each. Same approach as the Z but it’s better for small sample sizes.

0
#143856

luke skywalker
Participant

OK, good start so far. One thing I would suggest is to draw a picture to see where along the horizontal axis you are (Z numbers here) and then how much of the curve you have accounted for.
In this case, when asking about 475 or less, you’ve arrived at a Z of 1.68, but since we’re really looking at a potential result less than the mean, it’s -1.68 (left side of the picture). The area – or cumulative area corresponding to that Z score is 0.4535, or about 45%.
The question wants to know how likely it is to see some volume less than 475. So that’s beyond your Z score, or further left on the picture. If you take 0.5 – 0.4535 you get slightly less than 5%. That’s how likely it is to see a result less than 475.
Hopefully by drawing a picture and labeling it with the information given, you’ll have an easier time makking sense of these word problems.

0
#142569

luke skywalker
Participant

Thanks everyone! This thread has made my day. I was convinced that it’s just the place where I work that has this sense of entitlement – wanting perfect and complete answers nicely bundled on a hypoallergenic silver platter with no effort extended. Knowing that it’s the world at large makes me feel like I’m part of something bigger.
Now I can tell confused belts “It depends” with a sense of zen like calm.

0
#142443

luke skywalker
Participant

Did you put gas in it? Even Japanese motors generally need fuel.

0
#142439

luke skywalker
Participant

0
#142127

luke skywalker
Participant

Keep in mind that these tests may look for differences in medians between samples, but the formulas to generate the test statistics are different, and the underlying distributions that actually determine the p-value are different.

In your case, look for outlier points in some samples – graph the data and assess how similar the groups are. Mood’s is more robust when the data samples have outlying points and Kruskal-Wallis has more power to detect differences when the samples are basically the same shape.

0
#141045

luke skywalker
Participant

Not sure I’d bet my career on a single qualitative data point. The story is consistent with American manufacturing culture over the past several decades – that we are soft and lazy and the lean hungry other folks work circles around us.
Mmm Hmmm.

0
#140854

luke skywalker
Participant

You know, it’s funny. We all decided at some point to leave where we were to see what the BOA program was. Some undoubtedly chased a larger paycheck, others, especially early on enjoyed the challenge of a start up program, and some were just plain happy to find a job in the area after their previous employers hit rough times.
Then along the way we have all made our decisions to stay or leave. And yes, I can attest that Darth is correct and some have “been asked to leave.”
Jack Welch had a ranking system of A, B, and C players, where as we mostly know, the C players get mustered out annually. Here we have a similar system of A, B, and C, but the meanings are different.
When we encounter unfavorable conditions at work (say the previous ‘ledaership’ scheme):
A players say, I don’t like this and decide to quit.
B players say, I don’t like this, but I love my paycheck, so I will sacrifice my integrity, suck it up and stay. (yes, sir, right away sir!)
C players say, I don’t like this, but this is part of why I came here, so I’ll work to change the way it works.
We all manage our careers and make our own decisions. Chasing greener grass is part of corporate America, but most know that the grass looks greener because it’s over the septic tank.
I hope your decision to leave had good results for you and your family. I do however seriously doubt that your new employer is really any better from a holistic program viewpoint.

0
#140702

luke skywalker
Participant

Good to hear about your meeting. Had a good meeting here this week as well. We can catch up this weekkend or next week on the phone, which ever works best.

0
#140696

luke skywalker
Participant

Wow, that has about as much to do with MBBs and Six Sigma as many of Daddy Darth’s posts.
Impressive.

0
#140534

luke skywalker
Participant

If there are indeed clusters, or non-independent questions in your survey, you can use factor analysis to identify them. Minitab has the capability to do it.
Ideally you would consider this upfront when designing the instrument, but it’s great to know even after the data starts rolling in.
Good luck with it.

0
#139827

luke skywalker
Participant

If you’re assessing inventory levels, you may want to determine if you’re looking at weight, volume, or length as measures or if you’re counting how many items there are. If counting, and you have fewer than 30 observations in each subgroup, then I would imagine you will find a great many subgroups to be non-normal – due to the nature of the data and the sparcity.
You might also look into the various normality tests as they have different approaches and the default Anderson-Darling test offered in Minitab may not be the most appropriate for you.
As a quick look, you could use boxplots, displaying both the median line and a mean symbol to look for skewness in the subgroups. The whiskers can be a good indicators too.
Hope this helps a bit.

0
#139679

luke skywalker
Participant

Thanks for letting us inquire. I’ll be sure to send an Outlook invitation when I’m ready to do so.
Besides, why sit in a coffee shop when there’s a perfectly good pub around the corner?
Tell me the name of the pub.

0
#139675

luke skywalker
Participant

Hey, let’s chat offline. I can see some of what you’ve asked about here and have actually though through it a bit.
But, since the kids are playing upstairs (“Milton” and “Chuck”), I’d prefer to keep program specifics out of the cheap shot arena.
Hopefully I can address your other e-mail at that time as well.
Luke

0
#139632

luke skywalker
Participant

I guess it depends on where you go to school and what you study. Engineering students should have a pretty wide array of companies that sponsor co-op programs and many of which have Six Sigma programs.
I’d check with the university to see about intern/co-op opportunities and be specific in asking about companies with Six Sigma or Lean programs. It won’t hurt to ask.

0
#139627

luke skywalker
Participant

Sure – I’ve worked with interns who certified here and at my previous employer. The interns were either on a  single semster stint, or on a recurring co-op program. If you could identify a company that sponsors the latter, you could conceivably get through BB with an appropriate level of application (project work).
Seems many of the academic programs focus on the classroom experience because well, that’s what they do, and miss the point that Six Sigma and Lean are applied methodologies – and are applied for the purpose of improving businesses.
It’s like taking a class with a professor who came to teach after a career in industry and taking the class with a professor who recently completed their PhD. One has done it, one has read about it.
You can decide which is the richer learning experience.

0
#58822

luke skywalker
Participant

Some programs would have you carry a red hot brazier across 30 feet of rice paper. If you can do that without tearing the paper, you would show mastery, grasshopper.
Of course the industry has yet to standardize on that apporach, so you may find varying criteria.

0
#139527

luke skywalker
Participant

Great feedback so far on the website design. One additional thing to consider (that we ran into about 4 years ago) is to look beyond the GUI. A very real measurement is page loading speed – which affects both user satisfaction and volume, or throughput on the site. Even if the actual content is great, slow loading will tend to drive users away.
Traditional DMAIC will be able to address this, and additional support from a more DFSS themed approach will perhaps even better enable it.
In any event, you can describe a clear Y=f(x) model, measure the Xs and Ys and define benefit thorugh a mix of tangible savings, potential revenue generation, customer/user retention, etc…

Hope this helps.

0
#139470

luke skywalker
Participant

Ummm….”Tall”????
Must be the boots, Dark Helmet.

0
#139291

luke skywalker
Participant

With sincerity, I hope you all did have a good and happy Fathers day. I would have called or written but had to get new batteries installed in this mechanical hand – thanks for that by the way. Everytime the humidity goes up, this thing goes on the fritz (don’t buy spare parts on Tatooine).
Even though you have gone to pasture, I can definitely hear your theme music playing here, which indicates the presence of yet another Sith. Should be fun.

0
#139148

luke skywalker
Participant

in classroom E.

0
#138885

luke skywalker
Participant

With the proper amount of beer, the bamboo stick seems like a light saber. That’s why it’s important to engage these hobbies in the correct order.

0
#138879

luke skywalker
Participant

Well, I hear the guy running the education team used to work with Darth…does that give you any insight?
Seriously, it’s a good program and will challenge about anyone’s technical and political acumen.
You can reach me offline at everydaoandzen@yahoo.com

0
#138841

luke skywalker
Participant

What is this? Good Darth, Bad Darth? Usually when you see a simple, if naive post, you employ your nearly patented scorched earth personna. This is indeed a refreshing change. Maybe you could leave the dark side behind after all….

0
#58811

luke skywalker
Participant

Pete,
A couple of ideas…First, MTB is not real helpful for non-normal capability analysis, but is rather decent at distribution fitting, so this is not too difficult. If you have fit the data to a logistic model, you could use the cumulative density function to determine the actual area under the curve beyond your spec limit. That would be your effective ‘defect rate’, which enables you to get a DPM or perhaps  DPMO and from there it’s a short hop to sigma level.
Of course, I would look at it all umped together, but anticipate needing to assess the 2 strains of subprocesses independently to better understand the root causes within each.
Hope this helps.

0
#138694

luke skywalker
Participant

I’m honestly not sure if there accepted threshold values for skewness and kurtosis – given that we are dealing with inferential stats. I’ve seen ‘experts’ state an acceptable range from -1.0 to +1.0 as being ‘normal’, yet other ‘experts’ seem to get queasy when skewness goes beyond a +/- 0.7, etc.
I do know of normality tests that have been developed using these stats, but not in relation to the A-D test. And there seems to be industry specific preference for the normality test of choice.
I would rely more on the array of normality tests (AD, KS, RJ, SW, etc) and a normal probability plot to assess the data considering that a set of data could have skewness and kurtosis values indicating normality – but either a weibull or log-normal condition could fit there.

0
#138649

luke skywalker
Participant

Wilson,
Now that the chuckle patch has gotten off on a tangent…
Are you looking for additional ways to leverage skewness and kurtosis to determine normality, or are you looking for additional normality tests? Depending on which, a Google search may be helpful if you know what you’re looking for (NIST site is great), or you may just need to look within the software package you use for available tests.
Best of luck.

0
#137757

luke skywalker
Participant

“Black” said the Pot to the Kettle.
What method did you elders decide was most effective for attaching photos to your posts? That only took about 3 weeks to resolve.

0
#137680

luke skywalker
Participant

Oh, and believe it or not, The Cartoon Guide to Statistics is a decent book for explaining this stuff in absorbable terms. Lots of stats books out there, just need to find the right themes. Psych stats and business stats books tend to provide clearer examples than engineering stats, but that will vary by individual. For me, cartoons work wonders as Darth can attest to.

0
#137679

luke skywalker
Participant

Well, sort of. First, Pete is on the right path in his desire to find a difference between the 2 processes. A good first step would be to check the normality of each sample. If either or both are not normal, a nonparametric test like Mood’s Median could work. If both samples pass a normality test then you can use either a 2-sample t-test or a 1-way ANOVA as suggested earlier.
Now the cool part. ANOVA does require (assume) equal variances in order to function and that’s of more concern than non-normality. You can run the 2 sample t-test more easily, but need to check the variances first via an F-test. Both JMP and Minitab can perform this. The reason you care is that you can select or de-select options in the t-test dialog to direct the program to calculate the t-statistic correctly based on whether the variances are equal. If the ARE equal, Darth is correct and a pooled standard deviation is used. If they ARE NOT equal, the individual variances are considered to provide a better test statistic. Either way, both JMP and Minnie can get you there with solid results.
Best of luck.

0
#137265

luke skywalker
Participant

Well, before you get tooooooo wrapped up in your advanced degrees, would it perhaps make sense to remember what the question was?
Seems this guy wants to build a survey, probably using a 5 category Likert style scale (hopefully with definitions for each anchor) and would like to know what makes BBs tick.
The use of ANOVA or t-tests is common – probably more so from a rationalization of ‘that’s what we’re familiar with’ rather than it being the most correct approach. Since it’s a survey with limited responses, technically, Darth is correct (yes, that hurts) in that the data will most likely not be normal and is not continuous, or really even interval in nature. I always thought Welch’s test handled ANOVA based issues where there were either grossly different sample sizes or unequal variances (generally a p-value less than 0.01). The intent is there though – to assess the difference in responses.
Maybe all of your advanced degree programs covered issues like alpha and beta risk? What adverse effect is produced by using non continuous, non normal data (haven’t even addressed sample sizes yet) in a test that explicitly states normality as an assumption ( t-test, ANOVA and Bartlett’s)? What is the reduction in power of the test in that case, and how could that affect the business once decisions are made and programs are either launched or scrapped?

0
#135651

luke skywalker
Participant

Depending on the response or Y data that you want to use for this analysis, there are 2 good methods and neither requires you to transform the X data.
If your Y is continuous, you can use a general linear model (GLM) approach – which will test the Xs like ANOVA would, but it can handle both discrete and continuous Xs (covariates).
If the Y data is discrete, you can use logistic regression. It will work whether you have binary, nominal, or ordinal data.
Most of the stat software packages will be capable of these analyses.

0
#135478

luke skywalker
Participant

Good observation about the Dark side. Since my daddy Darth had not jumped into this one yet, thought I’d have a go at it.
BTW, I’m an MBB too. The way the sigma level calculation works on this site, you would have to perform so much intermediate work to prep your data to arrive at the sigma level, it’s really not value added.
Transforming the data to fit the normal is somewhat outmoded. Why not just fit the data to the log-normal  and get the proper parameters so you can determine the actual area non-conforming? That would seem to be an easier and more correct way to arrive at sigma level for this type of data.
I’m just having some Friday fun, but on a more serious note, those of us who are MBBs should feel some inclination to help innocent requests – even from lazy people as the original poster seems to be, since they managed not to see the icon at the top of the screen for ‘sigma level calculator’.
May the Force be with you…

Luke (wearing black and swinging the light saber)

0
#135461

luke skywalker
Participant

If I could manage to search the site and use the provided sigma level calculator, would it provide a believable answer if my data appeared to be log-normally distributed? You sound like you know what you’re doing, so I hope you can provide some guidance here.

0
#133681

luke skywalker
Participant

Awww, the kind and nutruring ways of Daddy Darth. Makes me misty.

0
#131124

luke skywalker
Participant

Now Pappy, it wasn’t that long ago that you were a part of the rebellion here…Some of us are still here fighting a good fight.
I do see far too much emphasis on looks over substance from the light-bulb factory folks who work here though, and that can happen at any company.
The comments on checking the candidate’s skills are spot on – getting a nice piece of paper or a plaque means nothing, BUT being able to answer 10 simple questions can show some substance, wouldn’t you agree?

0
Viewing 100 posts - 1 through 100 (of 137 total)