THURSDAY, APRIL 17, 2014
Font Size

MBBinWI

  • What does your collection plan look like? Are you looking at historical data, or is this data that you are currently collecting? One of the first things we need to do is determine homogeneity of the data. I don’t want to make assumptions because I don’t know much about the data you collected. 2,000 data points could be okay, but, if these…[Read more]

  • Time to process these contributions would be most useful? I’d not bother with VA ratio (not much transformation of the product occurs).

  • Chris Seider replied to the topic Big Non-Normal Data in the forum General 5 days, 12 hours ago

    If you must, you can convert a ppm defective to a Ppk but I wouldn’t recommend that “exercise”.

    If you still find majority aren’t normal….there could be many reasons for this and how to proceed is a long answer (including how good your gage is–as my friend and colleague @MBBinWI suggested).

  • Chris Seider replied to the topic Big Non-Normal Data in the forum General 5 days, 16 hours ago

    See if you can stratify with some demographics and see if normality can be seen. Often when you have multiple devices/processes/units producing a product/service and you combine all of the data together, you’ll fail the AD test for normality.

    Define your problem as ppm defective and life gets easier.

    DON’T forget to look how your data changes…[Read more]

  • @MBBinWI Homework question? ;)

  • Chris Seider replied to the topic FTY=e-DPU in the forum Methodology 1 week ago

    LOL…..

    How about this kind of flying item?

  • Chris Seider replied to the topic FTY=e-DPU in the forum Methodology 1 week ago

    It’s RTY….not FTY.

  • What a horrible corporate learning environment you are placed into. No guidance–how ridiculous!

    Heck, maybe they could redesign the question. For one, they haven’t defined the number of opportunities involved in the process so technically, all they are doing is having you assume one opportunity to turn a simple ppm defective into dpo. Also,…[Read more]

  • I was a bit off….got yours but keep in mind AC :)

  • I didn’t get close to what you got…

    Don’t forget to rerun your model after eliminating grossly unstatistical terms. However, if you did really get your scenario, most would say to keep A, B, and C along with the AB interaction–as long as you are happy with the residual analysis

  • Impossible to find similar parts? I’d say you must look harder.

    Also, don’t fall into the amateur trap of doing an MSA on the scales involved…technique will be too important to evaluate the measurement SYSTEM.

  • @Heiko-Demski — Have you checked out our Implementation section of content? http://www.isixsigma.com/implementation/

    Some of the categories of content include topics that sound as if they’d be of interest to you: change management, deployment structure and success factors (and more!)

  • Chris Seider replied to the topic Certification in the forum General 2 weeks, 5 days ago

    You’ll get potentially lots of input on this but it’s been covered in other strings on this blog and I’m sure Katie will find you one. Certification by itself isn’t given nearly the credence that certification with project experience. If you can show how you used Lean Six Sigma tools and the progression of those tools to drive a business result…[Read more]

  • Good detailed response.

    1. I don’t like to encourage p-charts for the use as you’ve indicated.
    2. I didn’t call YOU a racist but suspect we could agree on nuances and word choice better in person over a […]

  • 1. You used a p-chart inappropriately. P-charts like all control charts are a tool meant to see if a process is changing over time–not categories. A more appropriate analysis tool is a chi-squared analysis for […]

  • @WBH, then don’t use the term belts in your organization. Easily handled but I’m not sure if the folks in the martial arts equate their talents as offensive…more defensive.

  • You can use your two rigs as 2 “operators” in the gage R&R study. I always conservatively question when someone says the operator doesn’t have an impact on the measurement–I’ve seen too many surprises.

  • Load More

Login Form