iSixSigma

Cpk for a service process

Six Sigma – iSixSigma › Forums › Old Forums › General › Cpk for a service process

Viewing 14 posts - 1 through 14 (of 14 total)
  • Author
    Posts
  • May 30, 2006 at 5:17 pm #43573

    Nani
    ★ 10 Years ★
    Participant

    How can I measure Cpk for a service process? Can we get a Cpk control chart of a “mistakes” made in a process like a call center or other “service processes”?

    0
    May 30, 2006 at 5:35 pm #138384

    Heebeegeebee BB
    ★ 10 Years ★
    Participant

    What are your associated KPIV’s/KPC’s/KQC’s?
    Do you have a target, goal or metric?
    If so, what is it/what are they?
    Are you hitting that target with repeatability and reproduceability?
    Does your data exhibit tight “shot groups”?
    get my drift? (no pun intended)

    0
    May 30, 2006 at 6:09 pm #138388

    Elbrin
    ★ 10 Years ★
    Participant

    Cpk requires a clearly defined spec. limit, most service processes do not have this.  If it is truely important to the overall performance of the orgainzation, get your leadership, better yet your customer, to establish these limits.  If not quit doing it, it’s waste.  Either way it requires leadership to make a decision, good luck.

    0
    May 30, 2006 at 6:46 pm #138389

    Pepe
    ★ 10 Years ★
    Participant

    As you know the services are catloged like a discrete data, and not follow a normal distribution, for that is not posible to get a CPk at first time.
    You need to convert the discrete data to continuos and be carefully with this.
    Saludos,

    0
    May 30, 2006 at 7:33 pm #138392

    Romel
    ★ 10 Years ★
    Member

    You can use Cpk for continuous data metrics on a call center like AHT, among others. If you intend to monitor defects on calls, you can only use the accuracy or yield formula which is good outs/inputs then compare the results of the yield in the 6sigma Cpk table since the data is more on attributes. You can also use control charts on both.

    0
    May 30, 2006 at 8:36 pm #138395

    Bucktheblackbelt
    ★ 10 Years ★
    Participant

    What do you do if you have non-normal data due to gauge resolution (i.e., measuring weights in grams but the scale’s resolution is only in increments of 5) which will give the appearance of multi-modal, non-normal data.  Can I still manually calculate to determine what percentage of samples fall within 1-, 2-, 3- standard deviations from the mean? Of course I check for the control of variance first before determining distribution profile. RIGHT?!

    0
    May 30, 2006 at 9:29 pm #138397

    Heebeegeebee BB
    ★ 10 Years ★
    Participant

    Elberin and Peppe,
    Show me the data that indicates that “most” service (read: Transactional) processes don’t have firm specs or are limited to Discrete data…   It may be true in Nani’s case, but to make such sweeping statements really ppoints to your limited experience with Six Sigma in transactional environments.
    What a load of Bovine Excretory Material!!!
    Nani, answer the questions I posted for yourself and then the matter should become clear.
     

    0
    May 31, 2006 at 7:54 am #138417

    Peppe
    ★ 10 Years ★
    Participant

    Heeb…, I’m often wrong, but  Pepe isn’t me.  Peppe (=joe) is Italian name, Pepe (?) isn’t.
    Rgs, Peppe 

    0
    May 31, 2006 at 4:00 pm #138435

    Heebeegeebee BB
    ★ 10 Years ★
    Participant

    Peppe, My bad…Sorry about the confusion.
    My note was intended for “Pepe”.
     

    0
    June 1, 2006 at 9:54 pm #138526

    Nani
    ★ 10 Years ★
    Participant

    Yes, we have metrics and goals.  We have as a goal 4 defects per month and we measure how many mistakes we make per month so I can graphic these.  can we use these numbers for a Cpk calculation?

    0
    June 1, 2006 at 10:00 pm #138527

    Nani
    ★ 10 Years ★
    Participant

    Thanks!  Where can I get a 6sigma Cpk table?

    0
    June 1, 2006 at 10:04 pm #138528

    Nani
    ★ 10 Years ★
    Participant

    Heebeegeebee,
    Thanks for your support!  I wanna clarify the process I’m working on:
    We make set ups for products in a computer system, which means we have to be really carefully in our daily job, any mistake could mean a big problem for production area.  So we want to see if our process is capable to work with only 4 mistakes per month.  We do measure every mistake found.  Can I get a Cpk for this? Or is there any other way to see if my process is capable? Thanks again :)

    0
    June 1, 2006 at 11:14 pm #138531

    Heebeegeebee BB
    ★ 10 Years ★
    Participant

    Are your measurments discrete or continuous?
    Do you have a defined spec?
    You need to make sure you have continuous data.   Cp/Cpk is all about accuracy/precision.
    Are you hitting the target (spec) with lose shot goroups, or are you maintaining tight shotgroups, centered on the target?
     

    0
    June 1, 2006 at 11:16 pm #138532

    Heebeegeebee BB
    ★ 10 Years ★
    Participant

    google it.
    -or-
    Use the search function above (in the right-hand corner)

    0
  • Author
    Posts
Viewing 14 posts - 1 through 14 (of 14 total)

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