iSixSigma

Help… Data Analysis Needed-Quickly

Six Sigma – iSixSigma Forums Old Forums General Help… Data Analysis Needed-Quickly

Viewing 7 posts - 1 through 7 (of 7 total)
  • Author
    Posts
  • #152995

    Adam
    Participant

    Create a casue and effect diagram (fishbone) where your effect is a late shipment.ย  For all of your causes, find the once that first appear to be most significant and write them down.ย  Then, for every time you had a late shipment, find out what the cause was.ย  For instance, if you list a cause as “place of shipment”, for every time you had a late shipment, you would list where it was shipped from.ย  Compare these answers to what the answers are when the shipment was not late and see if you can see a pattern.

    0
    #152996

    greeny
    Participant

    Thank you so much for the quick response. This data is all shipments from 2006 so a retro ishikawa won’t work. Also, I’m looking for a statistical analysis.

    0
    #152999

    BTDT
    Participant

    Woops – I mean box plot – thanks Adrian.

    0
    #46363

    greeny
    Participant

    I have shipping data that needs to be analyzed and I don’t know what to do. The data set includes: port of departure, port of entry, country of departure and arrival, custom’s release date, transpo carrier, and days late (this is the response variable). I need to determine what causes shipments to be late. The factors aren’t continuous. I’m having a brain fart on what to do. Any help is greatly appreciated.

    0
    #153001

    Deep
    Participant

    greeny:
    Copy paste the data here or send me a copy. i can have a look at it and try to make some conclusion out of the data…
    microsoft_sucks0

    0
    #152997

    Adrian P. Smith
    Participant

    Hi,
    I would do boxplots of days late vs each of the potention Xs (port of departure, port of entry, country of departure, country of arrival and transpo carrier).
    The customs release date is harder to analyse since it of course depends on the shipment date. You need to know the shipment date and work out the “cycle time” from shipment to customs release and then analyse thata against the same list of variables.
    You may also want to look as some of the Xs in pairs, e.g. transpo carrier + country of arrival to see whether performance of some carriers varies by country of arrival. Other obvious pairings are port of departure/port of entry and country of departure/arrival.
    Other graphs which may give you some clues would be main effects plot and a multivari plot.
    Hope this helps.
    Rgds,Adrian
    http://adrian.smith.name

    0
    #152998

    BTDT
    Participant

    Greeny:Pareto charts by subgroup. T-test or ANOVA to confirm.Cheers, BTDT

    0
Viewing 7 posts - 1 through 7 (of 7 total)

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