iSixSigma

Subgrouping Issue

Six Sigma – iSixSigma › Forums › Old Forums › General › Subgrouping Issue

Viewing 4 posts - 1 through 4 (of 4 total)
  • Author
    Posts
  • September 6, 2007 at 3:17 pm #48053

    Shimuk
    ★ 10 Years ★
    Member

    It is said that subgroups should be formed from time ordered sequence of observations and should depict “small windows of time”.
    I have daily 1 observation of a continuous type taken over 3 months. If I want to understand whether the process is in control, should I do an I-chart with 90 data points OR should I do subgroups of days (Mon….Sun) and do an X bar chart with 21 data points ? X bar is obviously more sensitive and robust…..but am I not violating the “window of time” theory ?

    0
    September 6, 2007 at 3:30 pm #160838

    Evil
    ★ 10 Years ★
    Participant

    Shimuk,
    My suggestion would be to use a X-mR chart with rules 1, 5 & 6 turned on in MINITAB. That will give you reasonable sensitivity to special cause.If you can convince yourself that the variables in your process do not change day to day over the span of 4-5 days, then you won’t desensitize the chart by including more variation within the subgroups.hope that helps…
    E

    0
    September 6, 2007 at 3:46 pm #160840

    jediblackbelt
    ★ 10 Years ★
    Participant

    Agreed.  I would think about it this way.  If you use the Xbar chart with 1 week as your subgroup then you will tell difference from week to week and if there is a within variation, correct?
    However, you won’t be able to see if it is an issue that happens every Wednesday and then drops down.  The X-bar range will wash that out and your average could stay the same.  I am just a bigger fan of the XmR unless you really understand your Xbar chart. 
    my 2cents.

    0
    September 6, 2007 at 4:00 pm #160841

    Shimuk
    ★ 10 Years ★
    Member

    As I mentioned in the post, I am envisaging subgrouping by day and NOT by week……that will give me 3 data points for each “day” over the span of 3 months data. Contrary to washing out, it should highlight the day to day variation if any. However, this way, the subgroups are not time ordered….I am picking 4 observations each separated by a week to create a subgroup.

    0
  • Author
    Posts
Viewing 4 posts - 1 through 4 (of 4 total)

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