+ Reply to Thread
Results 1 to 2 of 2

Thread: control charts-type I error

  1. #1
    Points: 4,356, Level: 42
    Level completed: 3%, Points required for next Level: 194

    Posts
    7
    Thanks
    0
    Thanked 0 Times in 0 Posts

    control charts-type I error




    My control chart has a type I error rate of 0.0027 (3 standard deviations). I am comparing 5 months of signals to my baseline UCL. Am I inflating this error rate because I am doing multiple comparisons to the same UCL? The literature sends mixed signals. Some don't feel that a control chart is a true hypothesis test, while others would calculate the type I error to be 1-(0.9973)^5=0.013.
    Which is correct? Is the error inflated or not?

  2. #2
    TS Contributor
    Points: 13,042, Level: 74
    Level completed: 48%, Points required for next Level: 208
    JohnM's Avatar
    Posts
    1,948
    Thanks
    0
    Thanked 5 Times in 4 Posts

    I don't consider control charts a good tool for hypothesis testing - they're meant to be a visual tool (that's why they're a chart...) to allow people to detect a process shift.

    To answer your question, I feel that 0.0027 is the probability, on a 3-sigma chart, of getting a point out of control over the long run, if the process is in control.

+ Reply to Thread

           




Similar Threads

  1. Replies: 1
    Last Post: 09-12-2010, 12:10 AM
  2. Type 1 error and type 2 error
    By khhyatii in forum Statistics
    Replies: 1
    Last Post: 08-28-2009, 04:22 AM
  3. Control Charts-R
    By yon in forum Other Software
    Replies: 0
    Last Post: 07-06-2009, 07:03 PM
  4. Type I/Type II error
    By letsstophere in forum Statistics
    Replies: 0
    Last Post: 02-09-2009, 02:18 PM
  5. Replies: 2
    Last Post: 11-09-2006, 11:36 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts






Advertise on Talk Stats