iSixSigma

Collecting Historical data

Six Sigma – iSixSigma Forums Old Forums General Collecting Historical data

Viewing 9 posts - 1 through 9 (of 9 total)
  • Author
    Posts
  • #36122

    Prithvirajan
    Participant

    My project is – Reduction of prototype development leadtime. The shift in the data collection method (from manufacturing to transactional) is quite big. I have sarched the forum for “analysing transactional data” , tips were available to analyse data but I could not find any for collection of data.
    The problem is the data which I have is very erratic and most of the data is missing or it was not there. Whatever data is available , I am not convinced that I can rely on this data for some analysis. 
    Is it really necessary to rely heavily on Historical data ?
    or is this the only way to analyse?
    Is there any other way with which I can proceed with my project?
    { All my SSV are measurebles in terms of days/time}
    Please Help.
    Regards,
    Prithvi
     
     
     
     
     

    0
    #103260

    SD
    Member

    Prithvi,
    analyzing historical data is one of the most critical steps.  If you are trying to do a project to implement a change in your end results, then you definetly require historical data.  I did one project sometime back on reducing past due buckets and I was facing the same problem.  what i did was, i took out an average and went ahead….i was able to get a decent result which could have been better if i had the proper data…i am not sure if you can risk it…but you can still try….
    thanks,

    0
    #103262

    Prithvirajan
    Participant

    Dear SD,
         Thanks, for your reply.
    The Average part is a bit unclear to me . What I understood was , take the lead time taken for a particular process/step for the entire product range and then take the average, is this what you had meant??.( & make that average as your base line……)
    Is it possible for you to send me your project for my reference?? If it is confidential atleast your Data collection part.

    0
    #103263

    SD
    Member

    Prithvi,
    you are more or less at par with what i am talking about. If you can send me your email, I can send across the data analysis to you and can highlight the changes you might want to make. However, i wont be able to send across the entire report. confidential. But data I can,
     
    thanks,
    SD

    0
    #103265

    Prithvirajan
    Participant

    My email Id is [email protected].
    Regards,
    Prithvi

    0
    #104051

    RonJ
    Member

    Prithvi,
    We

    0
    #104052

    RonJ
    Member

    Prithvi,
    We a

    0
    #104054

    RonJ
    Member

    Prithvi,
    Sorry about the previous posts; don’t know what happened. We did a similiar project and found the historical data that existed was minimal and did not go into the detail that we needed. So, we identified the current steps in our process and the key players then created a database and had the key players enter the completion date for their process. We collected data for about 3 months then began analysis. We did find that some of our “folklore” was incorrect … bottlenecks were other than where we expected them.
    That’s what we did.
    Good Luck

    0
    #104191

    Jonathon L. Andell
    Participant

    You have good instincts.
    It is crucial to establish the “integrity” of a data stream before you analyze what it might tell you about a process. If you have a valid reason to think that historical data are unreliable, you are right to dusregard it. If there never was a dependable Operational Definition for historical data entry, it just might be useless.

    0
Viewing 9 posts - 1 through 9 (of 9 total)

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