How to Validate Data.. Without Data

Six Sigma – iSixSigma Forums General Forums General How to Validate Data.. Without Data

Viewing 6 posts - 1 through 6 (of 6 total)
  • Author
  • #53500



    I was tasked to consolidate and personally mange a railcar ( trains ) program at a Corporate Level.

    The process of when sales orders were entered, railcars were ordered, railcars arrived, railcars loaded, and railcars delivered to customer was being performed 8 different way by 8 different warehouse/plants. Complete chaos ( hence here is where I came in ).

    It has been 1 year since I have managed this on a daily basis. I have any piece of data necessary to build some sort of analysis of how we are doing better than before.
    Which is 1000% better since I have taken over.. but without ‘data’ to validate this…

    Since I do not have ANY data from before hand, how can I quantify this change?

    Alot of process controls were put into place ( by me ), and the entire process is streamlined now.

    Any ideas??


    Robert Butler

    I’m not trying to be offensive or “cute” but if you don’t have any data how is it possible for you to say that things are 1000% better? A percentage claim of this type would mean you would have to know what you had when you started (note this is not the same thing as what you had before you started) and what you have now.



    You are EXACTLY right!
    That is exactly what I am I trying to ‘quantify’.

    I truly believe in data and numbes to validate a ‘success’..
    This project was a no brainer…

    Take 8 difference processes of how refrigerated railcars are managed, and let 1 person ( me ), at the corporate level manage it.

    I put in alot of process controls ( lead time requirements, verifying routes, verifying rail providers ).

    So can I truly consider this being a ‘project’ or just fixing the problem by consolidating the 8 process into 1?



    There’s got to be some type of historic data. Even if nobody kept track, there had to have been a recognized “average” time, maybe even a best / worst type of range. If this could be broken down for each category you identified, that’d be great. You could poll the guys in scheduling / planning and establish an estimated range. You could also ask your customers and maybe get a better estimate. Make sure it’s very clear where the information to build this range came from.

    Now, since you’ve changed the system, how long do the individual and accumulative processes take? You could even ask those customers again and make sure they are seeing the differences, too.

    You could also evaluate how many people or man-hours were required every month / week to do the monitoring previously and what was the reduction. HR might have that type information.




    Zero, other than…

    What we have always done, “aint working”.

    its not how long is it taking, it is the consistency of railcars showing up when ordered.

    So it is going to be very hard to ‘quantify’ this as a six sigma project. Just another atta boy award..





    You have not tried hard. You have accounting data behind all these transactions.

Viewing 6 posts - 1 through 6 (of 6 total)

You must be logged in to reply to this topic.