iSixSigma

Project Selection Performance

Six Sigma – iSixSigma Forums Old Forums General Project Selection Performance

Viewing 5 posts - 1 through 5 (of 5 total)
  • Author
    Posts
  • #29481

    Patrick
    Participant

    I would like to measure our ability to select six sigma projects.  After a project ranking process based on potential benefits, resources, alignment with business priorities, etc, we came up with a ranking of potential projects.  This ranking is then submitted to a management committee for review and action.  What we would like to do is follow up on the validity of this ranking system by comparing it to the “actual” project launches.  For example, let’s say we recommended projects A,B,C, and D to be launched this year.  At the end of the year, projects B,C,E, and F were launched: we would have 2 out of 4 potential defects.  For the subsequent year we recommended projects E, F,G, and H but G,H,I and K were launched.  Then out of cumulative 8 projects, we now have only 2 defects.  I want to keep this as simple as possible, can anyone think of a better way?

    0
    #75676

    Mike Carnell
    Participant

    Patrick,
    In your question you mentioned that you evaluate projects in terms of potential benefits, resources, etc. and then rank them and submit them to management. I would assume the evaluation process makes the management decision fairly simple. It would seem that the ability to select projects is a function of how well the projects are characterized prior to ranking. Maybe the way to improve the selection process is to track the initial evaluation and compare it to the result. It could tell you where the evaluation process is working well and where it isn’t. If the process improves the improvement should flow through.
    I just seems like a gage study (a little unconventional but still a measurement system).
    You might get a little Hawthorne effect on the evaluations if people realize they will be evaluated for accuracy later on.
    This is just an opinion. Good luck.

    0
    #75686

    Patrick
    Participant

    That’s interesting, Mike.  I may conduct a Gage R&R to see the results of the selection (by management) versus the evaluation (by the team).  However, I still want to attribute a sigma value to the ranking process itself so that I can determine whether or not this evaluation is any better than the “unstructured” way projects were selected in the past…
    Thanks again for your input.
    Patrick

    0
    #75711

    Mike Carnell
    Participant

    Patrick,
    I am really just shooting from the hip on this one but I can see applications for hypothesis testing, doing a one sample t test using the projected evaluations to the actual. You might be able to do a sigma calculation if you treat the evaluation as a target value and do the analysis similar to the Taguchi loss function. You might even be able to modify the Cpm calculation in Minitab to help (it will give you a capability but it will not be in terms of a sigma). It is going to depend on how you treat some of this. People will have to understand that sandbagging and rolling in later with big dollars is hurting the evaluation process as much as overestimating it.
    I think your plans to evaluate the selection process is great. Not enough of the people involved in the selection process are being held accountable for the data or lack of data they generate or use.
    Let us know how it works out.

    0
    #75845

    Mark Hammond
    Participant

    Patrick,
    Just my thought …  It sounds like you are going to have a such a significant improvement that presenting the data alone may lead an objective viewer to make the obvious connection.  If they are not objective viewers, analysis like GR&Rs will help, but you are still going to have all kinds of smoke thrown up anyway.
    On the other hand, if you are not likely to have arguements — get started.  You will not have data to analyze for a year after you change methods.  The last thing you or your sponsors want is to see a long delay to figure out an analysis method for what they see as a no brainer improvement.  As the gambler said, “There’ll be time enough for counting when the dealing done.”

    0
Viewing 5 posts - 1 through 5 (of 5 total)

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