iSixSigma

Need Help on Preparing DCP and Further Analysis

Six Sigma – iSixSigma Forums General Forums Tools & Templates Need Help on Preparing DCP and Further Analysis

Viewing 7 posts - 1 through 7 (of 7 total)
  • Author
    Posts
  • #54146

    Vineet Aggarwal
    Participant

    Hi,

    I am working on a project for quality of video / audio used in media industry whose data is capture in the form of ‘Very Good’, ‘Good’, ‘Average’, ‘Poor’, ‘Very Poor’. The Customer wants the quality atleast Good or above. A ppt is attached for having an idea. Define phase has been prepared as of now. Would appreciate if someone could guide how this discrete data can be analysed further and if the DCP (What to Measure?) that is prepared as of now is ok.

    Thanks in advance

    0
    #193914

    Vineet Aggarwal
    Participant

    If I get little more specific I need help on preparing operational definition, performance standard, defect definition. Should I convert the raw data into date wise form and prepare all the above based on that or I can use words in these components for example, Performance standard :- Quality of footage >= Good.
    Please guide.

    0
    #193921

    Chris Seider
    Participant

    Can I suggest someting? You aren’t being aggressive enough in your vision as the team leader. Shoot for at least a reduction of 50% of footage that has to be redone. If you think that’s too undoable, consider gathering data on REASONS for reediting of footage and then shoot for a 68% reduction of the REASON. When I mean reason, I mean the category of defect as would be found if you pareto’d the count of reasons.

    You will have to pareto the reasons anyways even if you want to reduce the % of footage to be reedited. 10% is not ambitious. Shoot for breakthrough results.

    My 2 cents. FYI, grats on opening up your project to online mentoring.

    0
    #193922

    Chris Seider
    Participant

    I meant 68% reduction of the top reason. My emphasis of reason means the defect type and NOT the root causes.

    0
    #193928

    Vineet Aggarwal
    Participant

    Hi Chris,

    I completely agree with you and shall try to redefine my goal. Actually the reason behind setting this non-aggressive goal was that this project is being done for this team first time on experiment basis in which they want to have an idea about the concept of how DMAIC works and see the results. My aim with this project is to complete it as soon as possible. Setting higher goal would have lead to take more time to complete. Further buy-in and support is dependent on this project. Another limitation is that we have limited data in hand at present and planning to collect more detailed data will come under debate unless concerned people see their own benefit in this.
    Would appreciate if you could guide how to achieve higher goal considering current time and resource limitation and gain buy-in too.

    0
    #193934

    Prabhu V
    Participant

    Hi,
    Based on your reply, I would like to suggest the following:-

    Assumption: Quality related parameters are high important than quantity related parameters.

    Op. def.: In order to decide the footage quality you may define the binary points for your quality definitions like Ok or Not Ok.

    Like if any footag’s Sharpness – Ok, Contrast – Ok, Brightness – Ok, Colour – Ok, Proper frame – Ok, etc. then the Quality rating of that footage is Very Good otherwise depends of priority you may define the Op. def for Good, Average, Poor and Very Poor. (the other parameters like Defect def etc may be same with your observation)

    By doing so you can get quality rating on quality definition for your current defects (rejected footages), On which data you may do suitable analysis as you have mentioned (like Parato etc.,)

    I hope this will project more information with less efforts.

    0
    #196727

    john
    Guest

    Just a couple things to consider: (1) Your initial focus should be on answering the following suite of questions..what do we make…who do we make it for…what do they want…why to they want it. You really want to understand the “why” here before you get into anything else, as this opens up a whole spectrum of opportuntiy that you will miss if you simply ask “How can we improve the quality of the existing product?” (2) If you answer the following questions and it leads you back to your initial desire to simply improve “quality” in the current product offering, then you should focus on how your customer is defining value and how they are measuring it, both in terms of operational definitions and method. Seek as much precision as possible, meaning using continuous data as much as possible (ie avoid 5 point Likert scales whenever possible) and using a CTQ drilldown exercise to get past nebulous terms like “better quality”, “improved sound”, “less distortion”, etc. Now you know what you are currently making for each of your key customer segments, what each segment wants (and doesnt want) from your current product, and why they are buying it. In addition, you have a precise and accurate understanding of how THEY are defining and measuring your current offering which you will use to identify spec limits and improvement goals based on current process capabilities/entitlement. Best of luck

    0
Viewing 7 posts - 1 through 7 (of 7 total)

You must be logged in to reply to this topic.