+ Reply to Thread
Results 1 to 5 of 5

Thread: How to amend one variable in a Full Factorial Design during experimentation

  1. #1
    Points: 12, Level: 1
    Level completed: 23%, Points required for next Level: 38

    Posts
    3
    Thanks
    0
    Thanked 0 Times in 0 Posts

    How to amend one variable in a Full Factorial Design during experimentation




    I have a 2 level, 4 factor DOE set up for 16 runs. These are all continuous factors. I'd like to amend one of the factors to a narrower range. I am currently 10/16 runs complete and am wondering if there are any options available before I have to end up repeating the first 10 experiments. I looked into folding, but I'm not really sure if I can apply that (blocking effect from different weeks, equipment, etc). Are there computer designs that could amend the original design and produce a comparable design resolution?

    Any help is greatly appreciated!

  2. #2
    Devorador de queso
    Points: 97,410, Level: 100
    Level completed: 0%, Points required for next Level: 0
    Awards:
    Posting AwardCommunity AwardDiscussion EnderFrequent PosterActivity Award
    Dason's Avatar
    Location
    Tampa, FL
    Posts
    12,981
    Thanks
    308
    Thanked 2,639 Times in 2,254 Posts

    Re: How to amend one variable in a Full Factorial Design during experimentation

    Why do you want to make it narrower?
    I don't have emotions and sometimes that makes me very sad.

  3. #3
    Points: 12, Level: 1
    Level completed: 23%, Points required for next Level: 38

    Posts
    3
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: How to amend one variable in a Full Factorial Design during experimentation

    The range was made too wide and was studied previously. The purpose of the experiments was to narrow the ranges for better understanding of the process.

  4. #4
    Points: 12, Level: 1
    Level completed: 23%, Points required for next Level: 38

    Posts
    3
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: How to amend one variable in a Full Factorial Design during experimentation

    I'd also be interested to know how to widen a range in the middle of a full factorial DoE assuming blocking is necessary. Just in case something similar happened. I'm not sure if that's easier or harder.

  5. #5
    TS Contributor
    Points: 15,152, Level: 79
    Level completed: 61%, Points required for next Level: 198
    Miner's Avatar
    Location
    Greater Milwaukee area
    Posts
    1,185
    Thanks
    34
    Thanked 408 Times in 366 Posts

    Re: How to amend one variable in a Full Factorial Design during experimentation


    If you change the levels part way through the experiment, you should be able to analyze the experiment using GLM or regression. However, the experimental design is no longer orthogonal nor balanced, and there are repercussions associated with that. The most obvious is the loss of power. While you will probably still detect strong effects, you will likely miss all weak effects and a percentage of moderate effects. If you are trying to build a model and optimize, your model will not be as good and your optimization off as a result.

+ Reply to Thread

           




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