+ Reply to Thread
Results 1 to 9 of 9

Thread: What is Experimental Design

  1. #1
    Points: 1,821, Level: 25
    Level completed: 21%, Points required for next Level: 79
    Buckeye's Avatar
    Location
    Ohio
    Posts
    102
    Thanks
    31
    Thanked 4 Times in 4 Posts

    What is Experimental Design




    Hi all, it's been awhile since I've posted. Is there a concise description of what experimental design is in regards to statistics? I'm trying to make sense of a course description. Is it simply the way to conduct an experiment in order to gather data? Please elaborate.
    "I have discovered a truly remarkable proof of this theorem which this margin is too narrow to contain." Pierre de Fermat

  2. #2
    Omega Contributor
    Points: 38,334, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    6,998
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: What is Experimental Design

    Well, there has to be an experiment, AKA an intervention of some type. So you have observational studies, where you don't intervene just observe and you have experimental, where some type of intervention happened. Then you can have additional descriptors, say randomized experimental design.
    Stop cowardice, ban guns!

  3. The Following User Says Thank You to hlsmith For This Useful Post:

    Buckeye (12-20-2016)

  4. #3
    Points: 1,821, Level: 25
    Level completed: 21%, Points required for next Level: 79
    Buckeye's Avatar
    Location
    Ohio
    Posts
    102
    Thanks
    31
    Thanked 4 Times in 4 Posts

    Re: What is Experimental Design

    Thank you hlsmith. I have another question. Perhaps someone can interpret this quote by Ronald Fisher:

    "To call in the statistician after the experiment is done may be no more than asking him to perform a postmortem examination: he may be able to say what the experiment died of."

    Does this relate to experimental design?
    "I have discovered a truly remarkable proof of this theorem which this margin is too narrow to contain." Pierre de Fermat

  5. #4
    Omega Contributor
    Points: 38,334, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    6,998
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: What is Experimental Design

    If you have ever done analyses for other people's projects this almost feels like a truism. You could partially equate this to randomization, but I think of it more as any bad study design or systematic error within a study.


    Most times a very well designed and executed study may only require a simple statistical approach (say a t-test given treatment assignment effectively randomized). Though, if investigators did something wrong our forgot to control for a confounder, then more advanced approaches are needed. I actually jokingly say something like this when someone knocks on my door asking for help after they have conducted the study. These instance may require you to think outside the box and get creative in order to identify the causal mechanism.
    Stop cowardice, ban guns!

  6. The Following User Says Thank You to hlsmith For This Useful Post:

    Buckeye (12-20-2016)

  7. #5
    TS Contributor
    Points: 14,811, Level: 78
    Level completed: 91%, Points required for next Level: 39
    Miner's Avatar
    Location
    Greater Milwaukee area
    Posts
    1,171
    Thanks
    34
    Thanked 405 Times in 363 Posts

    Re: What is Experimental Design

    Quote Originally Posted by Buckeye View Post
    Thank you hlsmith. I have another question. Perhaps someone can interpret this quote by Ronald Fisher:

    "To call in the statistician after the experiment is done may be no more than asking him to perform a postmortem examination: he may be able to say what the experiment died of."

    Does this relate to experimental design?
    In addition, I see under-powered studies due to inadequate sample sizes, restrictions on randomization, failure to block when appropriate, using a low resolution design with a lot of confounding/aliasing, botched runs and more.

  8. The Following User Says Thank You to Miner For This Useful Post:

    Buckeye (12-20-2016)

  9. #6
    TS Contributor
    Points: 14,811, Level: 78
    Level completed: 91%, Points required for next Level: 39
    Miner's Avatar
    Location
    Greater Milwaukee area
    Posts
    1,171
    Thanks
    34
    Thanked 405 Times in 363 Posts

    Re: What is Experimental Design

    Quote Originally Posted by Buckeye View Post
    Hi all, it's been awhile since I've posted. Is there a concise description of what experimental design is in regards to statistics? I'm trying to make sense of a course description. Is it simply the way to conduct an experiment in order to gather data? Please elaborate.
    Here is a clear, yet simple explanation of experimental design.

  10. The Following User Says Thank You to Miner For This Useful Post:

    Buckeye (12-20-2016)

  11. #7
    Fortran must die
    Points: 58,790, Level: 100
    Level completed: 0%, Points required for next Level: 0
    noetsi's Avatar
    Posts
    6,532
    Thanks
    692
    Thanked 915 Times in 874 Posts

    Re: What is Experimental Design

    Experimental design is how you set up the experiment in the first place, which in turn will set limits on how you can analyze the results. For example you chose whether to have random assignment or not. If you do many threats to validity will not exist, if you do not then you have to deal with those threats. Commonly it is not possible to have random assignment or even to sample randomly. So you create designs, or use existing ones to address the limitations that result as best you can. DOE also tells you what problems are created by different experimental designs (and what statistics are appropriate to use).

    DOE has little to do per se with statistics. In fact you could do DOE without any quantitative methods at all - although you probably would not. The best known work on this is "Experimental and Quasi-Experimental Designs for Research" by Campbell and Stanley. Its not an easy read, there is little that is simple about DOE.

    https://www.amazon.com/Experimental-.../dp/1111071861
    "Very few theories have been abandoned because they were found to be invalid on the basis of empirical evidence...." Spanos, 1995

  12. The Following User Says Thank You to noetsi For This Useful Post:

    Buckeye (12-20-2016)

  13. #8
    TS Contributor
    Points: 12,227, Level: 72
    Level completed: 45%, Points required for next Level: 223
    rogojel's Avatar
    Location
    I work in Europe, live in Hungary
    Posts
    1,470
    Thanks
    160
    Thanked 332 Times in 312 Posts

    Re: What is Experimental Design

    Hi,
    I would recommend "Statistics for Experimenters" by Box and Hunter. Not so heavy on theory as most and very practical.

    regards

  14. #9
    TS Contributor
    Points: 14,811, Level: 78
    Level completed: 91%, Points required for next Level: 39
    Miner's Avatar
    Location
    Greater Milwaukee area
    Posts
    1,171
    Thanks
    34
    Thanked 405 Times in 363 Posts

    Re: What is Experimental Design


    Quote Originally Posted by rogojel View Post
    Hi,
    I would recommend "Statistics for Experimenters" by Box and Hunter. Not so heavy on theory as most and very practical.

    regards
    Excellent recommendation. It is written from an Industrial Statistics perspective, and may not address common practices in other disciplines. Also, industrial experiments tend to be less noisy and consequently easier to replicate and verify.

    A second recommendation is Quality Improvement Through Planned Experimentation. If you can only choose one, go for Statistics for Experimenters.

  15. The Following 2 Users Say Thank You to Miner For This Useful Post:

    GretaGarbo (12-21-2016), rogojel (12-21-2016)

+ 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