+ Reply to Thread
Results 1 to 13 of 13

Thread: Combine Normal Distributions

  1. #1
    Points: 972, Level: 16
    Level completed: 72%, Points required for next Level: 28

    Posts
    1
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Combine Normal Distributions




    How do you combine multiple normal distributions with different standard deviations and means, to produce one representative normal distribution?

    Thank you!

  2. #2
    R purist
    Points: 35,103, Level: 100
    Level completed: 0%, Points required for next Level: 0
    TheEcologist's Avatar
    Location
    United States
    Posts
    1,921
    Thanks
    303
    Thanked 607 Times in 341 Posts

    Re: Combine Normal Distributions

    Depends on how you want to combine them. So you could have supplied a little bit more information (it helps to read this).
    I am going to assume you mean the sum of multiple normally distributed random variables. In that case you need to look into the Normal sum distribution.
    The true ideals of great philosophies always seem to get lost somewhere along the road..

  3. #3
    TS Contributor
    Points: 7,284, Level: 56
    Level completed: 67%, Points required for next Level: 66

    Location
    MD, USA
    Posts
    465
    Thanks
    8
    Thanked 11 Times in 11 Posts

    Re: Combine Normal Distributions

    This is assuming n > 30 for those distributions?

  4. #4
    Devorador de queso
    Points: 95,889, Level: 100
    Level completed: 0%, Points required for next Level: 0
    Awards:
    Posting AwardCommunity AwardDiscussion EnderFrequent Poster
    Dason's Avatar
    Location
    Tampa, FL
    Posts
    12,937
    Thanks
    307
    Thanked 2,630 Times in 2,246 Posts

    Re: Combine Normal Distributions

    Why would you need n > 30 if you knew the original populations were normal?

  5. #5
    R purist
    Points: 35,103, Level: 100
    Level completed: 0%, Points required for next Level: 0
    TheEcologist's Avatar
    Location
    United States
    Posts
    1,921
    Thanks
    303
    Thanked 607 Times in 341 Posts

    Re: Combine Normal Distributions

    Quote Originally Posted by Outlier View Post
    This is assuming n > 30 for those distributions?
    "combine multiple normal distributions" does not sound like combining samples rather it sounds like combining pdfs to me.
    The true ideals of great philosophies always seem to get lost somewhere along the road..

  6. #6
    Devorador de queso
    Points: 95,889, Level: 100
    Level completed: 0%, Points required for next Level: 0
    Awards:
    Posting AwardCommunity AwardDiscussion EnderFrequent Poster
    Dason's Avatar
    Location
    Tampa, FL
    Posts
    12,937
    Thanks
    307
    Thanked 2,630 Times in 2,246 Posts

    Re: Combine Normal Distributions

    Oh ok. Now I see where Outlier was coming from. I'm with The Ecologist on this one though. Now let's just hope they do mean the sum of normally distributed random variables and not the product!

  7. #7
    TS Contributor
    Points: 7,284, Level: 56
    Level completed: 67%, Points required for next Level: 66

    Location
    MD, USA
    Posts
    465
    Thanks
    8
    Thanked 11 Times in 11 Posts

    Re: Combine Normal Distributions

    I guess I assumed you can have normal distributions with less than 30 samples. I'm probably mixing up several concepts. I'd think a distribution with more samples than another one should be weighted differently when combined, like a weighted average. Maybe not.

  8. #8
    Points: 2, Level: 1
    Level completed: 3%, Points required for next Level: 48

    Posts
    1
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Combine Normal Distributions

    I dont get this. If you have two distributions with wildly different means, and very tight standard deviations, why would you get another normal distribution? It seems like you would just get a bimodal distirbution, or two normal pdfs sepearated by a gap.

  9. #9
    Omega Contributor
    Points: 38,396, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,001
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: Combine Normal Distributions

    Curious of some examples where you may want to add two normally distributed variables together.

  10. #10
    Devorador de queso
    Points: 95,889, Level: 100
    Level completed: 0%, Points required for next Level: 0
    Awards:
    Posting AwardCommunity AwardDiscussion EnderFrequent Poster
    Dason's Avatar
    Location
    Tampa, FL
    Posts
    12,937
    Thanks
    307
    Thanked 2,630 Times in 2,246 Posts

    Re: Combine Normal Distributions

    The OP never made their way back to this thread so I don't think we'll get clarification on what they actually were interested in. For everybody else I think a big problem so far is that it's not sufficiently clear what you're trying to convey.

    Quote Originally Posted by SCE1000
    If you have two distributions with wildly different means, and very tight standard deviations, why would you get another normal distribution?
    It depends on what you're doing to combine them. If we're just assuming some mixture distribution then your intuition seems fine. If we're taking the sum of two normal random variables with the given means and variances then the resulting variable will have a normal distribution.

    @hlsmith - I'm not sure what you were asking exactly.
    I don't have emotions and sometimes that makes me very sad.

  11. #11
    Omega Contributor
    Points: 38,396, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,001
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: Combine Normal Distributions

    Just curious what two variables they were thinking of adding together.

    Trying to wrap my head around if they would be colinear or not. Or what types of scales or units they may have.

    Also trying to picture how two variables might be uncorrelated and how their sums would look normal. Say I have heights and I am adding it to some made up variable where those with centrally located heights had extremes values on the other variable. An example would help me understand how in this scenario the sum would be normal. I get the addition of correlated variables but don't intuitive get the sum of two totally different scales.

  12. #12
    Super Moderator
    Points: 13,151, Level: 74
    Level completed: 76%, Points required for next Level: 99
    Dragan's Avatar
    Location
    Illinois, US
    Posts
    2,014
    Thanks
    0
    Thanked 223 Times in 192 Posts

    Re: Combine Normal Distributions

    hlsmith: Look at the theory underlying the ratio of a sum of U-Statistics and see how this ratio is asymptotically normally distributed.

  13. #13
    Points: 3,051, Level: 34
    Level completed: 1%, Points required for next Level: 149

    Posts
    8
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Combine Normal Distributions


    I ran across this thread while looking for what I believe was the OP's intended question. Below I provide an example and question with that assumption. In my case, I'm interested in knowing how to combine, say, three independent normally distributed variables. That is, I want to know the resultant additive value of three consecutively-occurring, independent, normal functions at a certain confidence.

    Let's say a different normal curve describes each of the following activities, in minutes: how long it takes to drive to the dentist (mean=5, stdev=2), how long I spend there (60, 15), and how long it takes to drive back home, fighting traffic (15, 5).

    If I wanted to know with 95% confidence that I had allocated enough time in my schedule to accommodate the total trip, how would I find the value of the joint distribution at +2 standard deviations?

    This is a simplified but relevant example of what I'm trying to solve, and I think it's in the spirit of the OP's question. Thanks in advance for any help!

+ 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