+ Reply to Thread
Results 1 to 4 of 4

Thread: Understanding and Calculating Credible Intervals

  1. #1
    Points: 104, Level: 2
    Level completed: 8%, Points required for next Level: 46

    Posts
    2
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Understanding and Calculating Credible Intervals




    Hi,

    I am having trouble understanding and calculating credible intervals. My goal is to understand confidence intervals vs. credible intervals.

    From what I understand, the credible interval is a range of values (parameter?) that is fixed (not really) based on the prior probability distribution. I am assuming that the posterior probability distribution is the same as the prior probability distribution.

    I found something that I am trying to understand about credible intervals,but I got a bit lost in the wording. Is there an equation that helps calculate credible intervals?

    I am trying to decipher what the attachment means.
    Attached Images  

  2. #2
    TS Contributor
    Points: 18,889, Level: 87
    Level completed: 8%, Points required for next Level: 461
    CowboyBear's Avatar
    Location
    New Zealand
    Posts
    2,062
    Thanks
    121
    Thanked 427 Times in 328 Posts

    Re: Understanding and Calculating Credible Intervals

    Quote Originally Posted by bmac1 View Post
    I am assuming that the posterior probability distribution is the same as the prior probability distribution.
    Nope, they're different. I think the first thing to do here would be to try and get a good conceptual understanding of how confidence intervals and credible intervals differ (before moving on to calculations). This thread should get you started: https://stats.stackexchange.com/ques...dible-interval
    Matt aka CB | twitter.com/matthewmatix

  3. #3
    TS Contributor
    Points: 22,408, Level: 93
    Level completed: 6%, Points required for next Level: 942
    spunky's Avatar
    Location
    vancouver, canada
    Posts
    2,135
    Thanks
    166
    Thanked 537 Times in 431 Posts

    Re: Understanding and Calculating Credible Intervals

    Quote Originally Posted by bmac1 View Post
    I am assuming that the posterior probability distribution is the same as the prior probability distribution.
    Quote Originally Posted by CowboyBear View Post
    Nope, they're different.
    My guess is that, in this very specific and extremely contrived example, OP is assuming the same function form between the prior and the posterior distribution? I mean, almost all introductory texts I've seen about Bayesian statistics take the classic "coin tossing" model example because the Beta distribution acts as a conjugate prior here.

    But related to the specific question of:

    Quote Originally Posted by bmac1 View Post
    Is there an equation that helps calculate credible intervals?
    Most of the time the answer is "no". These things are complicated and usually need to be approximated via computer methods, unless you're dealing with an example that is so contrived and simple that you can actually work it out by hand.
    for all your psychometric needs! https://psychometroscar.wordpress.com/about/

  4. #4
    Points: 104, Level: 2
    Level completed: 8%, Points required for next Level: 46

    Posts
    2
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Unhappy Re: Understanding and Calculating Credible Intervals


    Quote Originally Posted by spunky View Post
    My guess is that, in this very specific and extremely contrived example, OP is assuming the same function form between the prior and the posterior distribution? I mean, almost all introductory texts I've seen about Bayesian statistics take the classic "coin tossing" model example because the Beta distribution acts as a conjugate prior here.

    But related to the specific question of:



    Most of the time the answer is "no". These things are complicated and usually need to be approximated via computer methods, unless you're dealing with an example that is so contrived and simple that you can actually work it out by hand.
    I have read that stackexchange forum many times and still don't understand the jargon.
    I think what I am having trouble with is how the prior, beta-binomial poster distribution, and the beta distribution fit in the credible interval?

    I understand that you need to know what your prior distribution is before getting the credible interval.

+ Reply to Thread

           




Tags for this 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