1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Bayesian Stats - Finding a Posterior Distribution

  1. Mar 3, 2013 #1
    1. The problem statement, all variables and given/known data
    Let x be the number of successes in n independent Bernoulli trials, each one having unknown probability θ of success. Assume θ has prior distribution θ ~ Unif(0,1). An extra trial, z, is performed, independent of the first n given θ, but with probability θ/2 of success. Show that [itex]f(\theta | x,z=0) = c(\theta^x (1-\theta)^{n-x} + \theta^{x}(1-\theta)^{n-x+1})[/itex] where [itex]c = \frac{1}{B(x+1,n-x+1)+B(x+1,n-x+2)}[/itex]

    2. Relevant equations

    3. The attempt at a solution
    [itex]f(\theta | x,z=0) \propto f(x,z=0|\theta)f(\theta) = f(x|\theta)f(z=0|\theta)f(\theta) = \theta ^x (1-\theta)^{n-x} (1 - \frac{\theta}{2})[/itex]
    But from here, I can't seem to get it into the desired form, leading me to think I've done something incorrect. Where am I going wrong?
  2. jcsd
  3. Mar 3, 2013 #2


    User Avatar
    2017 Award

    Staff: Mentor

    ##(\theta^x (1-\theta)^{n-x} + \theta^{x}(1-\theta)^{n-x+1})## can be written as ##2 \theta^x (1-\theta)^{n-x}\left(1-\frac{\theta}{2}\right)##
    I think if you calculate f(\theta | x,z=1) and normalize both properly, you get the correct result.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted

Similar Discussions: Bayesian Stats - Finding a Posterior Distribution
  1. Stats distribution (Replies: 1)