Q&A Bank Part 1 Q1.12(i)

Discussion in 'CT6' started by DMF, Jul 14, 2018.

  1. DMF

    DMF Member

    Hi,

    If I am not mistaken, when working out the likelihood function you multiply the function for each x in the sample - e.g. the solution 1.14 (i) multiplies the function for eact t in the sample. (I am ignoring constants as the likelihood function is used to determine the posterior).

    However, in the solution for Q1.12(i) the binomial distribution is not multiplied out for the sample of n. Why is that?

    Thanks
     
    Last edited by a moderator: Jul 15, 2018
    DMF, Jul 14, 2018
    #1
  2. John Lee

    John Lee ActEd Tutor Staff Member

    A single binomial distribution already contains a sample of n iid Bernoulli trials.

    For the situation described one binomial X~Bin(20,p) gives the likelihood.
     
    John Lee, Jul 23, 2018
    #2
  3. DMF

    DMF Member

    Thanks - this is really helpful.
     
    DMF, Jul 27, 2018
    #3
    John Lee likes this.

Share This Page