Banner Ad 1

Collapse

Announcement

Collapse
No announcement yet.

normal approximation

Collapse
This topic is closed.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • normal approximation

    can anyone here feed us with questions relating to normal approximation of discrete random variables?.when do we add or subtract the 0.5?

  • #2
    When the exam question tells you to continuity correct.

    Comment


    • #3
      Originally posted by Krieger
      When the exam question tells you to continuity correct.
      That's probably the best answer.

      From a theoretical point of view, though, think about what "continuity correction" means. If you're doing a normal approximation with a discrete distribution, that means that you're approximating a distribution that can only take on integer numbers with a distribution that can take on any numbers. So, if you want the probability of the actual distribution being between 1 and 3 inclusive, you have to notes that the normal distribution can give you a number like 0.756, which rounds to 1. So, you have to consider the number being between 0.5 and <3.5, including 0.5, but not including 3.5. That's what the continuity correction stands for.

      Comment


      • #4
        Continuity correction

        Originally posted by jagazo
        can anyone here feed us with questions relating to normal approximation of discrete random variables?.when do we add or subtract the 0.5?
        The best way to understand this is to draw the histogram of the distribution of you are approximating together with the normal distribution you are using for approximation, on top of each other, and mark the regions, for example,
        where Pr (X < number) for both. You will see why you need a correction. An explanation in detail is in my ASM manual.

        Yours,
        Krzys' Ostaszewski
        Want to know how to pass actuarial exams? Go to: smartURL.it/pass

        Comment

        Working...
        X