I'm slightly confused with when I should use a Poisson distribution.

If I have a case where:

  • I have the sample mean of the number of events happening in a particular day (say it is 1.27).
  • Five days are randomly selected from a sample.

Can I use the Poisson process to find the probability that the sample mean now is more than 0.3?

  • $\begingroup$ I'm not completely sure I understand your description, but it sounds like you're trying to construct some kind of interval for a Poisson mean for some unobserved time period based on a sample for some other time period. Is that correct? $\endgroup$
    – Glen_b
    Commented Feb 20, 2014 at 13:53
  • $\begingroup$ Thanks for the speedy reverts Glen. I'm not quite sure if Poisson is the game here. What i have on hand is a probability distribution of the number of events happening on a particular day - from which I've worked out the mean. What i want to find out is the probability that this mean is going to be more than 0.3 if 4 days are randomly selected from a sample. Is the sample size important? $\endgroup$ Commented Feb 20, 2014 at 14:01
  • $\begingroup$ Oh, I think I see. You've been given some probability distribution for a random variable representing the number of events per day. You computed the mean of that distribution (a population mean). Then you start talking about a sample mean over several days of observations. Is that it? (There are several strange things about your wording that make it especially confusing.) $\endgroup$
    – Glen_b
    Commented Feb 20, 2014 at 14:30
  • $\begingroup$ Spot on Glen! You're really brilliant. Yeps. Thats what Im trying to self learn and solve $\endgroup$ Commented Feb 20, 2014 at 14:35
  • 1
    $\begingroup$ I will say that there's no hint in the little information you give that the Poisson comes into this at all. A little more detail might clarify that. $\endgroup$
    – Glen_b
    Commented Feb 20, 2014 at 22:33

1 Answer 1


Here is how I understand the Poisson distribution: the distribution is used to describe events happening over a given range(i.e. a time period)

For the PDF distribution you can use it for things like the probabilities for:

  • Number of mutation in DNA strands per unit of time.

  • N peaks in stock in a over a given period of time

For the CDF distribution you can use it for the probabilities of:

  • 3 or less robberies in a given length of time.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.