0
$\begingroup$

This may be a silly question, but what if you set p=1 in the binomial distribution probability mass function, doesn't the PMF go to zero? Does this mean the probability of k successes in n trials is zero despite the probability of success in one trial is 100%?

$\endgroup$
  • $\begingroup$ Can you give the relevant formula which you are referring to? $\endgroup$ – Andreas Sep 8 '17 at 17:24
2
$\begingroup$

There's 2 cases to look at here.

If $k < n$, we have that the probability of $k$ successes in $n$ trials will be $0$. This is because the binomial distribution parameter, $p$ refers to the probability of success for each individual trial. In the case that $p = 1$, we have that every single trial will be a success. This means that in $n$ trials we will have $n$ successes and so $k$ being less than $n$ guarantees that the probability of having $k$ successes in $n$ trials is $0$.

On the other hand, if $k = n$, using the logic from above, we can find that the probability of $k$ successes in $n$ trials is 1.

Therefore, the PMF is not $0$, as the entire mass of the function is contained at $k = n$.

However, as stated in Rohan's answer, having $p = 1$ doesn't really agree with the idea of a binomial distribution.

$\endgroup$
  • $\begingroup$ Hi. It's just confusing. Because on Wikipedia's page: en.wikipedia.org/wiki/Binomial_distribution, they say, $p \in [0,1]$. $\endgroup$ – Thomas Moore Sep 8 '17 at 22:01
  • 1
    $\begingroup$ Since p is a probability, it's restricted to be between 0 and 1 inclusive. Intuitively though, when you think about a binomial distribution, each trial is representing a success or a failure, with p representing the probability of success. When p is 1 (or 0), it's just a trivial case, because every single trial will be a success (or a failure). $\endgroup$ – Green Sep 8 '17 at 22:15
1
$\begingroup$

Note that a binomial distribution has two possible outcomes: success or failure. As a result, the random variable $X$ in a binomial distribution denotes the no. of successes.

If the probability of success, $p=1$, then it is no longer a binomial distribution!!

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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