0
$\begingroup$

I have the interarrival time $t$ follows a normal distribution of $N(8,4)$. I'm trying to find

  1. $P(t < 0)$
  2. The probability that the 16th and the 9th customer arrive within 55 minutes of each other

For 1, it would be this right?

$P(t < 0) = \int_{-\infty}^{0} \frac{1}{2\sqrt{2\pi}} e^{-(t-8)^2/(2 * 2^2)}$

For 2, I'm a bit confused. Is this a Poisson distribution then? Because t already follows a Normal distribution.

Can I have some helps with 1 and 2 please?

$\endgroup$
7
  • 1
    $\begingroup$ In general, if $t\sim\mathcal{N}\left(\mu, \sigma^2\right)$ where $\sigma > 0$, then by standardising, we have $P(t < a) = P\left(Z < \frac{a-\mu}{\sigma}\right) =\Phi\left(\frac{a-\mu}{\sigma}\right)$, where $Z$ is a standard normal random variable and $\Phi$ is the standard normal CDF. If you have a table of standard normal CDF values, you can use this to calculate this expression. Your integral expression is also correct, but how are you expected to calculate it? $\endgroup$ Mar 7 '19 at 22:22
  • $\begingroup$ Isn't an interarrival time a non -negative random variable? How can it be normally distributed?. $\endgroup$ Mar 7 '19 at 23:30
  • $\begingroup$ @MinusOne-Twelfth Oh I was going to use the CDF for that but I thought I can only use it if I were to calculate $P(t \le 0)$. This is me not really knowing CDF and PDF as well. $\endgroup$
    – PTN
    Mar 8 '19 at 0:17
  • 1
    $\begingroup$ Possibly "for simplicity", we are modelling the interarrival time as normally distributed (maybe we want to model it as peaked around $8$ minutes for example), and maybe the question asking for $P(t < 0)$ at the start is asking us how likely it is that under this model, we get an interarrival time being "absurd" (negative); if this probability turns out to be very small, then it is perhaps not too much of a concern that we can get negative times from this model. That's my best guess, anyway. $\endgroup$ Mar 8 '19 at 0:51
  • 1
    $\begingroup$ @PTN We can use the CDF for $\le$ and also for $<$ (since we are dealing with a continuous random variable here, the answer is the same either way). $\endgroup$ Mar 8 '19 at 0:52
1
$\begingroup$

Part 1 looks right to me. I would guess that Part 2 is asking for the probability that the wait times $t_{10:16} = \sum_{i=10}^{16} t_i$ are less than 55 minutes, that is $P(t_{10:16} < 55)$. Assuming that the arrival times are independent, then $t_{10:16}$ is also a Gaussian $N(7 \cdot 8, 7 \cdot 4)$ and hence we can find $P(t_{10:16} < 55)$ just as in part 1.

$\endgroup$
2
  • $\begingroup$ That makes sense. I just have questions at how did you get the $7$, and why is it $t_{10:16}$ and not $t_{9:16}$ $\endgroup$
    – PTN
    Mar 8 '19 at 0:19
  • 1
    $\begingroup$ The 7 is because there are 7 terms in the sum (we count starting at 10). I used $t_{10:16}$ because $t_9$ is irrelevant -- it is the waiting time for the 9th person to arrive, but the second problems asks us to start counting when the 9th person arrives. $\endgroup$
    – dskeletov
    Mar 8 '19 at 23:02

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.