1
$\begingroup$

In a homework problem, I am given a Poisson distribution with lambda = 1 as null hypothesis, lambda greater than or equal to 2 as an alternate hypothesis, and 3 as a test statistic. I am instructed to reject the null hypothesis if I observe that test statistic. We are describing traffic accidents in the problem, so the question is "What is the probability that you are committing a Type I Error if you reject the null hypothesis (lambda=1) when you see 3 accidents?" I'm then asked for the power of the test.

My first thought was to use (e^-lambda)*((lamda^3)/(3!)) to get the probability of 3 given a lambda of 1. But it seems like it has to be more complicated than that. Would I need to get the sum of the probabilities of 1 and 2, with lambda = 1, and subtract that from one?

Would someone mind putting me on the right track? (I hope this question was intelligible).

(By the way, I know it's extremely annoying to have the formula written in words rather than symbols--I'm brand new to the site and under sort of a time crunch, but I'll get it right in future posts. I apologize for the irritation!).

Thanks very much!

$\endgroup$
1
$\begingroup$

Your second suggestion is right, with a little detail incorrect. A reasonable test is to reject the null hypothesis if the observation is $\ge 3$. So the probability of Type I error is $1$ minus the probability that, if $\lambda=1$, our observation is $0$ or $1$ or $2$.

Thus the probability of Type I error is $1-e^{-1}\left(1+1+\frac{1}{2}\right)$.

$\endgroup$
  • $\begingroup$ Thanks André! I guess I forgot that we might have NO accidents as well. $\endgroup$ – jonb Oct 19 '14 at 17:22
  • $\begingroup$ You are welcome. It is all too easy to forget about $0$, it is so tiny. I have seen this little slip a few hundred times. $\endgroup$ – André Nicolas Oct 19 '14 at 17:26

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.