0
$\begingroup$

Let's say a bus company has 500 markets with an average of 10000 passengers each day but some markets as large as 100000 and some as small as 300. How would I determine the simple random sample size of the entire population so that with 95% confidence I could determine if demand in a given market increased by 10%?

$\endgroup$
  • $\begingroup$ I think you're treating all markets as the same and you shouldn't. For example, if you sample 100,000 people from the population, you may get 0 from a certain market. With that, you cannot say anything with certainty about that market where you sample nobody. You can either treat each market as different units or each market as essentially the same thing but your final results will either be in reference to the population or each specific market. Which are you attempting to answer. $\endgroup$ – Matt Brems Nov 21 '15 at 20:24
0
$\begingroup$

Both... It looks like the government ran into a similar issue. See page 30 in the link below. http://www.regulations.gov/#!documentDetail;D=DOT-OST-1998-4043-0070

Another secondary issue is that most of the market sizes are fairly similar in size but have very different price points (distribution is not normal) so even a large sample from a MOE standpoint doesn't result in the sample matching the population...

$\endgroup$

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.