10

I have a function that returns an average wait time for a web user. That is, it gives an average time that an average user may stay on a web page, given the web resource length in words. I want to use this function (and the resulting average) in conjunction with a distribution to model an 'average web user' browsing the web. Which distribution(s) might be appropriate for this, and why?

Edit: I would also like to particularly know the viability of using an exponential distribution for this purpose.

Thank you

Ken - Enough about Monica
  • 543
  • 6
  • 18
  • Update: I am still looking for information on this. So far my (very amateur) opinion is leaning toward an exponential or weibull distribution, though I am not sure how right these guesses are, or what parameters to use with the distros. Information on that would be appreciated as well. Thank you. – Ken - Enough about Monica May 26 '14 at 09:12

1 Answers1

3

It looks like a Weibull distribution is appropriate, as seen here

Ken - Enough about Monica
  • 543
  • 6
  • 18