1

I am trying to compare the goodness of fits of an exponential and a log-normal distributions to some empirical data. Is it possible to use a chi-square test to do so, even though the data is quantitative?

In the section 2 of this paper (http://physics.umd.edu/~yakovenk/papers/PhysicaA-370-54-2006.pdf) the author use the chi-square test for that. But no matter where I look, the chi-square is always used for categorical data.

Thank you very much for your answers.

Rodrigo Guinea
  • 123
  • 1
  • 4
  • 1
    You bin the data. The details matter, though: if you don't carry this out exactly the right way, you will get the wrong results. In an answer to a different question at http://stats.stackexchange.com/a/17148/919, I explain the details and provide worked examples. – whuber Mar 12 '17 at 18:34
  • Thank you @whuber, I understand. It's a shame I can't rank this. – Rodrigo Guinea Mar 12 '17 at 19:53

0 Answers0