I have a data set (2003 rows and 2 columns) containing: years (2011 and 2013) and contact with hospital (yes/no). The persons answering the question are independent between 2011 and 2013.

What test should I use to check if there is a statistical difference between the years?

What test should be used if contact with hospital had three or more alternatives?


1 Answer 1


If you don’t care about the ordered relationship of years, use a chi-square test. If further types of contact with the hospital have no order (ie no contact, had contact or uninsured) chi-square is still applicable.

If you need to know the effect of advancing years on hospital contact, use logistic regression. If you have a collection of counts of contacts per year (ie 0,1,2,...), consider a count model, such as Poisson regression.

  • $\begingroup$ Thank you for your response! What if I was to add a change that happened in 2012, so 2011 the change had not yet happened and 2013 the change had happened. And I would still have the question of difference between 2011 and 2013, that is to say is there a statistical difference before the change and after the change between 2011 and 2013 for persons having contact with the hospital or not? Would it still be the same sort of suggested tests? $\endgroup$
    – Anno
    Dec 3, 2017 at 16:50
  • $\begingroup$ Yes, you create a variable representing the era before and after the change and then proceed with chi-square or logistic regression. $\endgroup$
    – Todd D
    Dec 3, 2017 at 17:25
  • $\begingroup$ Please accept an answer if it answers your question. $\endgroup$
    – Todd D
    Dec 3, 2017 at 20:49

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.