We do a monthly fundraising letter at my nonprofit. I know how to set up tests in R to compare whether the proportion of response and amount received between two mailings is significantly different, but I would like to compare all twelve for the last year. I figure there is a test to do that, but I don't know what it would be.

Can someone suggest a way to look at 12 samples at once?

  • 1
    $\begingroup$ what tests would you do for the comparison of just two? $\endgroup$ – Glen_b Nov 26 '13 at 2:16
  • $\begingroup$ I would use a t test to check the donation amounts,and use R's prop.test function to check the response rate (how many responses to the total mailed out). Doing that for each combination would be very repetitive and have some additional error associated with it. $\endgroup$ – chrisfs Nov 26 '13 at 8:29
  • $\begingroup$ Yes, it would have the problems you suggest, but it gives me a better idea how to talk about how to do it. $\endgroup$ – Glen_b Nov 26 '13 at 8:59
  • $\begingroup$ I think Anova is what I am looking for in a case like this. en.wikipedia.org/wiki/Analysis_of_variance $\endgroup$ – chrisfs Nov 27 '13 at 20:01
  • $\begingroup$ Possibly you might use ANOVA for amount received, though even there it may not satisfy the assumptions; for example it may well be quite skew. ANOVA is unlikely to be suitable for proportions. There are a number of issues to cover. $\endgroup$ – Glen_b Nov 27 '13 at 22:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.