3
$\begingroup$

I have the number of dispensed items per year and the cost per item - the graphs look as if when the number of dispensed items increase the cost comes down per item (red = cost, blue = number dispensed). For example,

enter image description here

enter image description here

I want to do some a statistical test to see whether or not the number of items dispensed in a year increasing forces the cost down / if they are correlated, but I'm not sure which test to use.

$\endgroup$
1
  • $\begingroup$ You may do a trend analysis or a correlation analysis. $\endgroup$ – Ayalew A. Apr 14 '15 at 12:40
3
$\begingroup$

If the assumptions of randomness, fixed distribution, fixed location, and fixed variation hold, you can perform an independent samples t-test. If the assumptions are violated, try a Mann-Whitney test.

If you are just interested in correlation, try a Pearson test. If the assumptions are violated, try a Spearman test.

If you would like to perform a statistical test with time series in mind, consider a Granger causality test.

$\endgroup$

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.