I have a distribution of values for each of 12 months. When I draw the boxplots (one boxplot for each month) the median gradually increases, but the boxes overlap.

Which test do I have to use in order to tell whether this seeming increase is significant or not? Also, how would I quantify the increase? e.g. if between January and Feb the median increased by 1, but the box plots for these two months largely overlap, is that even meaningful?

  • $\begingroup$ You would likely get a better answer if you told us more about your data - for example, are these the same units observed each month, or a new set of units each month? Are you interested primarily in whether the median is different, or the mean, or something else? $\endgroup$ Feb 15 at 15:28

You could use time series regression, and test whether the slope/trend has a significant coefficient after accounting for months. This page might be a good place for more details.


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.