I am trying to conduct a statistical test of the difference between two shares expressed in percentage points. While I could use a regular t test if the difference was between two means, with shares I am not sure what to use.

The specific example refers to protest events. I can calculate what share of the events have been organized by parties and what share have been organized by other organizations. I would like to know if the difference between the two is statistically significant.

In this literature relative shares are most commonly used. Means of absolute numbers are not very useful. I am grateful for any hints!

  • $\begingroup$ You seem to be talking about tests of proportions. One common approach (as long as sample sizes are not too small) is via chi-squared tests in a 2x2 table of counts, though some people do a two-sample proportions test using a z-statistic (the two should be equivalent). If you want a one-tailed test the chi-squared won't work. However, I am not completely sure I have correctly understood your question. $\endgroup$ – Glen_b May 23 at 1:23

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.