I would like to compare the total number of public M&A transactions observed in two separate years (e.g. 2021 vs 2022). Given the data set for each year is complete (i.e. a population), can I simply compare the counts and determine an increase or decrease in activity? Or should I view each count merely as samples of a population of all M&A activity (all-time) and use hypothesis testing to demonstrate the two samples are likely to come from different populations before concluding a significant difference?



1 Answer 1


There are two issues here:

One is what you are interested in. If you are only interested in those two years, then you don't need a hypothesis test. Hypothesis tests are inferential statistics that go from a sample to a population. You have the population.

OTOH, if you are using those two years as a sample, then you might want a hypothesis test, but, depending on exactly what it is and what data you have, you probably won't be able to do one, because your N = 2. (Again, that would depend on what you are testing).

  • $\begingroup$ Thanks for your thoughtful response, Peter, it’s much appreciated. If I could break down the annual count in each year to weekly or monthly totals within each year (to boost N), do you think I could use a Poisson rate test to demonstrate the rate in one of the years was significantly different from the other? Thanks again, Kerry. $\endgroup$
    – Kerry
    Aug 6, 2023 at 19:09
  • $\begingroup$ Why would you do that? People get hung up on "significance" but, if you have the population then, in effect, p = 0.000 $\endgroup$
    – Peter Flom
    Aug 6, 2023 at 23:32
  • 1
    $\begingroup$ Ah - I see. That clears everything up for me and answers my original question. Thanks, Peter. $\endgroup$
    – Kerry
    Aug 7, 2023 at 9:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.