Comaring one year to previous years

#1
I'm trying to compare the number of events in this year to the previous three years. There is a decline that looks significant( previous three years ,16,11,14 and this year -2)-
What is the best and correct way to present it statistically? T-score?
 

Karabiner

TS Contributor
#2
You could first conduct a one-sample Chi² test with 4 categories.
The Null hypothesis would be that each category (year) has the same
0.25 share of the 43 events (in the population where the sampe is
drawn from).

Mind that the p values could be distorted, if you first noticed
some interesting looking pattern in the sample data and then decided
to perform a tailored statistical test to show that this is statistically
significant.

With kind regards

Karabiner