User blog comment:Homuhomu123/Estimation of Radar Modifiers on Artillery Spotting Chance/@comment-25534439-20141213144852

Your sample size is rather small, but large enough to be interesting. It's not the most statistically appropriate method, but a standard error calculation puts your DA/CI rate's +/- intervals at 4.8%, 5.8%, 4.1%, and 4.7%, respective to the ships (90% confidence; a better method takes intergrals and it's too early at the moment). You have a fairly solid basis to differentiate the Air and Surface radar CI chance of 24% vs 30%, but your other conclusions aren't so solid, statistically.

On the one hand you need more tests, on the other I note you've already poured a lot into this.