Comparing odds (random)

Jara

New Member
#1
Hello to everybody,
I would like to ask how to compare the odds (the random responding).
Let's image the situation of the test where are two types of events - target and non-target.
Person need to respond to target event and must not respond to non-target events.
Target events are occurring in 20 % of cases (i.e. 20% of events are targets).
The person responded in 120 times. That means that if he/she would be answering randomly, responses to target events should be present in (around) 24 times.
The results showed that person responded to target event 29 times (rest to non-target event).
How I can test if the person still responded randomly (29 times to target 91 times to non-target (complete random would be 24 target 96 non-target)).
By other words random chances for target answerings are 20% and the person answers to target in 24% cases (how to statistically test if it is significantly different).
Thank you very much for the hints :)
Jara
 

hlsmith

Less is more. Stay pure. Stay poor.
#2
Would it be appropriate to compare this to predicting coin tosses. So they should get 50% right given the number of flips approaches infinity and you want to be able to say the person correctly predicted the number of flips better than chance? If so, look at binomial tests.