I’m studying the probability of committing a crime after several specific events (A, B, C). I have 50 subjects, each of them has a probability to commit a crime after events A,B,C.
I want to show this in a 3 bar plot, a bar per event. The bar amplitude corresponds to the mean of the probabilities of committing a crime after each event (taking into account all subjects).
Im representing the standard deviation (SD) in little bars, as error bars. Some bars, cross the value 1, because mean+SD exceeds one. What might be happening? Is this possible and reasonable?
