The larger the experiment example the more reliable the results would be. Take for example if you analyze something using only 2 examples and one confirms your theory and the other doesn't. That would have you believe that it occurs 50% of the time. If you analyze a hundred of them you can see more correct results.