In our search for information we almost always run into studies that publish their results using statistics. It is important to understand what these numbers mean. These results apply to the population as a whole and mean nothing to an individual case. They can be validly applied as guidelines for treatment but contain no guarantees. Also, they are not probablilites because they are calculated from experimental data. (To calculate the probablility that a certain treatment is, say 90% effective for a cure, we would have to understand everything about
the physical, medical and chemical processes involved.)
Example: The probability that a coin flip comes up heads is 1/2. We know this because we understand the simple physical law that governs this action. Yet, if we flip a coing 10 times, it could come up heads 2 or 7 or 9 times. If we used these experiments and called the results probabilities, the probability of a head would be 2/10 or 7/10 or 9/10.
The principle that validates the use of experimental data in studies is that if we do the experiment a very large number of times, the result will approximate the probability. Yet any single result of the experiment would probably not match the overall result.
The media thrives on eye grabbing sensationalism. The following link demonstrates how statistics are used to alarm people and sell ad space.
http://www.latimes.com/news/opinion/la-oe-tavris17apr17,0,3924193.story