Why is probability important to a scientist?

Answer 1

Some but not all scientific models are based on the ability to determine the likelihood that a given experimental outcome has happened by chance alone. If you have an accurate understanding of how the variables in the experiment change when nothing in particular is affecting them, then you have a way to establish some confidence that your outcome is the result of your experimental procedure and not the result of purely random events. The experimental 'lingo' is that the researcher has to determine if the 'Null Hypothesis' can be rejected. The Null Hypothesis is that the experimental outcome is not significantly different from what you would expect if the experiment had no effect at all.

As an example, if the probability in the natural world is that some event will happen by chance only one tenth of one percent of the time, then when I observe that event as my experimental outcome, I can be reasonably sure that my procedure has brought about the event; it is so unlikely that it happened by chance. It is not perfect, but few scientific procedures are. This also highlights the importance of replicating studies or of doing meta-analyses of experimental data gathered in many experiments to further reduce the likelihood that observed outcomes are nothing more than chance events.