Question

Contrary to popular media reports, scientists rarely prove a theory. Instead, the language of science requires a person to report that a theory is supported, partially supported, or not supported. Based on your knowledge of statistics, explain why the scientific community uses this language.

Answer 1

What scientists really prove or rather demonstrate through their research, is basically the result of a hypothesis test. Given the evidence at hand, they work out the probability of a certain explanation for a phenomenon/observation being more or less probable. Hence, with the sample collected/available, statistics allows scientists to explain the theory using the principles of inference, like hypothesis testing.

This is reasonable, as given one rarely has the entire population at disposal (say, all the possible cases of cancer ever spotted), the best one can do is to carefully analyze the sample at hand, with appropriate randomization, and come up with inferences which help solve the problem at hand (say, devise the most promising diagnosis for a certain type of cancer). Given the tools of hypothesis testing, one can establish procedures which are expected to give favorable results with a high probability or low failure ratio. For example, the most accurate course of treatment for a disease as per cases encountered/solved historically. Hence, statistics plays a vital role in applying the findings from a smaller sample to the wider population.