Science is a collective, multi-layered process consisting of three steps. First is the individual scientist testing hypotheses according to the norms of their field. Second, the results of the individual scientist undergo peer-review and are published for the community to evaluate. At this point a result may be considered preliminary, but not proven.
Third, important claims are then re-tested in the "crucible of science" -- they are either reproduced by independent scientific groups or they have their implications tested to insure consistency with the existing body of scientific knowledge. After enough tests/reproductions, a consensus emerges that the idea is correct.
