A single measurement result is not a proof--this has been shown again and again in science. We can only really rely on a research result when it has been measured several times, preferably by different research teams, in slightly different ways. In this way, errors can usually be detected sooner or later.