The term science means at least two things: (1) the application of the scientific method, and (2) the human endeavor to understand our physical world by scientific investigation.
The human scientific endeavor is an ongoing accumulation of evidence-based knowledge about the natural world, continually building upon the experiments and findings of many generations of scientists. It relies upon the proper conduct of scientists, the proper design and interpretation of experiments, and the proper consensus by scientists about the meaning of those experiments.
All scientific understanding is evidence-based, but interpretation can vary about the meaning of any piece of evidence. All scientific views are open to revision if convincing evidence is found that counters our current understanding. Invalid conclusions about the nature of the world are eventually overturned.
Science, as a human endeavor, is cultural. It is a way that humans are trained to think about the world and how to ask testable questions. A scientist "belongs" to a worldwide scientific community that shares its standards and methods in order to best assess the findings of its members.
Scientific evidence is expected to be testable, repeatable, and falsifiable. Results of an experiment are considered relevant only if the experiment can be repeated and if there was a fair chance that the expected outcome would not occur. New ideas tend to be scrutinized more heavily than familiar or accepted ones.
Odd results from an experiment can often prompt a scientist to reject that experiment completely, on the assumption that something went wrong. The experiment will usually be repeated, but if the odd result is again encountered, one of two things will happen depending on the personality of the scientist:
One type of scientist will be convinced they should have seen the expected result. They will not question their original hypothesis but instead will redesign their experiment, or abandon it entirely. Such a scientist usually reports only those experiments that "work" (i.e., that end with an expected result), which tends to confirm the scientist's understanding of the system being studied.
Another type of scientist, however, might question whether their original hypothesis was valid in the first place. This is a far more exciting line of questioning because the hypothesis itself was a logical application of current knowledge. If the experiment was well designed, the strange result could mean that current knowledge of the system might be incorrect, or at least insufficient.
This later kind of scientist tends to be more rare, for he/she will report a negative result but will take the time to explore what it means in the context of a revised understanding of the subject knowledge. Important shifts in human understanding of natural systems often derive from "failed" experiments, when a good scientist questions our basic knowledge and finds it wanting.
