The Scientific Method:

• Generate a hypothesis: This may be by any method, from deductive reasoning to intuitive, inductive leaps of logic. By whatever method, a testable hypothesis is formed, which clearly and objectively postulates a connection between observations. If a hypothesis is put forward which is not testable by objective means, it is not science, but philosophy.
• Perform experiment: This may be a traditional experiment, where objects, plants, or animals are subjected to changing conditions -- or it may consist of simple, methodical observations. The purpose of this step is to collect data which will hopefully either confirm or contradict the hypothesis.
• Analyze data: The data gathered in the experimental/observational phase is compiled and analyzed. Generally, statistical tests are run on the data in order to determine aggregate properties of the data such as mean and standard deviation.
• Interpret data: The compiled and analyzed data is interpreted, and perhaps conclusions can be drawn. If sufficient data has been collected that would tend to either confirm or contradict the hypothesis, a definite statement about the likelihood of the hypothesis' being correct can be made.
• Publish results: In order to be of the most use, results should be shared with others whenever possible. There are so many diverse fields of scientfic study in the modern age that no single scientist, no matter how brilliant or dedicated, can possibly claim expertise in every scientific discipline. Science thrives on honesty and openness; even experiments with inconclusive results can be useful in the long run, since over time statistical analyses can be performed which may yet be able to draw conclusions.