Science is defined by the methods used, not by what is being studied. It is the method that differentiates science from other human endeavours and gives it authority. This method involves making observations and asking scientific questions. A scientific question is one that can be answered by experiment. The 'scientist' comes up with an answer, called a hypothesis. This hypothesis must be falsifiable for it to be considered scientific. This means that there has to be a way to disprove it via repeatable experiment, if it isn't true. If such experiments continually fail to disprove the hypothesis, it will eventually become accepted as a scientific theory. When people stop arguing about it, it becomes a scientific law. All hypotheses, theories and laws could still be disproved and most probably will be, especially in physics.
Some fields of study lend themselves to the scientific method more than others. I can think of two reasons for this. One is that the scientific method applies very high standards which some fields of study could not meet. The other is that by it's nature, science only deals with the nature of things as they are now. It does not deal with questions of history because questions of history do not lend themselves to repeatable experiment. If an experiment can only be carried out once, or a thousand times, or a million times, then it is not scientific. For it to be scientific it must be possible for any future researcher to repeat the experiment and obtain the same result (provided the nature of the universe hasn't changed of course). This does not diminish the value of history (or art, or law) as a field of study, as they each have and continue to yield valuable insights into us and our environment.
If someone makes an observation, comes up with an hypothesis, then makes more observations to check whether it agrees with the hypothesis, and continues making such observations indefinitely, that is not science. Historians can get away with that method. That is how most knowledge and understanding of the world arose prior to the emergence and institutionalisation of the modern scientific method. But it isn't scientific. The power of science, it's authority, it's contribution to and influence over modern society is derived from the power of the scientific method. The scientific method is actually the slowest in aquiring new information or understanding, at least in the short term. That is why science is boring in practice. But it is also the surest method, which is why the combined efforts of so many scientists can combine into so much academic progress.
Nearly all fields of study employ the scientific method at least some of the time. Auto mechanics use it to solve difficult problems with multiple potential explanations. No fields of study employ the scientific method exclusively. For example, there was a lot of progress made in chemistry before the modern scientific method arose. People figured out how to make explosives by making observations, stumbling across discoveries and tinkering – this was an evolutionary process, not a scientific one.
An experiment must have a single controlled variable – the one changed by the experimenter, and a measured variable. This is why so much biology has difficulty with science – because of the ethical and logistical difficulties in carrying out such an experiment. Much of biology (especially the parts relevant to natural resource management) relies on 'natural experiments,' however their scientific value is diminished because of all the uncontrolled confounding variables. Psychologists and doctors also have to be careful with controlling all variables, as the observer can influence the outcome. That is why they use double blind experiments.