What exactly is science? Thinkers from Aristotle to Larry Laudan have pondered some version of this question and come up with different explanations or criteria. Most of us have seen the linear method containing steps like coming up with an idea, testing it and evaluating the results. This, however, is an oversimplification and the real process of science. This does not mean that rules of thumb like falsifiability is wrong or unhelpful, however, science is much more dynamic, combining interacting features such as exploration, testing, interaction with society and communicating results in the peer-review literature.
There are many factors influencing the start of scientific exploration. It could be a practical problem, like how to reduce smog from a major city or treat malaria, it could be due to the development of new technology, such as biotelemetry equipment that reduces effects of measurements in animal research, or simply unintentional accidents, like Alexander Flemming’s discovery of penicillin. It can also be due to being put into a research group, talking to academic advisers, reading an exciting new article in the peer-review literature and so on.
When designing an experiment in many areas of science, a null hypothesis is often tested. This is usually stated as no statistically significant different between two groups, two treatments or two physical variables. Results can often be confusion, hard to interpret or inconclusive. This spurs further research, trying to fix the problems or approach the issue from a different angle. Often an experiment needs to be useful, coherent and cost-effective.
Research is often published in the peer-review literature, presented at scientific conferences, discussed with colleagues and used to build upon the state of knowledge by improving explanations and models or replicating already existing research in order to critically evaluate it.
Some of this knowledge can be used to improve our life and society, from finding the link between smoking and lung cancer and treating infections to renewable sources of energy and designing smarter computers.