Science students and teachers are often asked to define science. The traditional natural sciences include subjects such as astronomy, biology, chemistry, geology (often called Earth science in schools), and physics. Then there are the social sciences such as anthropology, psychology, sociology, and others. What is the common thread that makes these areas of study science?
Scientific methodology, or the scientific method, is a process for learning about the world or universe around us. One good definition of science is anything that follows scientific methodology. Both the natural and social sciences mentioned above use scientific methodology to help advance knowledge.
The scientific method as it is typically taught in schools is an idealized approximation of how scientists really work. Students may memorize, often with no real understanding, a series of steps that constitute the scientific method. Scientists however apply these steps differently in different situations and for different areas of science.
For example, chemists and physicists can do properly controlled experiments that change only one variable at a time. Astronomers can not do experiments; they must simply watch as the universe unfolds before their telescopes. Biomedical researchers use controlled double blind studies to reduce possible bias in clinical observations. Social scientists study such complex human interactions that they must resort to statistical studies of large sample sizes. These scientists are all, however, following scientific methodology appropriate to their areas of study.
Falsifiability and Scientific Methods
Given these variations among areas of science, the first step in scientific methodology is to observe nature in some way via observations or simple experiments. Scientists then try to explain the observed natural phenomena via a scientific model, hypothesis, or theory. Theories, hypotheses, and scientific models are approximately the same thing with differences beyond the scope of this article. Theories tend to be more encompassing than scientific models or hypotheses. The rest of this article will use the term model to cover all three.
All scientific models must be testable and falsifiable. In other words even the best and most elegant scientific models must be subject to being proven wrong by experimental or observational data.
Scientists use the model to make very specific predictions about what should happen in a certain situation and perform experiments or observations to test the prediction. If the experiment or observation agrees with the model’s prediction, then the model passed the test. Scientists will tentatively accept the model as being correct but continue to test the model.
The most important identifying feature of real science however is what happens if the model fails the test. If the data disagree with whatever the model predicted, then the data proved the model wrong. Data falsified the model. Scientists must then change the model. Changes to the model might range from minor tweaks to a completely new model that incorporates and agrees with the new data.
This new model is however not the final answer. Scientists continually make new predictions based on their models and continue to test their models with increasingly more stringent tests. Scientists only accept a model or theory as being correct if they continually fail to prove the model or theory wrong.
Science does not however prove a model or theory correct. The best scientists can do is to fail to prove that a model or theory is wrong. There is always the possibility that a new more stringent test will prove that a scientific model or theory is wrong. Good scientists will abandon even the most elegant theoretical models, when compelled by new data.
What is Science?
The key distinguishing feature of science is falsifiability. All good science is subject to being falsified, or proven to be false, by experimental data or observations. Anything that can not be tested and subjected to the possibility of being proven wrong is not science.