The scientific method
So this blog is about science. But what is science and how does it work?
Science is the process in which scientist acquire knowledge based on empirical and measurable evidence. But what does this really mean? What is the scientific method and how do we define the scientific method? Here I discuss the steps of the scientific method. The scientific process starts with asking questions.
Scientist are much like young children, they see things around them (observation) and question how it works or why it happens. Just as parents do not always have answers to the questions of young children, scientist cannot always find the answer in the existing literature. For example, I work on extraembryonic membranes (membranes around the insect egg). I already know that almost every insect egg produces these membranes and I also know that in one specific beetle, the eggs can also survive without this membrane. But if these eggs can survive without this membrane, why is it there?
After a question has been formulated, the scientist will try to form a hypothesis. A hypothesis is a statement which, if true, leads to certain predictions. A very important aspect of this hypothesis is that it should be possible to test it. This means that a hypothesis should be falsifiable (can be disproven). It is important to understand that in most sciences (math is an exception) it is impossible to prove without a doubt what is true, normally scientist will gather evidence until their hypothesis is very likely to be true. I will explain this in an example. Take the next “wrong” hypothesis:
- This membrane is around the egg because it’s pretty.
There are several things wrong with this hypothesis. It is not particularly precise, who thinks it is pretty? Secondly, it is not falsifiable. I can ask the beetles as many times as I want whether they think eggs with membrane are prettier than eggs without a membrane but as soon as one answers I will have a heart attack and won’t survive to tell anybody about it. Furthermore, it does not lead to any predictions. Why would an insect egg want to be pretty? A better hypothesis would be:
- This membrane protects the egg from drying out.
Why is this hypothesis better? Well, it leads to testable predictions. If it does protect the egg from drying out, then eggs without this membrane should perform worse in dry conditions than eggs with this membrane. So it is possible to disprove (falsify) the hypothesis, if eggs without this membrane do not perform worse in dry conditions this hypothesis cannot be true. So why can I not prove that this hypothesis is true? Even if eggs without this membrane perform worse at dry conditions, there are always possible other factors that I did not take into account, these are called confounding factors. A good scientist will take as many as he or she can think of into account, but you cannot take the factors into account which you do not think of. If I for example only looked at dry conditions and found that eggs without this membrane performed worse than eggs with this membrane I might conclude that this membrane protects the egg from drying out. But maybe the lack of this membrane also leads to a worse performance at high humidity, which would show that the lack of the membrane in general reduces the success of the eggs and it has nothing to do with humidity. That is why I would test the success of eggs at different humidities. If I find that eggs without a membrane perform worse only in dry conditions and eggs with this membrane do not have this problem, than I provided evidence that my hypothesis is likely to be true.
Science in real life
In reality, scientists start with a question and a hypothesis. Then they design experiments that are suitable to test their hypothesis. However, in real life, experiments often go wrong. Or they clearly disprove the hypothesis. Furthermore, for each question answered, 10 new questions arise. So a scientist is constantly updating his or her ideas. If the first hypothesis is disproven, what would be another possibility? And how can he or she test that hypothesis? If the scientist is content with the data and thinks it shows that the hypothesis is very likely to be true then it needs to be published. But I will get back to peer-reviewed publishing later. Below the scientific progress is outlined in a figure.