What is science relying on? What is its foundation and history? Why should we believe in science? Is it highly corrupted and based on who pays win? Let’s see ….
In the last year with the COVID situation, science has been put in the front line of the news and many scientists were put on the spot light to explain to the people what was going on. This led to the expression of the different opinions of diverse scientists showing that sometimes there is no consensus among them and also revealing conflict of interests for some of them. This was exemplified with the publication in the Lancet of an article based on wrongful data developed by a company making profit of such data production. Despite peer-reviewing, this article was published but rapidly removed thanks to the effort of many scientists to prove the incoherence of the underlying data.
Now the public is divided on who to believe and to some extent on how science is corrupted or completely wrong.
- Here, I will not discuss about this particular topic of the COVID-19 pandemic, linked with the different conflicts of interests that some scientists have, or the use of science for political reason.
- But I will try to give you a small overview of the scientific methodology how it has arisen, evolved and what are the actual standards.
Historically, science was a branch of philosophy known as natural philosophy or the philosophy of nature. For long, philosophers tried to come up with certain methods in order to prove or disprove theories or explain certain observed phenomena. Famous philosophers came up with methodologies such as Aristotle with his empiricism and syllogism, Roger Bacon, Francis Bacon introducing a new system of logic, René Descartes with his Discourse on the Method famous for ‘I think therefore I am’, and David Hume with his ‘skepticism’. In the 20th century, the scientific methodology was then further developed and consolidated notably with Karl Popper with his ‘critical rationalism’. Briefly, he states that we cannot prove a theory to be true but on the contrary we can prove it to be wrong.
From all this historical development of scientific methodologies, how can we describe the modern scientific methodology:
- 1st: scientific methodology should be based on empirical data (or observations). As long as we do not produce or gather data, our claims are just theories: no data = theory. Nevertheless, theories can be formulated based on data and still be considered as theories if they cannot be universally proven or other theories could also explain the data.
- 2nd: scientific hypotheses should be written in order to be disproved. The most famous example for that comes from Popper: we cannot prove that the sun will always rise, however we can disprove that every day the sun will rise, if one day the sun does not rise then our hypothesis will be disproved.
- 3rd: scientific methodology should rely on experimental design tending to disprove one’s own theory. This is linked with the above point and is based on the fact that if you want to prove something to be true then you can always find a way to do it by designing your experiment in such a way that your hypothesis will not be proven wrong.
- 4th: All scientific experiments must be reproducible. This is highly important, since it is a core of the modern scientific methodology because we cannot prove an hypothesis but just disprove it. The more an experiment is repeated giving the same result the stronger the hypothesis becomes.
- 5th: An experiment has to be designed to account for all the possible outcomes, to foresee all its limitations, and to give a clear answer to your original questions. This task is the most difficult in science, as many hidden factors may influence the outcome of the experiment or there are too many factors that were not accounted for (aka confounding factors) or the choice of variables was not good enough to provide you a clear answer.
- 6th: When several hypotheses are in conflict, the principle of parsimony is applied. This means that if there are several hypotheses the best one is the simplest one until gathering further evidence.
- 7th: Speculation should be kept to the bare minimum and the use of affirmative tense (stating facts) should tightly reflect your results and data.
So WHAT A SCIENTIST DO?!
More exactly here, I will describe how a scientific article is made. The first thing that scientists do is to READ A LOT of scientific articles and books to gather all the information on the topic he / she wants to investigate. From all this reading, they need to THINK ABOUT WHAT SUBJECT HAS NOT BEEN ADDRESSED or WHAT SUBJECT NEEDS FURTHER STUDIES. Then, they will design his question(s) and hypotheses following the 2nd point described above. When their question and hypotheses are well established, they have to design the experiment which will allow them to answer their question and disprove their hypotheses, following point 3 above. Often, the development of the question and hypotheses and the design of the experiment is not done alone but often together with colleagues and else, which will refine both question and experimental design to foresee possible shortcomings. Then they perform the experiment, gather data, and then analyze the data often using statistics. From those analyses, they will draw results and use them to answer their original question. Most of the time, the results obtained does not provide a definitive answer but on the contrary bring more questions. From their results together with their question, scientists will put back their experiment into a broader perspective and discuss what their results mean along former studies and theories. Finally, they need to draw all the limitations of their own experiment and results and provide ideas on what the next step should be. What I have just described is more or less the structure of any scientific articles:
- the introduction where the topic is introduced in its entirety gradually describing why you came with your question and hypotheses
- the material and methods where the experiment is described in sufficient details in order to be reproducible, how the data was gathered and analyzed
- the results where all the results are presented often within figures to ease the comprehension
- the discussion where the results are discussed within a broader context and along with previous studies, theories, and models, what those results mean and how they impact our current knowledge of the topic, what are the limitations of the current experiment and what can have been done better, and finally what directions the future research should follow.
This was my first post and hopefully many will follow!!! I hoped you enjoyed reading it and found it useful. Please do no hesitate to provide me feedback, if you think I missed something, something is wrong, you want more details on one particular point and so on …
PS: this article has been written by B. Fouks and should be solely considered as an opinion. The statements, views and opinions expressed in this column are solely those of the author and do not necessarily represent those of the funding body (EU MSCA-IF) and the University (WWU).