James B. Conan remarks in his book Science and Common Sense, “[t]here is no such thing as the scientific method. If there were, surely an examination of the history of physics, chemistry and biology would reveal it.”
In addition to being a scientist and noteworthy scholar (and, incidentally, one of the main scientists who developed napalm), Conant was very concerned with the popular misconceptions of scientific inquiry by non-specialists and a “popular science” author. Unlike most other such authors, though, Conant didn’t seek to popularize some field of science. In the book referred to above an in others (Modern Science & Modern Man, On Understanding Science: A famous scientist explains science and its role today, etc.) Conant wrote about what “science” actually involves compared to what most have thought it to who are not practicing scientists.
For those who know something of the history & philosophy of science literature of the 20th century (e.g., Kuhn’s paradigm shift idea, Popper’s falsification approach, the feminist and other radical challenges to scientific epistemology from the 60s onwards, etc.) it is worth noting that before Popper had published his second academic work on the nature of science Conant had already published at least 5 books intended to explain science to non-scientists or addressing science education.
I have a short personal story that cuts straight to the heart of the issue. Very early in my graduate career the head of my lab ran a weekly graduate seminar (some post-docs were there as well) largely devoted to criticizing a wide-spread theory that he disagreed with. At one point during the seminar, a graduate student scathingly criticized a peer-reviewed study supporting the theory that the head of the lab regarded as so utterly flawed, but surprisingly he actually reined her in. He said that while the study was flawed, it was an improvement on previous studies by the same author and similar researchers and that it addressed many of the “flaws” we had covered in studies in previous weeks. Here comes the interesting/important part. He went on to say that it could be future studies continued to find the evidence for embodied cognition and do so without any flaws. In that case, he stated, we would have to realize that the methods used by neuroscientists were inadequate.
Now, as anybody who has taken even high school science classes can tell you, The Scientific Method (TSM) says that if you continually confirm some hypothesis, then you accept it as theory (at least until it is falsified). So why was this distinguished professor, with an academic pedigree few could match, so blatantly rejecting the basis for TSM? I asked him how we could determine when the methods were the problem vs. the theory (that is, given many experiments in some field of science that all support the same theory, how can we determine whether findings reflect reality or poor methods)? He replied with something to the effect of “when you have a really good reason for thinking that certain evidence should exist but you don’t find it, it’s because your methods are wrong.”
Apart from introducing a component that isn’t part of TSM (i.e., after confirming a hypothesis, determining whether the means of confirmation are sound), this part of real scientific research has a much more pernicious, nuanced, and devious side. Scientists who have a “really good reason” to suspect to find X result do so because they aren’t really testing hypotheses. They use their theories to develop hypotheses, design experiments, and interpret the results in order to contribute to these same theories.
The most dramatic example of this was the fall of classical physics. Simply put, before Einstein’s work in 1905, scientists beginning especially with Young had confirmed that light was a wave for over a century. Not only that, it was an electromagnetic wave and fit nicely into classical electrodynamics as a wave. Then Einstein showed empirically that light isn’t a wave but a “particle” (made of discrete units we now call photons). The problem was that physics of that time (what we now call classical physics) held that something could be a wave or a particle but not both. Now, if it weren’t for the simplicity of the experiments, physicists would still be arguing over whether light was a wave or not just the way scientists in various fields have argued for decades over which of at least two mutually exclusive theories is (more) accurate.
Luckily, the experiments were simple and the only possible answer was the realization that the experiments were testing to see if light was a wave or a particle because theory said it must be one or the other. The theory, though, was wrong. Any method used to confirm that light was a wave or a particle was doomed from the start because it would involve trying to confirm what didn’t (and doesn’t) exist.
To varying degrees, this is an issue with all scientific research: scientists “confirm” theories using those theories not just to develop how they will confirm them but even how they interpret their results. This notion that scientists develop a hypothesis, try to disprove it, and if they can’t it becomes theory is thoroughly mistaken. TSM, as presented even in college textbooks, is at best a 19th century relic that lead to an epistemological crisis when the oldest, most venerable field of the sciences (physics) went from being “nearly complete” to “nearly completely in ruins.”
There is no Scientific Method. Scientists conduct research (ideally, anyway) using logic, interpretation of probability, and in general that unintuitive kind of analytic thinking that typifies formal argumentation and mathematical reasoning. They also, though, rely on frameworks (theories) within which all their work is conducted, and nothing in The Scientific Method hints at this, let alone attempts to explain it.