Yes, but I don’t think it is reading about Popper on Wikipedia that is the source of this stuff. Kids get told in the “introduction to scientific method” section of their first year university science courses, or maybe even when doing their science fair project at elementary school. The people who teach it probably know better, but they need to keep it simple. Popper probably isn’t even mentioned much of the time. It is not like he actually invented hypothetico-deductivism.
He’s pulling your leg; in my experience, GIGO reliably answers questions about whether a field involves falsifiable predictions by supplying a falsifiable prediction.
Nope, just like **Gorsnak **and **njtt **point out, appealing to Popper is not at the levels some propose, in fact I say that there is even a denial of sorts regarding the limits many scientists already know exist regarding falsification, and some indeed abuse the methods proposed by Popper.
Nah :), the reaction I saw here shows that I was correct before regarding the true perspective and how most scientists see falsification. I see it as touching my funny bone more like any nerves.
Well – in the spirit of that XKCD comic – how about, say, prescribing a particular antidepressant to treat folks with specified symptoms? Do you figure we could make predictions about how those chemicals would stack up against a placebo in a proper double-blind study? Could we hypothetically falsify those predictions?
Definitely. Just see Lakatos’ The Methodology of Scientific Research Programs. Here is a summary from the Stanford Encyclopedia of Philosophy. It is a wall of text, but one well worth reading.
My emphasis. This is exactly how the social sciences I am familiar with (political science and economics) work. For example, consider the Median Voter Theorem (Median voter theorem - Wikipedia). It is a core finding of rational choice political science, and empirically, it is completely wrong. But it’s not wrong because the theoretical groundwork that supports it is wrong; the MVT is just too frictionless to make meaningful empirical predictions. Political scientists did not abandon the entire framework when it became obvious that the conditions for the theorem didn’t hold. They gradually refined it and produced much more powerful results.
Actually that experiment has already been done and it turned out that antidepressants were no more effective than placebos. It doesn’t seem to keep doctors from prescribing them. Antidepressants were created by Biochemists, but psychologists have show no ability to actually figure out how they work or if they work. All they have are some rather subjective descriptions of symptoms that doesn’t seem to work much better than casting a horoscope.
Lakatos is off base. Science is always limited by the tools. He creates an imaginary example and thinks he is proving a point. There is always the possibly of experimental error, which is why scientific results have to be independently confirmed. That happened recently with the faster than light neutrinos.
String Theory actually is outside science right now, because we can’t built the tools to test it. It may become science some day when we build tools or it might be falsified. The limits of Science are defined by our tools, not by our theories.
I’m not following you. I sure wouldn’t mind a cite – but if, as you say, the antidepressants truly were no more effective than placebos, then why claim the predictions can’t be falsified? It sounds like something entirely relevant to figuring out the “if they work” point you mentioned at the end there.
So, by your somewhat idiosyncratic ideas about science, the observational sciences now get kicked out? Epidemiology is not a science now because we cannot construct experiments?
Actually, it was shown to be no better than placebo on patients who did not have moderate to severe depression, but shown to be effective on those who did.
Your point about doctors still giving them out for lesser illnesses is still on point, however.
That is correct. My main objection is that physicians prescribe antidepressants based on symptoms rather than actual observations of the brain chemistry. Physicians are not scientists yet, but there are some movement in that direction, like evidence based medicine. Of course, most physicians will remain technicians, but their actions will be based on actual science.
Your reading skills seem rather limited. The example we were discussing was from astronomy which an experimental science which generates falsifiable predictions.
I don’t know that much about Epidemiology, but don’t they do animal experiments?
Astronomy isn’t an experimental science, because we have no control over the things we’re studying. If you want to learn about the evolution of a star with a particular metallicity content, say, you can’t just make one in the lab; you have to hope to find one that’s relatively close, and which has minimal confounding variables, and extrapolate from there.
You’re right. My reading comprehension must be very limited indeed. After all, I am just a social scientist.
How is astronomy experimental? How was Galle’s finding of Neptune, the critical event that corroborated Newtonian mechanics, experimental and not observational?
