Scientific Misconceptions
There are a few subtle but important common misconceptions about the scientific method which come up often enough that I'd like to address them here so that I can point to it later. In short: When someone says "Scientists perform experiments to prove that a theory is true," they are actually saying three wrong things.
The first is the misconception that all scientists perform experiments in laboratories. This is sometimes true, but not always. All scientists perform tests, not experiments. Some of those tests are controlled laboratory experiments, some are field experiments, some are natural observations, and some are observational studies. The goal in all cases is to make observations. This allows scientists to study things that don't fit into a laboratory, like supernovas or tribes of pygmies.
For example evolutionary biologists are scientists. A few evolve creatures in a laboratory, but most attempt to observe evidence of evolution in the natural world. A group of scientists studying devonian speciation realized that they ought to find an intermediate "fishapod" species spanning the gap between fish and tetrapods on Ellesmere Island in Canada. They looked, and found tiktaalik. Jennifer Clack said "It's one of those things you can point to and say, 'I told you this would exist,' and there it is." Validating a theory's extremely specific predictions with observations of physical evidence is as good as science gets, even though it happened thousands of miles from the nearest laboratory.
The second misconception is that tests prove that a theory is true. This is inaccurate. Tests are not just about proving whether a theory is true, because many theories are true to varying degrees. Good tests determine which hypothesis to favor: the null hypothesis or the alternative hypothesis. A test is not useful if its outcome is ambiguous; if its result doesn't indicate which explanation is more likely to be true.
For example the decomposition of mercuric oxide was evidence of Lavoisier's theory of oxygen, but it was also claimed by Priestly as evidence of phlogiston theory. Lavoisier needed to show that the gas produced in the decomposition was oxygen in order to disprove phlogiston theory.
The third misconception is that tests prove which theory is true. This is inaccurate. Theories predict observations, and tests reveal whether those predictions can be observed. Tests do not prove which theories are true, they allow theories to resist being proven false. A theory that passes a million tests is not necessarily "true" while a theory that fails even one test can be suspected of being false ( to some extent). This is why theories which are not falsifiable are not considered science - because unfalsifiable theories are not even wrong. If you can't be wrong you can't be right either.
For example Newton's law of universal gravitation consistently resisted falsification for almost 300 years, until observations like the Pioneer Anomaly indicated that a new theory was needed to correct it. We don't yet know what theory is correct, but we do know that universal gravitation isn't. In contrast Flying Spaghetti Monster theory can't be wrong because whatever we observe, no matter what it is, is always explained as the work of His Noodly Appendage. FSM Theory is unfalsifiable, not science, "not even wrong", and therefore does not have to be taken seriously by rational naturalists.
Inaccurate: "Scientists perform experiments to prove that a theory is true."
More Accurate: "Scientists observe tests to determine which theories are false."
I think these misconceptions are the result of insufficient elementary science education. Science is presented to young students in the simplified form of "theory, experiment, proof" for the same reason that we present math to young students in the simplied form of "addition, subtraction, multiplication and division." Curricula must be age-appropriate, and young students shouldn't be overwhelmed by hitting them with everything at once. We eventually learn that math involves quite a bit more than four basic operators, but for some reason relatively few people learn that "theory, experiment, proof" is oversimplified as well, causing many people to either think that science is only test tubes and lab coats, or that knowledge is more than test tubes so anything goes. Anything does not go. The strength of evidence - whether it be lab results or astronomical observations or archaeological specimens or field reports - must always be capable of being evaluated objectively and epirically.
The first is the misconception that all scientists perform experiments in laboratories. This is sometimes true, but not always. All scientists perform tests, not experiments. Some of those tests are controlled laboratory experiments, some are field experiments, some are natural observations, and some are observational studies. The goal in all cases is to make observations. This allows scientists to study things that don't fit into a laboratory, like supernovas or tribes of pygmies.
For example evolutionary biologists are scientists. A few evolve creatures in a laboratory, but most attempt to observe evidence of evolution in the natural world. A group of scientists studying devonian speciation realized that they ought to find an intermediate "fishapod" species spanning the gap between fish and tetrapods on Ellesmere Island in Canada. They looked, and found tiktaalik. Jennifer Clack said "It's one of those things you can point to and say, 'I told you this would exist,' and there it is." Validating a theory's extremely specific predictions with observations of physical evidence is as good as science gets, even though it happened thousands of miles from the nearest laboratory.
The second misconception is that tests prove that a theory is true. This is inaccurate. Tests are not just about proving whether a theory is true, because many theories are true to varying degrees. Good tests determine which hypothesis to favor: the null hypothesis or the alternative hypothesis. A test is not useful if its outcome is ambiguous; if its result doesn't indicate which explanation is more likely to be true.
For example the decomposition of mercuric oxide was evidence of Lavoisier's theory of oxygen, but it was also claimed by Priestly as evidence of phlogiston theory. Lavoisier needed to show that the gas produced in the decomposition was oxygen in order to disprove phlogiston theory.
The third misconception is that tests prove which theory is true. This is inaccurate. Theories predict observations, and tests reveal whether those predictions can be observed. Tests do not prove which theories are true, they allow theories to resist being proven false. A theory that passes a million tests is not necessarily "true" while a theory that fails even one test can be suspected of being false ( to some extent). This is why theories which are not falsifiable are not considered science - because unfalsifiable theories are not even wrong. If you can't be wrong you can't be right either.
For example Newton's law of universal gravitation consistently resisted falsification for almost 300 years, until observations like the Pioneer Anomaly indicated that a new theory was needed to correct it. We don't yet know what theory is correct, but we do know that universal gravitation isn't. In contrast Flying Spaghetti Monster theory can't be wrong because whatever we observe, no matter what it is, is always explained as the work of His Noodly Appendage. FSM Theory is unfalsifiable, not science, "not even wrong", and therefore does not have to be taken seriously by rational naturalists.
Inaccurate: "Scientists perform experiments to prove that a theory is true."
More Accurate: "Scientists observe tests to determine which theories are false."
I think these misconceptions are the result of insufficient elementary science education. Science is presented to young students in the simplified form of "theory, experiment, proof" for the same reason that we present math to young students in the simplied form of "addition, subtraction, multiplication and division." Curricula must be age-appropriate, and young students shouldn't be overwhelmed by hitting them with everything at once. We eventually learn that math involves quite a bit more than four basic operators, but for some reason relatively few people learn that "theory, experiment, proof" is oversimplified as well, causing many people to either think that science is only test tubes and lab coats, or that knowledge is more than test tubes so anything goes. Anything does not go. The strength of evidence - whether it be lab results or astronomical observations or archaeological specimens or field reports - must always be capable of being evaluated objectively and epirically.