On Skepticism and Science
I consider myself a skeptic, and while many people are concerned about what skeptics don't believe, I think it's more useful to talk about a skeptic's criteria for belief. I see skepticism as antonymous with credulity. Whereas a credulous person is liable to believe assertions as soon as they hear them, a skeptic requires evidence, particularly with regards to extraordinary claims.
In reality, it's often a matter of practicality. If someone claims, "So, I was in the store the other day..." It's highly impractical and unreasonable to respond with skepticism. "Were you? Were you really?" But if someone claims to have literally seen a ghost, that's a highly extraordinary claim, and it's reasonable to withhold belief in the absence of objective, independently-verifiable evidence.
I want to be clear here. In this example, I'm not trying to say that ghosts don't exist. I simply don't believe that they do. I concede that I might be wrong, and I'm happy to change my mind provided my evidentiary criteria are met. So far, they have not. I think reasonable skepticism starts with a lack of belief that something is true rather than the presence of a belief that something is false.
Science is a way of skeptically interrogating the universe with a fine understanding of human fallibility.
I think it's also good to define 'extraordinary' here. We all generally live ordinary, mundane lives. Physics generally behaves as we've come to expect it to. Magical, supernatural events tend not to be a daily experience. Even those who claim to have had such experiences can usually count on one hand how many they've had, and those experiences are interesting to hear about precisely because they are beyond our ordinary experiences. When skeptics say, "Extraordinary claims require extraordinary evidence," this is generally what they mean.
Skepticism is a philosophy that is embedded into science. Carl Sagan said, "Science is a way of skeptically interrogating the universe with a fine understanding of human fallibility." Does that mean scientists themselves are perfect? By no means. They are still just as susceptible to bias as any other human, but science as an enterprise is reliable, particularly over longer time scales.
Some fundamental components of the scientific method drive that reliability. The first is the falsifiability of a hypothesis. If it isn't possible for a hypothesis to be unambiguously refuted by testing, it is weak and not likely to reliably tell us anything about the universe.
As a contrived example, if you were to hypothesize that a coin would come up heads every time you flip it, that is a trivially falsifiable hypothesis. It doesn't tell us anything particularly earth shattering about reality, but it is easy to test, and just getting tails once refutes it.
The existence of ghosts, on the other hand, is very difficult to falsify. We can't say with any certainty that there are conditions under which ghosts should reliably appear or otherwise provide evidence that they are or ever were there. Weak hypotheses are ripe for moving the goalposts. "Oh, the ghost didn't show up this time because their last appearance weakened them, or they didn't want to, or a skeptic was present and was disturbing the energy fields the ghost uses to communicate…"
Peer review is another component essential to making the scientific method reliable. People who have the requisite domain knowledge but who are not necessarily invested in the conclusion of particular research perform a review to assess the strength of the hypothesis, the completeness of the test methodology, and the soundness of the conclusions. Peer review acts as a bias filter and can reveal things that the original researchers couldn't see because they were too close to the project.
It's worth mentioning here that there are some concerns about the peer review process nowadays, with publications that have pay-to-play schemes. There's also the fact that some publications prefer to review 'sexy' topics over 'boring' ones. These should absolutely be aggressively tackled to keep science reliable and to preserve the public's confidence in that reliability.
All knowledge is provisional in science. It pays to keep an open mind, though not so open that just anything can fall in there.
That said, over the long term, improperly reviewed and published work is unlikely to stand. Particularly significant findings rarely ever exist in a vacuum, and science is an enterprise where current research builds on what has been done before. If there was unethical or simply biased research that doesn't reflect the true nature of reality, it will likely be discovered eventually as new scientists try to carry that research further.
That point segues nicely into the final component of the scientific method that ensures it's reliability: reproducibility. If results of a particular experiment cannot be duplicated, it calls the conclusions that were previously drawn into question. If you hear about a paper that draws particularly shocking or extraordinary conclusions, it's worthwhile to wait for results to be duplicated before simply accepting those conclusions as facts.
As an example, in September of 2011, a 170-member team working with a particle detector in Italy claimed to detect neutrinos traveling at speeds slightly faster than light. For those familiar with the theory of relativity, these results should be particularly astonishing. Faster-than-light travel is effectively impossible if relativity is true. Relativity has been demonstrated to be true time and time again using various methods, so skepticism at results that seem to violate it is reasonable.
In five separate cases, other teams failed to duplicate the original team's results, instead verifying that neutrinos do NOT move faster than light. It turns out, a loose fiber optic cable affected the original team's timing system, resulting in a delay that rendered their results moot.
But even though the example above suggests that well-attested ideas in science should be preferred over new ideas, it's necessary to clarify that there are no 'sacred truths' in science. If a new idea explains the universe better, makes better predictions, and the evidence holds up to objective scrutiny, it will eventually supplant the older idea. All knowledge is provisional in science. It pays to keep an open mind, though not so open that just anything can fall in there.
Ultimately, the skeptical framework that science is built on, coupled with its self correction mechanisms, makes it the most reliable method humans have ever devised to understand the universe. That does not mean that all scientific knowledge constitutes absolute, incontrovertible truth. It could someday be discovered that every conclusion rendered by science thus far has been incorrect. Even so, its reliance on evidence and objectivity as fundamental criteria means belief in its conclusions are thoroughly rational and justified. And if such a revolutionary discovery were made, no one would be more excited about the new horizon of exploration that lay before humanity than the scientific community.