Science is performed by humans, and like anything done by humans, it has some flaws.
- Replication crisis
- Publication bias — researchers have a tendency to publish mainly positive results, and withold negative results. If one later reviews all available publications, the published material tends to support a hypothesis, regardless of whether it's actually true.
- (science reporting) Studies that show correlation are often reported in a way that implies that causation might have been proven (not to intentionally deceive, but just to "sex up" their stories). Then, due to the plebian confusion regarding causality-vs-correlation, readers tend to misinterpret the implication as fact.
- It's often impossible to draw clear lines between protoscience and pseudoscience — there may be no universal set of rules that qualifies something as scientific
- well-known hoaxes
- Bogdanov Affair, a deliberate hoax intended to target weaknesses in the peer review system. "The Bogdanov's papers consist of buzzwords from various fields of mathematical physics, string theory and quantum gravity, strung together into syntactically correct, but semantically meaningless prose."
- the Sokal affair was a similar (but less damning) version that preceded it
- Project Alpha — attempts to review paranormal activity in a scientifically-valid manner are shown to be flawed by two hoax spoon-benders. (lesson: it's outstanding that some scientists want to seriously review fringe-science, but in those cases, it is especially important to avoid pathological science)
- of course, there are countless examples of scientific misconduct for personal benefit only