Since science is a human enterprise, it is open to error, misinterpretation and, rarely but notoriously, fraud and fakery. Here’s a rundown of a few science mishaps, misapprehensions and debatable interpretations in recent years.

1. Faster-than-light neutrinos. In September 2011, scientists with an experiment called OPERA reported that neutrinos pumped from the CERN physics facility in Geneva had arrived at another laboratory in Italy some 60 nanoseconds faster than light could have made the same trip. But within months it became clear that a loose cable had thrown off the measurements.

2. Arsenic-based life. Scientists looking for a “shadow biosphere” reported in 2010 that they’d found bacteria in Mono Lake, Calif., that appeared to have swapped out phosphorous for arsenic in their DNA. Critics pounced on the NASA-funded research. The journal Science, which published the original paper, later published studies showing that the bacteria could not, in fact, survive without phosphorous.

3. Gravitational waves from cosmic inflation. This is an active controversy: Scientists at the South Pole saw signals of polarization in ancient radiation that were consistent with predictions that an early, violent, inflationary spasm of the big bang would leave behind such fingerprints. But a similar effect can be produced by foreground dust. So is this the whisper of cosmic creation or just schmutz in our eyes? More data needed!

In March, scientists from the Harvard-Smithsonian Center for Astrophysics announced their discovery of gravitational waves created at the dawn of the universe. These waves were created in a period of rapid expansion called cosmic inflation. This new evidence could prove the definitive confirmation of the inflation theory. But other researchers are not convinced. (Courtesy of Nature Video)

4. Goldilocks planets. In 2010, astronomers reported they had discovered a “Goldilocks” planet, not much bigger than Earth, that orbits a star in the habitable zone where it’s not too hot and not too cold for water to be liquid at the surface. But the Goldilocks planet, Gliese 581G, may also be not too real: A new report in Science, published Thursday, says that this congenial planet and another one believed to orbit the star Gliese 581 aren’t actually there. The report says the star has internal processes that created the illusion of planets. The new interpretation of the data remains in dispute, however, as the original team of scientists stands by its claim.

5. STAP stem cells. The journal Nature this week retracted two papers that reported claims that ordinary mouse cells could be exposed to a lightly acidic environment and turned into stem cells. The editors of Nature cited extensive errors, plagiarism and the inability of other researchers to duplicate the experiments. It is not clear now that the phenomenon of “stimulus-triggered acquisition of pluripotency” actually exists.

Nature’s editors, in retracting the papers, discussed the rash of recent incidents calling into question the reliability of scientific studies:

“Underlying these issues, often, is sloppiness, whether in the handling of data, in their analysis, or in the inadequate keeping of laboratory notes. As a result, the conclusions of such papers can seem misleadingly robust. Another contributory factor lies in selection bias behind the data presented, whether implicit because the experiment was not randomized or blinded, or explicit in the deliberate selection of data that, usually with honest good intentions, are judged to be representative.”

A reminder: Science as an enterprise is very good at catching its mistakes. That’s why science has worked so well for 400 years. But the most important science is done on the edge of the knowable, where results can be ambiguous and open to interpretation. If it was easy, it wouldn’t be great science.