Why does so much science turn out to be wrong?

It’s Wednesday: eat more kale because it’s a superfood and super good for you. It’s Thursday: put that kale away, it can cause problems like chronic fatigue and consuming it can be toxic to your health.

Kale’s utter goodness or fearsome badness should be a fairly simple and straight-forward matter of science. But contrasting studies have both praised and condemned the leafy green.

How is it possible for science to be wrong so often?

University of British Columbia microbiologist Rosie Redfield believes that bad science has been around forever and it’s not a new phenomenon. But because of the Internet, when scientists do get their facts wrong, it’s simpler and faster to discover the errors.

“With social media and with everything online, it’s a lot easier for find things and it’s easier for us to talk to each other,” Redfield said.

Redfield was one of the main scientists who helped debunk a controversial scientific paper put out by NASA and the journal Science back in 2010. The paper upended the belief of what had been the chemical basis of living things.

The NASA-funded research at the time was cited as changing the fundamental knowledge about what comprises all known life on earth after scientists conducted tests in the harsh environment of Mono Lake in California.

There, they discovered the first known microorganism on earth able to thrive and reproduce using the toxic chemical arsenic.

The microbe grew with arsenic instead of phosphorus in its cell components. Before then, the chemical rules of life had dictated that six essential elements—carbon, hydrogen, nitrogen, oxygen, phosphorous and sulfur--were needed for all living things, including bacteria.

The arsenic discovery was so momentous and caused such a scientific commotion that calls to NASA from the White House and Congress asked whether a second line of earthly life was the result of the findings.

Because cells could grow in the lab setting using arsenic instead of phosphorus, according to the NASA research, that opened up the possibility that it was a different life form, perhaps from aliens.

However, it all turned out to be false.

Redfield had tried to replicate the research in her lab at the University of British Columbia but could not. It later turned out that the results in the NASA-funded research was contaminated.

Redfield says problems emerge when scientists fall in love with their hypothesis. In the arsenic case, Redfield thinks that bad science happened because the researcher wanted to believe in the result instead of setting out to prove a hypothesis.

“It was the perfect storm. Everything that could go wrong did go wrong. Everyone who was supposed to quality check the processes slipped up. Her supervisor trusted too much that she knew what she was talking about. The reviewers slipped up and the journal picked reviewers that may not have been very good.”

Redfield said the slip-ups that caused the bad science were all human errors that will happen again.

She tells her students that the most important thing they can do as scientists is not to assume they are right.

“What makes you a good scientist is, it doesn’t matter how smart you are; it is to be able to say, ‘maybe I’m wrong’.”

In 2010, science writer Adam Marcus and science journalism instructor, writer and medical doctor Ivan Oransky started Retractionwatch.com to keep track of the retractions in research reporting as a window into the scientific process.

“It was like turning over a rock and finding a lot of ants swirling around. There were a lot of stories about fraud and misconduct and honest error too,” Oransky said.

All papers are now online which wasn’t true 15 years ago, Oransky said, and plagiarism software can easily detect when there has been plagiarism and duplication.

Oransky said the major issue in scientific research these days is a reproducibility crisis. More than half of the studies in life sciences can’t be reproduced and held up to peer review, he said.

“Science is supposed to get things wrong,” he said. “Journals and journalists often use a particular paper as the gospel, as something that is always right.”

Oransky cites the response from John Maddox, the former editor of the prestigious journal Nature, to a question from a reporter once about how many errors are published every week.

“He said, ‘that’s easy. All of it.’ What he meant was over time everything gets overturned and we have different understanding of it. With very rare exceptions, that’s the way science is supposed to work.”