Science Journalist Matt Kaplan Calls for Greater Transparency in Scientific Research at Harvard Talk

Science journalist Matt Kaplan said scientists and journalists should be more transparent about how research works during a Thursday evening book talk at Harvard.

“I think the single most important thing that we can do, both as scientists and science journalists, is talk about sausage making,” Kaplan said, invoking the adage that laws and sausages are better not seen being made.

Kaplan spoke about his book — “I Told You So!: Scientists Who Were Ridiculed, Exiled, and Imprisoned for Being Right” — which traces historical and modern examples of scientists whose ideas were dismissed or attacked before later being validated.

The talk was hosted by the Faculty of Arts and Sciences’ Division of Science, Harvard Library, and the Harvard Book Store and moderated by human evolutionary biology professor Daniel E. Lieberman ’86, whose class Kaplan audited during his time as a Knight Science Journalism Fellow at Harvard.

Kaplan said the idea for the book began with his reporting on a 2015 psychology paper by French neuroscientist Jean Decety, which suggested children from religious homes were less charitable than children from nonreligious households.

Kaplan had published a piece on the findings in The Economist, after which Canadian psychologist Azim Shariff contacted him, asking to get in touch with Decety to ask about his results. Shariff identified a coding error that invalidated the results after Decety had freely shared the data with Shariff. Decety subsequently retracted the paper, Kaplan explained.

“I was struck by that, because this is how science is supposed to work, right?” Kaplan said.

But history, Kaplan argued, shows that science has rarely worked that way — a problem he said was thrown into sharp relief during the Covid-19 pandemic. Kaplan said that multiple medical experts shared promising ideas about the coronavirus with him but refused to go on the record — fearing professional retaliation from supervisors or colleagues in adjacent fields.

“I had one person in the third week of the pandemic say, ‘I think it's a cytokine storm that's killing people,’” Kaplan said. “And I said, that's great. I would really like to write that up. And the response was, you can't, because my idea would be a threat to the work being done by my PhD supervisor or other virologists would not take me seriously anymore.”

That experience, combined with his historical research, led Kaplan to a bleak assessment of the current state of the field. During the talk, he cited data showing more than 23,000 scientific papers had been retracted in recent decades.

When Lieberman asked how the scientific community should address fraud, Kaplan said creating an “academic immune system” to monitor research would be ideal but difficult to fund.

With “extraordinary reluctance,” he said he leaned toward legislative penalties.

“It’s relatively cheap to create legislation that says if you engage in fraud and are caught and convicted by my peers, then you go to jail,” he said.

Kaplan also expressed doubt that artificial intelligence could effectively detect fraudulent research.

“If we start using AI to try to detect the fraudsters, guess what the fraudsters are going to do? They are going to use AI to cover their tracks,” he said.

Lieberman closed the discussion by arguing that the scientific method itself remained the best tool against bias.

“We need to teach people the scientific method, because even though we all have our biases, we all have our problems, it is a way to try to help overcome that,” Lieberman said.