An insider’s view of science reveals why many scientific results cannot be relied upon – and how the system can be reformed.
Science is how we understand the world. Yet failures in peer review and mistakes in statistics have rendered a shocking number of scientific studies useless – or, worse, badly misleading. Such errors have distorted our knowledge in fields as wide-ranging as medicine, physics, nutrition, education, genetics, economics, and the search for extraterrestrial life. As Science Fictions makes clear, the current system of research funding and publication not only fails to safeguard us from blunders but actively encourages bad science – with sometimes deadly consequences.
Stuart Ritchie’s own work challenging an infamous psychology experiment helped spark what is now widely known as the “replication crisis,” the realization that supposed scientific truths are often just plain wrong. Now, he reveals the very human biases, misunderstandings, and deceptions that undermine the scientific endeavor: from contamination in science labs to the secret vaults of failed studies that nobody gets to see; from outright cheating with fake data to the more common, but still ruinous, temptation to exaggerate mediocre results for a shot at scientific fame.
Yet Science Fictions is far from a counsel of despair. Rather, it’s a defense of the scientific method against the pressures and perverse incentives that lead scientists to bend the rules. By illustrating the many ways that scientists go wrong, Ritchie gives us the knowledge we need to spot dubious research and points the way to reforms that could make science trustworthy once again.
In this bracing indictment, Ritchie (Intelligence: All That Matters), a lecturer at the Institute of Psychiatry, Psychology and Neuroscience at King's College, charges that in recent decades the scientific establishment has become host to a "dizzying array of incompetence, delusion, lies and self-deception." He names outright "fraudsters" such as condensed matter physicist Jan Hendrik Sch n, who falsified data supporting his 2001 claim to have invented a carbon-based transistor to replace the silicone microchip, and shows how scientists can more subtly massage data by "p-hacking," or nudging experiment results to show significant effect. Ritchie sees software such as Photoshop as having significantly contributed to the rise of deceptions, as when, in 2004, South Korean biologist Woo-Suk Hwang manipulated photos of cells to back up his claim of having cloned a human embryo. Regarding popular scientific literature, he takes aim at overhype, using as an example Matthew Walker's 2017 bestseller Why We Sleep, for using what he considers to be thin evidence to insist sleeping eight hours a night is not only healthy but vital. On academe, Ritchie suggests the pressure to publish has also encouraged scientists to exaggerate results. Thorough and detailed, this is a sobering and convincing treatise for anyone invested in the intellectual credibility of science.