Bad Science

Ben Goldacre argues that the public's ability to evaluate health claims has been corrupted not primarily by a few charlatans but by a systemic failure to understand how evidence actually works: what a controlled trial is, why randomization and blinding matter, how placebo effects fool even sincere practitioners, and why cherry-picked positive studies mean nothing once you account for publication bias. Using homeopathy, nutritionism, and pharmaceutical marketing as recurring case studies, he shows the same reasoning failures recurring again and again, whether the culprit is a quack selling sugar pills or a drug company burying unfavorable trial results.

This matters because bad statistical reasoning has real consequences, not just wasted money: Goldacre documents patients abandoning effective treatments for malaria prophylaxis or childhood vaccination because alternative practitioners misrepresented the evidence, and he traces how science journalism's appetite for simple, dramatic stories routinely distorts findings that were actually cautious and provisional. His remedy isn't cynicism about all science but a specific, teachable literacy: understanding regression to the mean, meta-analysis, and why anecdote is not evidence, so readers can evaluate claims for themselves rather than trusting authority or headlines.

Who should read it

Anyone who reads health news, takes supplements, or has ever wondered whether a friend's miracle cure actually worked will benefit, as will journalists and students who want a rigorous but accessible introduction to evaluating evidence. It's especially useful before making any medical decision based on a news story or product testimonial.

About the author

Ben Goldacre is a British physician, academic, and science writer who wrote the long-running Bad Science column for The Guardian and has researched evidence-based medicine at Oxford.