Ben Goldacre: Using Benford's law, forensic statisticians can spot suspicious patterns in the raw numbers, and estimate the chances figures have been tampered with
Ben Goldacre: The same statistical errors – namely, ignoring the "difference in differences" – are appearing throughout the most prestigious journals in neuroscience
Ben Goldacre: Trials show that people who think they've done something healthy, even if they haven't, smoke more and believe they are invulnerable to diseases
Anecdotes are great – if they convey data accurately
Ben Goldacre
Ben Goldacre: Bad science: Channel 4 reported that a study in the Lancet showed a new drug had reduced the symptoms of Duchenne's muscular dystrophy. Unfortunately, the study shows no such thing
Ben Goldacre: We need a badger cull trial held in the real world ... to give us answers that matter, on results we care about, with the intervention we're actually using