When the underlying data they rely on is incomplete—and it often is—the growing use of machine learning tools in America's criminal justice system can have devastating effects.
Overdoses on opioid drugs—including heroin, prescription painkillers, and certain street drugs—were the top cause of unintentional, preventable deaths in the United States in 2017.
The reproducibility crisis is driven, in part, by invalid statistical analyses that are from data-driven hypotheses.
Environmental Protection Agency enforcement fell to 10-year lows in 2017, but 2018's numbers may be even lower.
City, state, tribal, and other law enforcement agencies reported more than 7,000 hate crimes to the FBI in 2017, according to numbers the FBI released on Tuesday.
Statisticians have shown that many scientific findings are wrong, and without an increase in statistical know-how for scientists it'll continue happening.
Modern statistics have made it easier than ever for us to fool ourselves.
The count of the dead in Syria and other global trouble spots matters a great deal, and so does the provenance of the numbers presented. Here's one exhaustive effort to get them right.