Computer algorithms can test the dodginess of published results

IN AN ideal world the data on which a scientific study is based should be, if not publicly available, then at least available to other researchers with a legitimate interest in asking. Sadly, this is not always the case. Though attitudes are changing, many scientists are still quite proprietorial about their data. They collected them, they reason, and thus they own them, and with them the right to analyse them without sharing them with rivals. This attitude, though selfish, is understandable. Bu...