Not really related to Physics, but more relevant for anything less direct than the hard sciences: don't forget about the evil twin: the dangers of false empiri

Not really related to Physics, but more relevant for anything less direct than the hard sciences: don't forget about the evil twin: the dangers of false empirical confirmation. People, organizations, or just inherent bias will juke the stats. In terms of human impact, it's probably far more detrimental.