Nature revealed last year that 2% of scientific papers – tens of thousands – are not as true as you’d think they are. Hannah looks into the situation at hand