A supervisor’s command to “prove it again” (that is, run another experiment) often leads female scholars to be much more careful in their work, but also less productive as a result. Therefore, the extra labor required for better writing extends the publication process an extra six months, further decreasing women’s output relative to men’s.