The analysis, led by researchers at the Annenberg School of Education at the University of Pennsylvania, showed that papers by men were over-cited and papers by women were under-cited relative to the citation rates you would expect if references were drawn randomly and gender was not affecting citation practices.