Dong Erico Yu, lecturer in the DACSS program, has published a new article with Yuehong Cassandra Tai of Penn State in Political Studies Review. The study uses original courtroom audio data, supervised learning, and a convolutional neural network to examine whether women and men judges differ in their Mandarin-language performance during criminal trials in China. The authors find little gender-based variation, challenging simple claims that women judges perform either better or worse than their male colleagues. They also find that more serious cases are associated with higher linguistic proficiency, a pattern consistent with strategic case assignment. The study shows how non-textual data, such as voice, accent, and speaking style, can provide new insights into judicial communication and institutional authority in a comparative perspective.