Abstract: Discrepancies in career success across genders persist, particularly at high-ranked positions in business, academia and politics. Given the importance of persuasion skills in those contexts, such discrepancies raise the question of whether differences in persuasion tactics and evaluations of persuasiveness across genders play a role. To answer this question, I investigate the persuasiveness of men and women in the well-defined setting of competitive inter-collegiate debate tournaments in a three-fold manner: (i) collect a data set of around 2000 verbatim transcribed debate speeches (roughly 3.5 million words) and associated speech scores from the most prestigious British Parliamentary style debate tournaments from 2008 to 2018; (ii) analyze the links between words, arguments and various speech performance measures in these speeches using natural language processing (NLP) techniques, i.e neural networks for classification and extraction of key phrases and arguments; and (iii) unravel, if any, the existence and magnitude of systematic gender differences in speeches and their evaluations. These insights in gender differences in speech delivery and evaluations could contribute to the design of more equitable policies in real-life competitive settings. In the seminar, I will discuss the rationale and analysis framework, the descriptive statistics of the entire data set and preliminary analysis of a selected sample.