The present study analyzes news articles about sexual violence published by online news media in Spain. Our goal is to get insights about the way in which sexual violence is covered and described, with a focus on biases described by previous research. We begin by collecting about 120,000 messages on Twitter ("tweets") posted during 2020 by 13 of the most popular online news outlets in Spain. Next, we use a supervised classifier to detect tweets pointing to articles related to sexual violence, finding nearly 500 of them. We group these into clusters of articles that are likely to refer to the same event, and extract a series of descriptive elements from them using regular expressions. Finally, we compare these descriptors with official statistics about sexual violence in Spain. Our conclusions find biases that are well aligned with those described in the literature about the coverage of sexual violence in the news, indicating that the type of automated analysis we perform can uncover these biases correctly. For instance, the media covers sexual assault cases much more often than sexual harassment cases, despite the latter being more frequent. More worryingly, crimes happening at home are under-represented in the media, and crimes happening in leisure spaces are over-represented. In general, rather than presenting a balanced view of different types of sexual violence, media outlets perpetuate and reinforce harmful preconceptions and myths.

PDF