01917nas a2200229   4500000000100000000000100001008004100002260000900043653003300052653002200085653001800107653001600125653001000141100001700151700001700168245008100185856006300266300001100329490000700340520132600347022001401673       2026                            d  c202610acomputational social science10adoctoral students10aMental Health10asupervision10ataboo1 aSaule Bekova1 aIvan Smirnov00a‘I Don't Want to Kill Any More Mice’: Taboo and Silence in PhD Education  uhttps://onlinelibrary.wiley.com/doi/abs/10.1111/hequ.70145  ae701450 v803 aImproving the experience and well-being of doctoral students requires a deep and nuanced understanding of their challenges. Traditionally, researchers have used reactive methods, such as surveys and interviews, to address these issues. However, some topics may be difficult to capture through these approaches, particularly those that are intentionally avoided or hard to discuss openly—what we might call ‘taboo’ topics. In this paper, we propose an approach to addressing this challenge by using nonreactive data sources, particularly from social media platforms and other online forums. We operationalise taboo as topics that are raised as explicitly anonymous questions by doctoral students, suggesting their prohibited or restricted nature. Our goal is twofold: first, to understand the whole range of challenges discussed by doctoral students on social platforms, and second, to determine which of these challenges are considered taboo by comparing explicitly anonymous and nonanonymous questions. By relying on large-scale computational analysis of social media data, our approach offers a comprehensive and unfiltered view of graduate student concerns and experiences. Our results highlight the prevalence of questions related to the taboo in academia, particularly concerning mental health and supervision.  a1468-2273