Exploring the Use of Topic Modeling and Sentiment Analysis in Research on Implicit Leadership Theories

Implicit Leadership Theories (ILTs) refer to individuals’ subconscious beliefs about the traits and behaviors of effective and followable leaders. These cognitive schemas shape how leadership is perceived, evaluated, and enacted. Traditional ILT research has relied on self-report surveys and interviews, which often face limitations such as subjectivity and small sample sizes. This paper explores the potential of topic modeling and sentiment analysis as scalable, data-driven alternatives for ILT research. Topic modeling enables the extraction of dominant themes from large textual datasets, allowing researchers to identify implicit leadership prototypes and cultural nuances without manual coding. Sentiment analysis complements this by quantifying the emotional tone in leadership descriptions, revealing implicit biases and evaluative framing. Combined, these methods offer a more comprehensive understanding of leadership perceptions in natural language use. The findings suggest that topic modeling and sentiment analysis can uncover patterns in leadership discourse that traditional methods may overlook. This integrated approach enhances the objectivity, depth, and cross-cultural applicability of ILT research, offering new methodological tools for studying leadership schemas at scale.