This study examines how personality traits, social capital, and entrepreneurial passion affect loan evaluators’ signal perception in Dutch microfinance. Using a mixed-methods approach, we first analyze over 14,000 loan applications of Dutch small businesses and start-ups through Latent Dirichlet Allocation (LDA) to identify signals. Drawing on perception and signal theory literature we hypothezise how the identified signals are interpreted by loan evaluators. Through a qualitative content analysis we find that traits like conscientiousness, extraversion, and passion positively influence signal perceptions, while indicators of social capital, such as strong and weak ties, enhance credibility. However, reliance on heuristics introduces biases, potentially reinforcing disparities for marginalized groups. This highlights the importance of managing subjective evaluations. By integrating signaling theory with cognitive perspectives, the study underscores how intentional and unintentional signals shape decision-making in resource-constrained contexts. Methodologically, it demonstrates the value of LDA for extracting insights from unstructured data. The findings inform entrepreneurs on presenting favorable traits and guide financial institutions in improving fairness and consistency in loan assessments
