A comparative assessment of fuzzy set-theoretic rule validation : analyzing fs/QCA software procedures and the modified consistency and coverage measures : evidence from ESG-stock return relations and online shopping behavior configurations

Abstract

This thesis investigates the robustness of fuzzy-set Qualitative Comparative Analysis (fsQCA) by comparing traditional rule validation in Ragin’s fs/QCA software with the modified fuzzy validation measures proposed by Stoklasa et al. (2017, 2018). While fsQCA is widely used for analyzing configurational causality, concerns remain about its ability to address logical contradictions and ambivalent configurations. To explore this, the study conducts a dual-case empirical analysis using ESG scores and stock returns from firms in the EUROPE STOXX 600, in addition to the online shopping behavior data from Pappas et al.'s research (2016).

Following standard fsQCA procedures, configurations are first evaluated using traditional consistency and coverage metrics. These are then reassessed using Stoklasa et al.’s modified consistency and coverage metrics (2017, 2018), namely F2, F3, and F4 in MATLAB. The comparison suggests that while fs/QCA software offers a structured approach, it may not fully capture borderline or contradictory patterns. Stoklasa et al.’s framework enhances diagnostic sensitivity and interpretive depth in these aspects.

This thesis contributes by demonstrating the value of the mentioned additional fuzzified rule validation as a complement to the standard process in fs/QCA software, with a practical procedure to apply it and by highlighting key considerations when relying solely on default software outputs. The findings support the idea of integrating the modified rule validation technique from Stoklasa et al. as a best practice to improve the reliability and transparency of fsQCA research.