Affective Publics, Strategic Framing, and Reputational Dynamics: A Sentiment Analysis of Directorate of Criminal Investigation on X (@DCI_Kenya)

Abstract

This study examines public discourse on X (formerly Twitter) directed toward the Directorate of Criminal Investigations (@DCI_Kenya) between July 2023 and June 2025, analyzing how sentiment, thematic framing, and institutional engagement shape digital reputation. Employing a mixed-methods design, the research utilized a census approach to analyze 165,842 public replies alongside qualitative and computational techniques, including multilingual text normalization and frame analysis. The findings reveal a structurally complex sentiment landscape: while neutral replies dominate in volume, they frequently mask sarcasm and distrust. Fear, despite its low frequency, exerts disproportionate influence by accelerating the viral reframing of institutional narratives. Thematic analysis indicates that posts regarding public safety and gender-based violence attract heightened emotional density, whereas justice-oriented frames encourage deeper interpretive engagement. Significantly, the study found the DCI’s communication to be largely non-dialogic, characterized by a formal tone, minimal follow-up, and limited responsiveness to emotionally charged feedback. Reputation emerged not as a static asset but as a dynamic interplay of early emotional convergence and response timing; delayed institutional replies often allowed negative frames to consolidate, while positive sentiment proved fragile without narrative reinforcement. The study contributes to scholarship on digital policing in African contexts and offers practical recommendations to improve emotional monitoring, narrative anchoring, and dialogic responsiveness, thereby enhancing institutional credibility.