The 2017 Kenyan elections saw a rise in social media monitoring of political violence compared to traditional sources. This study compares Twitter reports on violence with data from ACLED during this period.

Comparative Analysis

* Examines reporting strengths and weaknesses across two major election periods (August & Oct 2017)

* Focuses on geographical reach and timing accuracy as key dimensions

Key Findings

• Twitter data more concentrated geographically, especially in Nairobi and affluent regions during the immediate electoral period.

• ACLED reports show broader coverage across Kenya, including rural areas, with greater consistency before, during, and after elections.

• Both sources have limitations: neither fully captures overall violence levels; Twitter's advantage fades post-elections despite its timeliness.

What This Means for Violence Research

Combining social media monitoring and traditional data aggregation provides a more complete picture of political violence. The study advocates for mixed-method approaches in future research.