A comment on "'Stop the Count!' - How Reporting Partial Election Results Fuels Beliefs in Election Fraud"
This replication report examines the reproducibility of Vaz et al. (2024), who demonstrate the Cumulative Redundancy Bias (CRB)—the tendency to favor early-leading candidates and to suspect fraud when a late reversal occurs. Using the authors’ openly available raw data and analysis code, we successfully reproduced all reported results across seven preregistered studies. The only discrepancy identified was a minor difference in Experiment 5, which did not affect the results of the study. In addition, the authors conducted non-parametric tests for all experiments in the complementary documents. We verified and reported their results, concluding that the authors have followed good practices for the replication of their outcomes. The non-parametric tests support the original findings: early leaders were consistently evaluated more positively, and latelead scenarios reliably increased perceived manipulation, even when using real election data, presenting results as percentages, providing explanatory information, or accounting for partisanship. The evidence strengthens the conclusion that sequential reporting of partial results can distort perceptions of electoral integrity, underscoring the importance of communication strategies that better convey vote-count dynamics and uncertainty.