Abstract
Citation frequency is widely recognized as a crucial metric for assessing academic impact. Previous studies analyzing data from citation databases have observed a surname order bias-a phenomenon where the alphabetical ordering of researchers' surnames negatively impacts their citation counts. However, the underlying mechanisms driving this bias, the causality behind it, and its implications for in-text citation practices remain poorly understood. Therefore, the present research aims to address these gaps through two preregistered studies. Study 1 replicates and extends the work of Stevens and Duque (Psychonomic Bulletin & Review, 26, 1020-1026, 2019), using a larger sample of 446,755 articles and controlling for surname initial frequency and publication year. Study 2 is an experiment with 307 valid responses from academics holding doctoral degrees, manipulating both citation systems and surname alphabetical order. Consistent and robust findings emerged across both studies: articles authored by individuals with surnames appearing earlier in the alphabet were more likely to be cited. This effect was especially pronounced in the context of alphabetical citation systems, compared with numerical citation systems. The current research provides a testable, reliable explanation for the surname order bias and establishes a causal link between surname alphabetical order and citation frequency. Implications for theory and academic practice are discussed.