Algorithm Reliance: A Cross-Country Replication with Neuro Evidence

We propose a pre-registered replication and extension of Snyder et al. (2025) to investigate the behavioral and cognitive drivers of algorithm reliance. Beyond the original queueing system framework, this study introduces a cross-country dimension by comparing an advanced economy (the original U.S. sample) with an emerging market (China). This comparison allows us to test whether the significant global disparities in AI trust and acceptance moderate the "learning by using" process and the eventual adoption of algorithmic advice. Furthermore, we extend the original behavioral findings with neurophysiological evidence using eye-tracking technology. By identifying physiological indicators of the shift from deliberate, "slow" consultation to "fast" defaulting, we aim to map how workers manage limited cognitive resources, as conceptualized by Boyacı et al. (2024), under varying levels of system load. Our results seek to provide a more nuanced, culturally sensitive, and cognitively grounded understanding of human-algorithm collaboration in modern service environments.