Pedestrian traffic fatalities

Between 2010 and 2019, the number of Chinese students studying abroad and pedestrian traffic fatalities in the US both increased with a correlation of 0.9739. This is the kind of finding that gets a conference presentation titled something very careful, like 'Notes Toward a Preliminary Framework for Reconsidering.' The causal arrow has been described as 'unclear' by everyone who has looked at it, which is the academic equivalent of backing slowly out of a room.

US pedestrian fatalities actually bucked the overall traffic safety trend, rising from around 4,800 in 2010 to over 6,500 by 2019, driven by distracted driving, SUV proliferation, and urban planning that prioritized vehicle throughput. Chinese outbound student numbers grew from roughly 285,000 in 2010 to over 700,000 by 2019 before tightening visa policies and early pandemic effects began to bite. Both numbers were moving in the same direction — upward — over the same decade, which is sufficient to produce a high correlation with no mechanism required. They are two separate trends that happen to share a decade.

A correlation coefficient is a measure of shape-matching, not of contact. Two lines that rise together are geometrically similar whether or not they have ever been in the same room.