The Streetlight Effect is closely related to, but distinct from, survivorship bias: survivorship bias involves analyzing only the cases that remain visible after a filtering process, while the Streetlight Effect is about actively choosing an easy search location. It also overlaps with the availability heuristic, though that concerns how readily examples come to mind rather than where one deliberately looks for data.
Streetlight Effect is a cognitive bias and observational error in which people look for evidence, data, or solutions in the places that are most convenient to search, rather than in the places where the relevant information actually resides. The name comes from an old joke about a drunk man searching for his lost keys under a streetlight; when asked whether he dropped them there, he replies that he dropped them in the park, but the light is better here. The bias matters because it quietly shapes which questions get asked and which answers get found.
The effect works because search has costs. Accessible data, familiar methods, and measurable variables all reduce effort and uncertainty, so attention gravitates toward them. Over time the easy-to-measure becomes a stand-in for the important, a substitution people rarely notice because the resulting work still looks rigorous. Institutional incentives amplify the problem: funding, publication, and performance reviews reward producing results quickly, and the quickest results come from well-lit ground.
Medicine and public policy offer concrete illustrations. Drug safety monitoring has historically relied heavily on voluntary adverse-event reports, which capture only a small and unrepresentative slice of real harms, simply because that data stream exists and is cheap to query. Similarly, much of early psychology research drew on undergraduate volunteers—convenient, nearby participants—leading to concerns that findings described a narrow slice of humanity rather than people in general.
To counter it, separate the question “what do I want to know?” from “what data do I have?” and write the first down before consulting the second. Ask explicitly what evidence would be decisive, then estimate what it would cost to obtain it; sometimes the expensive search is worth it. Naming the gap between your metric and your actual goal, and acknowledging it openly in conclusions, keeps the dark areas visible even when you cannot search them.
