For example, in one study, subjects assigned the same probability to the likelihood of obtaining a mean height of above six feet [183 cm] in samples of 10, 100, and 1,000 men.
[2] Tversky and Kahneman explained these results as being caused by the representativeness heuristic, according to which people intuitively judge samples as having similar properties to their population without taking other considerations into effect.
A related bias is the clustering illusion, in which people under-expect streaks or runs in small samples.
[3] To illustrate this point, Howard Wainer and Harris L. Zwerling demonstrated that kidney cancer rates are lowest in counties that are mostly rural, sparsely populated, and located in traditionally Republican states in the Midwest, the South, and the West, but that they are also highest in counties that are mostly rural, sparsely populated, and located in traditionally Republican states in the Midwest, the South, and the West.
While various environmental and economic reasons could be advanced for these facts, Wainer and Zwerlig argue that this is an artifact of sample size.