Yes, self-reported results will have biases. We might also note selection bias if all the surveys went to neighborhoods in the same city or even the same state,

Yes, self-reported results will have biases. We might also note selection bias if all the surveys went to neighborhoods in the same city or even the same state, for example. Americans have (or at least had) cultural mores against discussing the topics of personal finance, income, and net worth at all. Where would you suggest starting? Statistics requires training and does have some intuitively nonobvious results, such as determining sample size[0]. If you prefer more interactive, Survey Monkey has a form[1] that lets you select population size, confidence, and margin of error. For the U.S. population of 325 million at 95% confidence level and 5% margin of error, the necessary sample size is 385 — not far from your figure Stanley’s. The topic warrants future study, but Stanley’s book does destroy the notion that all millionaires are a bunch of Paris Hiltons and Thurston Howells — for those readers who are willing to challenge their own prejudices. [0]: https://en.wikipedia.org/wiki/Sample_size_determination https://en.wikipedia.org/wiki/Sample_size_determination [1]: https://www.surveymonkey.com/mp/sample-size-calculator/ https://www.surveymonkey.com/mp/sample-size-calculator/