Apr 10, 2012
This article isn't specific to audit sampling, but to statistical sampling in general. It is very timely as the 1940 census was just released a few days ago. An interesting note in this article is that they placed the "offensive" question regarding income towards the end of the census form in 1940. Given that they expected the census taker to be kicked out after this question, what would the impact be on the rest of the data collected? Does it stand or is it negated as an incomplete form.
The article also touches briefly on flawed samples, specifically the Literary Digestpoll which predicted the easy victory of Alf Landon (who?) over Franklin D. Roosevelt in the 1936 Presidential Election.