"None of the answers I gave were precise. Quite a few of them were guesses or "close-enoughs." If the statisticians are using the data from 55,000 U.S. households to calculate the official, important, complete, and accurate final results, and 55,000 real-live, human, unique participants are approximating numbers and shrugging their shoulders about which radio button to select, then how helpful is that information?"
Really great writeup from Indi Young on her own experience as a survey respondent for the Census Bureau. Entertaining, insightful, and thought-provoking.
Read the whole thing here: Who Can Believe the U.S. Unemployment Figures?