Common (mis)beliefs about memory: a replication and comparison of telephone and Mechanical Turk survey methods.

摘要: Incorrect beliefs about memory have wide-ranging implications. We recently reported the results of a survey showing that substantial proportion United States public held conflicted with those experts. For survey, respondents answered recorded questions using their telephone keypad. Although such robotic polling produces reliable accurately predicts elections, it suffers from four major drawbacks: (1) is costly, (2) typically, less than 10 percent calls result in completed (3) do not reach households without landline, and (4) oversample elderly undersample young. Here we replicated our Amazon Mechanical Turk (MTurk) to explore similarities differences sampled demographics as well pattern results. Overall, neither closely approximated population, but they differed how deviated 2010 census figures. After weighting each conform demographics, though, two approaches produced remarkably similar results: In both surveys, people averaged over 50% agreement statements scientific consensus shows be false. The this study replicate finding discrepancies between popular experts surveys conducted on MTurk can produce representative sample population generates line more expensive techniques.