Journal of Official Statistics, Vol.8, No.1, 1992. pp. 97–103
Evidence of Anchoring in a Survey Recall Task
Carolyn M. Boyce and Marilyn C. Mauch
Abstract:The U.S. Department of Agriculture (USDA) asked farmers for bids to participate in a dairy herd buyout program. Once bids were submitted, USDA set a cutoff bid-$22.50-that achieved program purposes. Successful bidders were paid to slaughter or export their dairy herds and not dairy for five years. Unsuccessful bidders (bids over $22.50) did not participate in this dairy herd buy-out. About a year and a half after bids were submitted, a random sample of farmers was asked to recall their bids in a mail survey. Results were matched with USDA records of actual bids. Successful bidders, whose bids brought income and major life events (quitting the dairy), had unbiased recall of their bids. Recall for the unsuccessful bidders was significantly biased in the direction of the cutoff. They appeared to use the cutoff bid of $22.50 as a cognitive anchor, which they under-adjusted upwards, to reconstruct from memory the amount they bid.
Keywords:Response bias; underestimation; questionnaires; memory; judgmental heuristics.
Copyright © Statistics Sweden 1996-2018. Open AccessISSN 0282-423XCreated and Maintained by OKS Group