SRB Research Report
ANALYSIS OF RESPONSE BIAS IN THE
JANUARY 1993 CATTLE ON FEED SURVEY
Robert G. Hood
ANALYSIS OF RESPONSE BIAS IN THE JANUARY 1993 CATTLE ON FEED SURVEY
QUALITY ASSESSMENT, by Robert G. Hood, Survey Quality Research Section, Survey
Research Branch, Research Division, National Agricultural Statistics Service, United States
Department of Agriculture, Washington, DC 20250-2000, August 1993, Report No. SRB-93-07.
A reinterview survey was conducted in January 1993 in Illinois, Iowa, Minnesota, Nebraska and
the January 1993 Cattle on Feed Survey Quality Assessment
(COFQA), was the third and final step of a three-phase plan to implement an operational
program for cattle on feed inventory in the farmer-feeder
states. The purpose was to evaluate
the quality of the survey data
through response bias estimation and to gather cognitive
information on reporting problems associated with cattle on feed.
A subsample of respondents
to the January Agricultural Survey was recontacted for face-to-face
reinterviews in which a subset of the original questions was re-asked. Differences between the
reinterview responses and the initial survey responses were reconciled to determine a final "true"
value which was used to measure response bias. The January 1993 survey featured refinements
in the sample design to provide "richer" data from the small sample size. Both CATI (Computer
Assisted Telephone Interviewing) and non-CA TI domains were eligible for reinterview.
Overreporting of COF capacity was the only significant response bias detected at the five-state
Significant state level overreporting
for capacity was indicated in all five states
and for steer inventory in Iowa, while underreporting
for cattle on feed inventory in Illinois and
for steer inventory in Nebraska was indicated. Although many reporting problems were found