United States
Department
01
Aarieulture
National
Agricultural
Statistics
Service
RCllelU'ChDiYiaion
SRB Research Report
Number SRB-93-02
January 1993
EVALUATION OF mSTORICAL DATA USE IN
THE 1992 AUGUST YIELD SURVEY
Jeffrey T. Bailey
C - Significant difference in corn acreage expansions
S - Significant difference in soybean acreage expansions
EVALUATION OF HISTORICAL DATA USE IN THE 1992 AUGUST YIELD SURVEY, by
Jeffrey T. Bailey.
Research Division, National Agricultural Statistics Service, U.S.
Department of Agriculture, Washington, D.C. 20250. January 1993. Research Report
No. SRB-93-02.
ABSTRACT
The National Agricultural Statistics Service (NASS) is evaluating the use of historical
data in its agricultural surveys to improve data quality and reduce respondent burden.
Historical data are any data previously collected for an operation that could be used
to guide or check a current response. Previous research has shown (O'Connor and
Mergerson, 1992) that the use of historical data as a real-time editing tool in Computer
Assisted Telephone
Interviewing
(CATI) with the 1988 March Agricultural Survey
produced estimates closer to the "truth" for grain stocks. Assuming that historical data
used in the same manner will result
in responses closer to the '1ruth" for other
commodities
and surveys, we need only to determine
if it affects final response
significantly.
This research analyzed data from the 1992 August Yield Survey. The
differences between the responses prior to and after a historical data check were
examined to quantify the direct impact of historical data use. The results showed there
was a statistically significant difference between the initial and final August responses
for both corn harvested and soybean planted acres in about 1/2 of the states.
Soybean planted acres were significantly different at the U.S. level. States offset each
other for corn harvested acres so it was not significant at the U.S. level. The final
response for soybean planted acres was much closer to the June estimated acres
indicating