Page 242 - Contributed Paper Session (CPS) - Volume 7

CPS2073 Alban Çela et al.
5. Improvement of quality in the CA estimation
Calibration estimation, whereby sampling weights are adjusted to
reproduce known population totals, is commonly used in survey sampling. As
described in the article by Deville and Särndal (1992), calibration can be
treated as an important methodological instrument, especially in large-scale
production of statistics.
In Graph 1 is presented the difference between Census Raw data versus
Administrative data provided by MARDWA, as well as the difference between
Census Calibrated data versus Administrative ones for the total area. The total
area is defined as the cumulative sum of Kitchen garden; Arable Land;
Permanent grassland; Permanent crops; Unutilized agricultural area; Wooded
area; Non-agricultural area.
Graph 1: CA total area vs administrative data

Census Calibration vs Administrative 76.4%

Census Raw vs Administrative 61.5%

0.0% 20.0% 40.0% 60.0% 80.0% 100.0%

Results have been improved approximately 14.8 percent from the
calibration process. These results will directly impact positively the sampling
frame for future agriculture surveys.

6. Conclusions
Taking into consideration problems faced for the CA of 2012 as well as the
actions taken to improve quality of the indicators produced the following
conclusions are reached:
 Calibration uses for CA improved the CA estimation and they are more
in line with the historic data from the Ministry.
 In the preparatory phase of CA, the alternative methods for the logical
controls, extreme value analysis, post data collection checking of the
farms, post stratification needs to be evaluated, to improve the quality
of the estimations.
 The frame for the next Census of Agricultural needs to be based on an
Administrative data for the agricultural enterprises/big farms and
Population Census for household/farms

229 | I S I W S C 2 0 1 9