Page 34 - 2010 (3) Agricultural Census 2010, Punjab Report

serial numbers within Mouzal urban block were given manually by the staff of ACO.
Any serial number, Tehsil code missing or illegible was made correct during coding.
Then the questionnaires were tagged together for one Mouzal urban block to make
them ready for manual editing and further data processing steps.

MANUAL EDTTING:
Manual editing is.a process to make the data recorded on a questionnaire consistent
if inconsistencies arefound among data items within the questionnaire. For this
purpose manual editing instructions were prepared on the basis of theoretical mutual
relationships among data items. ln the light of those instructions manual editing was
performed by the staff of ACO to make the data ready for keying. Ambiguous and
illegible writings were also made correct during this step.

DATA ENTRY AND EDIT- CORRECTION:

After manual editing the data were entered in computers at Data Processing Centre
(DPC) by Keypunch Operators. Then the data were processed by the computer
programmers through already developed computer programmes for the purpose.
The programmers produced compute.q printouts as the first edits for each district.
These edits were checked at ACO with source documents and errors were made
correct, if found. The second edit wab produced by the programmers after
processing the errors fcund during the first edit and the process went on and on till
the data of a particular district were declared error free. The correction of data
through compuier edits tiil the data are error free with the help of source documents
is termed as edit-correction. After declaring data error free, tabulations were
produced.

TABULATIONS:

Tabulations are the actual presentation of data for data users. A tabulation plan
comprising of 90 tables was prepared at the time of census planning and
preparation. In the light of instructions already conveyed to the programmers for
each column and row of each table, they produced data tabulations for each district,
province and at country level as requested by the ACO.

ESTIMATION OF CENSUS RESULTS

As the census was based on sample enumeration, the sample observations were to
be extrapolated to the population level. Therefore, the raising factors determined at
various san,pling stages, representing the inverse of the probabilities of selection of
the respective household, were fed to the computer along with the other data. In
case of .a joint farm, 112, 113, 114 of the farm (or as the case might be) was
considered enumerated when one of the two, one of the three or one of the four
partners of a farm, respectively, was enumerated.

The terminology used in the estimation of aggregates along with the mathematical
formulae (as given below) for developing census estimates for three-stage sampling
plan (part-1 of the country) has been elucidated being the most complicated as
compared to the areas / regions of the country falling inpart-Z, part-3 and part-4.
Anyhow, the probability of selection as well as the procedure for working out census
estimates for part-2, part-3 and part-4 are implicit and may weli be understood fro-
the elucidated methodological formulae.
XXVI