Page 20 - 1960 (3) Pakistan Census Of Agriculture Vol 2 West Pakistan

-
~ ~

xv
-to note that the propaganda drive paved the way for a November, 1958 and it was decided to hold instead a.
-successful enumeration in which the farmers extended simultaneous census on a sample basis. Thus the
full cooperation. Census held in 1960 was taken on the basis of a 20 per-
cent sample of villages for the whole of West Pakistanc
'Type of Census except Kalat Division where census was taken in 1962.
on the basis of a 5 percent sample of villages.
13. An important question that engaged the
attention of the Census Organ}zation from the Sampling
very beginning was the type of the Census. The F.A.O.
expert who had visited Pakistan in 1952 had recommend- 16. A single stage sample of villages which was:
ed a progressive district-wise census to be completed in simple in design, was adopted. Complex design and multi-
five years. He discounted the idea of a comprehensive stage sampling had to be avoided as ·construction of the,
and simultaneous census, as he thought that since necessary frames, and sampling, could not be entrusted'
there was no previous experience of a Census of agri- to the field staff who were untrained in statisticai 1
culture in this country, it would be difficult to muster methods of that nature. A sample of villages was select-
sufficient number of trained persons to handle a gigantic ed with two-way stratific.ation by Assessment Circles anct
undertaking such as that of a _comprehensive simultane- by size of villages. All respondents resident in the
ous census. But his recommendation was not accepted sample villages were to be enumerated.
in view of the fact that a progressive census stretching
over five years was not likely to yield comparable data 17. In a sample census the occurrence of the·
for the various administrative units. Therefore, it sampling error is an unavoidabie factor and, therefore,
·was decided to hold a complete simultaneous census it was considered necessary to so design the sample-
of which the actual implications were yet to be tested. that the error be kept under a reasonable limit in order
In that direction a number of tests had to be carried that the usefulness of the data might not be impaired ..
-out from which a clear idea could be obtained as to Calculations were accordingly made on the basis of the
the magnitude of the task and the extent of financial test and pilot censuses and it was observed t1_1at the con-
and manpower resources required for an extensive ditions prevailing in the Province made it poss:ble t<>
undertaking of that nature. It was estimated that:- achieve that goal with a sampling fraction of 20 percent.
18. For the sake of sampling, the village was.
(i) Such a census would cost much more than taken as the unit which, in terms of respondents, meant
the sanctioned budget; the sum total of all the farmers and livestock holders.
residing within the boundaries of that village. Io_
(ii) About 30.000 enumerators and about 5,000 terms of land it meant total land operated by the resident
supervisors would be required to finish the farmers. In case any part of the farm of a resident far-
enumeration in 6 weeks as programmed; mer was located outside the village it was to be included
or alternatively in the enumeration in order to construct a complete
picture of the farm. Thus. the unit of enumeration
(iii) The period of enumeration would have to in the· census was the complete farms operated by the
be extended to 30 weeks if the number of resident farmers of the sample village.
enumerators was to be confined to the
available strength. 19. Since the village was adopted as the sampling:
unit a complete list of villages in the province had to
14. In the light of these considerl'l,tions the decision be prepared for the sake of a sampling frame. The-
to hold a complete simultaneous census had to be list was prepared out of reports sent by the officials.
reviewed. . It was agreed that the additional costs of the Revenue Department, containing information
required for that would not be available and besides, about the total area, cultivated area and culturable
it would not be possible to raise an army of 30,000 area of the viiiage. However, in the preparation oi
enumerators and 5,000 supervisors to work for a short the list, the villages wholly under Government forests.
period. Moreover, training and supervision of such and those forming part of urban areas were excluded.
a large staff would present serious problems which Thus the rural villages only, were sampled by
might affect considerably the quality of the census assessment circles and where there were no assessment
data. As regards the alternative of extending the circles such as in the Divisions of Hyderabad and
period of enumeration it had to be ruled out for three Khairpur, the taluka was retained as a stratum. The
reasons; firstly, the departments from which the field assessment circles or the talukas were joined together
staff for enumeration was to be borrowed could not to form one stratum of not less than 60 villages where-
afford to spare the -staff for more than a few weeks·; ver necessary. The villages within each stratum were
secondly, on account of the extension of the reference arranged by size of total culturable a:rea and were·
period the data collected over distant periods would subjected to random selection of one out of every
have become incomparable to some extent; and five. In all 6,865 villages were s~lected from the
thirdly, the popular support needed for a huge operation "settled" areas of the Province.
like the census could not be sustained for a period as
long as 30 weeks. . A separate sample of _515 villages was selected
from the "unsettled" villages of Quetta, Kalat and
15. In view of these considerations the proposal Karachi Divisions. In Quetta and Karachi Divisions.
for a complete simultaneous census was dropped in the sampling fraction of 20 percent was maintained.