Page 23 - Agricultural Census 2000 NWFP Vol II Part 1

SAMPLING PLAN This sample design was
basically the same as used in 1972,
It was considered necessary to 1980 and 1990 Agricultural
adopt different . sample designs for Censuses. However, minor
different parts of the country in view of improvements in the design and slight
the varied local conditions and changes in the size were made in the
availability of relevant information for design used for 1990 Agricultural
selecting a sample. Therefore, the Census to further reduce the
entire country was divided into three sampling variability without an
distinct parts for the purpose of appreciable increase in · the sample
sampling, detailed as below :- size.

The choice of three stage
a) Rural settled areas of NWFP,
sample design was based not only on
Punjab and Sindh Provinces;
sampling considerations but also in
where a three stage weighted and
view of a number of administrative
stratified sample was used.
considerations to control the non-
sampling errors. For example, the
b) Rural settled areas of Balochistan
selection of patwar circles in the first.
and Azad Jammu & . Kashmir;
stage made it possible to know, in
where a single stage weighted
advance, and regulate the number of
sample was used.
enumerators required. Similarly, the
selection . of two mouzas from each
c) Rest of the country, comprising
patwar circle helped in· equalizing the
entire urban areas and unsettled
workload per enumerator and
rural and tribal areas including
selecting clusters of households
Northern Areas; where a single
facilitated the enumeration work.
stage systematic sampl0 was
used.
This sampling plan called for
preparation of elaborate franles
In rural settled areas, sample involving great deal of effort and
size was determined at · sub-division
expense. The frames prepared
level and in rural unsettled areas at were:-
the district / agency / FR levels. In
urban areas, the sample size varied i) National Certainty Holdings (NCH)
according to the size of populace in list; comprising of all government
· an urban unit. In the three stage and private farms of 100 acres or
sample, patwar circles in the first more scattered all over the
stage and mouzas in the second country.
stage were selected with probability
proportional to size, whereas ii) Mouza list; containing only rural
households in the form of roughly mouzas arranged by patwar circles
equal clusters were seiected in the and kanungo circles showing
third stage with equal probability. mouza-wise total area, cultivated
area and number of · households.

(xx,)