Page 41 - 2000 (5) Agricultur Census 2000 Balochitan Province

sample mean. Suppose, if a a) Relatlve Variar:ce {R.V.}
number of estimates for an item Estimates at SubdivEsEom Level
are developed from all possibie
samples of same size drawn At the sub-division level, R.V.
with similar procedure from a can be estimated simply by first
population, their mean would be raising the sample abservaticns
equal to the mean based upon of each patwar circle to stratunr
complete count and their !evel using the overall raising
dispersion about that mean, factor for each holding but
measured in terms of variance, excluding all National Certainty
is called sampling variance. The Holdings {i{CHs}, and then
square root of sampling summing up to obtain an
variance is known as sampling estimate for each half stratum as
error.' under :-

Further, to compare precision of n 11 {Nr - N'r N"r
-
Qi
sample estimates ' of different lh'1 IIE L ) Yn,,jo'",
items, their samp!ing errors - jk
are brought on comparabie lt!l Pn,rjr.
'
terms by expressing thern
as percentages of relevant Q; {!fr \l
:s k
sample estimates. The resultant +I ) J fi'1jki!'
measure iS designated as eo- Zr n
K ltl rh'1j
:
efficient of va;-iaiion which can
be worked out directly from the
12 Q; tNr-N'r-N" x ) Yn,zir.r,
census data. For the Agrieuitural t-rr Fa
1h,2 L
Census 2000 co-efi!cie;ii of -/t/ aL4
IK ltt I Pn,zjr,
t
variation have been warked out
for a number oi imp**aglt iiems
TJ; Nt'L
and given in Appendix-ltl for f n'z:xy'
.Fr
fL {r L
enlighting the data u$ers ebout
k i" Prr'zi
accuracy of the concei':"rcrJ
estimate(s). The formuia used
then, R.V. for the sub-division
for estimating relative variance
estirnate cf aggregates of th*
{R.V.} is ciescrib*d in the atirlbute is cbtained by takirig
succeerjing parag raphs
"
square *f differ*nces between
esiimai*s fcr successi,;e pairs of

XXXVII