Upload
others
View
20
Download
0
Embed Size (px)
Citation preview
J. Stat. Appl. Pro.3 , No. 2, 269-293 (2014) 269 Journal of Statistics Applications & Probability An International Journal
http://dx.doi.org/10.12785/jsap/030216
An Almost Unbiased Estimator for Population Mean using Known Value of
Population Parameter(s)
Rajesh Singh1, S.B. Gupta2 and Sachin Malik2, *
1 Department of Statistics, Banaras Hindu University, Varanasi-221005, India
2Department of Community Medicine, SRMS Institute of Medical Sciences, Bareilly- 243202, India
Received: 11 Feb. 2014, Revised: 22 May 2014, Accepted: 25 May 2014 Published: 1 Jul. 2014
Abstract: In this paper we have proposed an almost unbiased estimator using known value of some
population parameter(s) with known population proportion of an auxiliary variable. A class of estimators is
defined which includes [1], [2] and [3] estimators. Under simple random sampling without replacement
(SRSWOR) scheme the expressions for bias and mean square error (MSE) are derived. Numerical illustrations
are given in support of the present study.
Key words: Auxiliary information, proportion, bias, mean square error, unbiased estimator.
1. Introduction
It is well known that the precision of the estimates of the population mean or total of
the study variable y can be considering improved by the use of known information on an
auxiliary variable x which is highly correlated with the study variable y. Out of many
methods ratio, product and regression methods of estimation are good illustrations in this
context. Using known values of certain population’s parameters several authors have
proposed improved estimators including [4, 5, 6, 7, 8, 9, 10, 11, 12, 13].
In many practical situations, instead of existence of auxiliary variables there exit some
∗ Corresponding author e-mail: [email protected]
c 2014 NSP Natural Sciences Publishing Cor.
270 R. Singh, S.B. Gupta and S. Malik: An almost unbiased population parameter(s)
auxiliary attributes (say), which are highly correlated with the study variable y, such as
i. Amount of milk produced (y) and a particular breed of cow ( ).
ii. Sex ( ) and height of persons (y) and
iii. Amount of yield of wheat crop and a particular variety of wheat ( ) etc. (see [14]).
Many more situations can be encountered in practice where the information of the population
mean Y of the study variable y in the presence of auxiliary attributes assumes importance.
For these reasons various authors such as [15, 16, 17, 18, 19] have paid their attention
towards the improved estimation of population mean Y of the study variable y taking into
consideration the point biserial correlation between a variable and an attribute.
Let
N
1iiA and
n
1i
iφa denote the total number of units in the population and sample
possessing attribute respectively,N
AP and
n
ap denote the proportion of units in the
population and sample, respectively, possessing attribute .
To estimate Y , the usual estimator is given by
2y1SfyVar
(1)
Define,
Y
Yyey
,
P
Ppe
,yi,0eE i
,CfeE 2
y1
2
y ,CfeE 2
p1
2
φ
p.ypb1φy CCρfeeE
c 2014 NSP Natural Sciences Publishing Cor.
J. Stat. Appl. Pro. 3, No. 3, 269-293 (2014) / www.naturalspublishing.com/Journals.asp 271
Where,
,N
1
n
1f1
,Y
SC
2
2y2
y
,P
SC
2
2p2
p
and
SS
S
y
ypb is the point bi-serial correlation coefficient.
Here,
N
1i
2i
2y Yy
1N
1S
,
N
1i
2i
2 P1N
1S
and
N
1iiiy YNPy
1N
1S ,
N
1
n'
1f, 2
In order to have an estimate of the study variable y, assuming the knowledge of the
population proportion P, [1] proposed the following estimator
p
Pyt NGR
(2)
P
pyt NGP
(3)
Following [1] , we propose the following estimator
α
321
321S1
KKpK
KKPKyt
(4)
The Bias and MSE expressions of the estimator S1t up to the first order of approximation
are, respectively, given by
c 2014 NSP Natural Sciences Publishing Cor.
272 R. Singh, S.B. Gupta and S. Malik: An almost unbiased population parameter(s)
p1
2
12
p1S1 KαV2
1)Vα(αCfYtB
(5)
1p1
2
1
22
p
2
y1
2
S1 αK2VVαCCfYtMSE (6)
Also following [2], we propose the following estimator
5454
5454
β
S2KpKKPK
KpKKPKλexp
P
p2yt
(7)
The Bias and MSE expressions of the estimator S2t up to the first order of approximation are,
respectively, given by
2
KλVβK
8
V2λλ
2
1ββ
2
βλVCfYtB
p2
p
2
222
p1S2 (8)
2
λVβC2KβλV
4
VλβCCfYtMSE 22
pP2
2
2
222
p
2
y1
2
S2 (9)
, and are suitable chosen constants. Also1K ,
3K ,4K ,
5K are either real numbers or
function of known parameters of the auxiliary attributes such as pC , 2, pb and
PK .
2K is an integer which takes values +1 and -1 for designing the estimators and
54
42
321
11
KPK
PKV
KKPK
PKV
We see that the estimator’s tS1 and tS2 are biased estimators. In some applications bias is
disadvantageous. Following these estimators we have proposed almost unbiased estimator of
Y .
c 2014 NSP Natural Sciences Publishing Cor.
J. Stat. Appl. Pro. 3, No. 2, 269-293 (2014) / www.naturalspublishing.com/Journals.asp 273
2. Almost unbiased estimator
Suppose, ytS0
α
321
321S1
KKpK
KKPKyt,
5454
5454
β
S2KpKKPK
KpKKPKλexp
P
p2yt,
Such thatS0t ,
S1t , WtS2 , where W denotes the set of all possible estimators for estimating
the population mean Y . By definition, the set W is a linear variety if
Wtwt2
0i
Siip (10)
Such that, 1w2
0i
i
and Rw i (11)
where, 0,1,2iwi denotes the constants used for reducing the bias in the class of
estimators, W denotes the set of those estimators that can be constructed from 0,1,2itSi
and R denotes the set of real numbers.
Expressing equation (10) in terms of e’s, we have
eeVeV
2
eV1we1Yt y11
22
1
1yp
2
eeVee
8
eV2
2
eV
2
eV
2
e1ew
y2
y
22
2
2
22
2
2 (12)
Subtracting Y from both sides of equation (12) and then taking expectation of both sides, we
get the bias of the estimator pt up to the first order of approximation, as
c 2014 NSP Natural Sciences Publishing Cor.
274 R. Singh, S.B. Gupta and S. Malik: An almost unbiased population parameter(s)
8
V2
2
1
2
VCwfYKV
2
V1CwfY)t(B
2
222
p21p1
2
12
p11p
2
KVK P2
P (13)
From (12), we have
2
eλVβeweαVweY)Y(t
φ2
φ2φ11yp
(14)
Squaring both sides of (14) and then taking expectation, we get the MSE of the estimator pt
up to the first order of approximation, as
p
22
p
2
y1
2
p QK2QCCfYtMSE
(15)
Which is minimum when
pKQ
(16)
where,
2
VwVwQ 2
211 (17)
Putting the value of pKQ in (15), we have optimum value of estimator as pt (optimum).
Thus the minimum MSE of pt is given by
2
pb
2
y1
2
p 1CfYtMSE.min
(18)
Which is same as that of traditional linear regression estimator.
c 2014 NSP Natural Sciences Publishing Cor.
J. Stat. Appl. Pro. 3, No. 2, 269-293 (2014) / www.naturalspublishing.com/Journals.asp 275
from (11) and (17), we have only two equations in three unknowns. It is not possible to find
the unique values for ,s'w ii=0, 1, 2. In order to get unique values of ,s'w i
we shall impose
the linear restriction
2
0i
Sii 0tBw (19)
where, SitB denotes the bias in the ith estimator.
Equations (11), (17) and (19) can be written in the matrix form as
0
k
1
w
w
w
tBtB02
λVβαV0
111
p
2
1
0
S2S1
21 (20)
Using (20), we get the unique values of ,s'w ii=0, 1, 2 as
1121
1P2
2
11211
1P11
11211
1P11121121P1112110
XAAV
AKw
XXAAVV
AKXw
XAAVV
AKVXAAVVXAKXXAAVVw
1
p
2
211
P2P
2
222
p1
2
11
V
KX
2
VAX
2
KVK
8
V2
2
1
2
VA
KV2
V1A
Use of these ,s'w ii=0, 1, 2 remove the bias up to terms of order 1no at (10).
c 2014 NSP Natural Sciences Publishing Cor.
276 R. Singh, S.B. Gupta and S. Malik: An almost unbiased population parameter(s) 3.Empirical study
For empirical study we use the data sets earlier used by [20] (population 1) and [21]
(population 2) to verify the theoretical results.
Data statistics:
Population N n Y P yC pC pb 2
Population I 89 20 3.360 0.1236 0.60400 2.19012 0.766 6.2381
Population II 25 10 9.44 0.400 0.17028 1.27478 -0.387 4.3275
Table 3.1: Values of ,s'w i
c 2014 NSP Natural Sciences Publishing Cor.
,s'w i Population 1 Population 2
0w
-3.95624
1.124182
1w
5.356173
0.020794
2w
-0.39993
-0.14498
J. Stat. Appl. Pro. 3, No. 2, 269-293 (2014) / www.naturalspublishing.com/Journals.asp 277
Table 3.2: PRE of different estimators of Y with respect to y
Choice of scalars
0w
1w
2w
1K
2K
3K
4K
5K
Estimator
PRE
(POPI)
PRE
(POPII)
1
0
0
y
100
100
0
1
0
1
1
0
1
NGRt
11.63
1.59
1
1
0
-1
NGPt
5.075
1.94
0
0
1
1
0
)0,1(1t
12.88
1.59
-1
0
)0,1(1t
5.43
1.95
1
0
1
1
)1,1(2t
73.59
0.84
1
0
1
-1
)1,1(2t
4.94
8.25
1
0
0
1
)1,0(2t
14.95
8.25
1
0
0
-1
)1,0(2t
73.48
5.58
0w
1w
2w
1
1
1
1
1
1
1
1
Pt optimum
241.98
117.61
c 2014 NSP Natural Sciences Publishing Cor.
278 R. Singh, S.B. Gupta and S. Malik: An almost unbiased population parameter(s)
4. Proposed estimators in two phase sampling
In some practical situations when P is not known a priori, the technique of two-phase
sampling is used. Let p'denote the proportion of units possessing attribute in the first phase
sample of size n' ; p denote the proportion of units possessing attribute in the second phase
sample of size nn' and y denote the mean of the study variable y in the second phase
sample.
In two-phase sampling the estimator pt will take the following form
Htht2
0i
idipd
(21)
Such that, 1h2
0i
i
and Rh i (22)
Where,
yt0d
m
321
3211d
KKpK
KKp'Kyt,
5454
5454
q
2dKpKKp'K
KpKKp'Kγexp
p'
p2yt,
The Bias and MSE expressions of the estimator 1dt and
2dt up to the first order of
approximation are, respectively, given by
2
pp31
2
p2
2
1
2
2
p1
2
1
2
p2
2
1
1d CkfmRCfRm2
Cf1)Rm(m
2
Cf1)Rm(mYtB
(23)
2
p3p1
2
p3
2
1
22
y1
2
1d Cfk2mRCfRmCfYtMSE (24)
c 2014 NSP Natural Sciences Publishing Cor.
J. Stat. Appl. Pro. 3, No. 2, 269-293 (2014) / www.naturalspublishing.com/Journals.asp 279
2
p23
2
pp23
2
p2
22
pp2
2
p2
2
p1
2d qCγRfCkγRfCfqCkqf2
C1)fq(q
2
C1)fq(qYtB
(25)
2
p3
2
1
2
y1
2
2d CfLCfYtMSE (26)
Where,
21
54
42
321
11
γAqL
KPK2
PKR
KKPK
PKR
(27)
Expressing (21) in terms of e’s, we have
φy1
2
φ
2
1
φy1φy1φ1
2
φ
2
1
1ypd e'emR2
e'R1mmeemReemRemR
2
eR1mmwe1Yt
φyφyφy2φφ2
2
φ
φφ
2
φ
2
φ
φ2 eqee'eeeγRee'γR2
e'1qqe'eqqe'
2
e1qqqew
Subtracting Y from both sides of the above equation and then taking expectation of both
sides, we get the bias of the estimator pdt up to the first order of approximation, as
2d1dpd tBtBY)B(t (28)
Also,
φ2φ2φφ2φ1φ110pd eγRe'γRqe'qewemRe'mRweY)Y(t
(29)
c 2014 NSP Natural Sciences Publishing Cor
280 R. Singh, S.B. Gupta and S. Malik: An almost unbiased population parameter(s) Squaring both sides of (29) and then taking expectation, we get the MSE of the estimator tpd
up to the first order of approximation, as
2
pp32
2
p3
2
2
2
y1
2
pd Ckf2LCfLCfYtMSE (30)
Which is minimum when
p2 KL (31)
Where 22112 γRqhmRhL (32)
Putting the value of p2 KL in (30), we have optimum value of estimator as pdt (optimum).
Thus the minimum MSE of pdt is given by
2
pb31
2
y
2
pd ρffCYtmin.MSE
(33)
which is same as that of traditional linear regression estimator.
from (22) and (32), we have only two equations in three unknowns. It is not possible to find
the unique values for s,'h i 1=0, 1, 2. In order to get unique values of s,'h i
we shall impose
the linear restriction
2
0i
idi 0tBh (34)
where, idtB denotes the bias in the ith estimator.
Equations (22), (32) and (34) can be written in the matrix form as
0
k
1
h
h
h
tBtB0
γRqmR0
111
p
2
1
0
2d1d
21
(35)
c 2014 NSP Natural Sciences Publishing Cor.
J. Stat. Appl. Pro. 3, No. 2, 269-293 (2014) / www.naturalspublishing.com/Journals.asp 281
Using (35), we get the unique values of s,'hi i=0, 1, 2 as
21211
1P2
21211
2P1
1
p
1
210
γRNNmRqN
NKh
γRNNmRqN
γRqKN
mR
kh
hh1h
where,
2
p23
2
pp23
2
p2
22
pp2
2
p2
2
p1
2
2
pp31
2
p2
2
1
2
2
p1
2
1
2
p2
2
1
1
qCγRfCkγRfCfqCkqf2
C1)fq(q
2
C1)fq(qN
CkfmRCfRm2
Cf1)Rm(m
2
Cf1)Rm(mN
Use of these s,'h i i=0, 1, 2 remove the bias up to terms of order 1no at (21).
5. Empirical Study
For empirical study we use the data sets earlier used by [20] (population 1) and [21]
(population 2) to verify the theoretical results.
Data statistics:
Pop. N n Y P p' yC pC pb n'
Pop.I 89 23 1322 0.1304 0.13336 0.69144 2.7005 0.408 45
Pop.II 25 7 7.143 0.294 0.308 0.36442 1.3470 -0.314 13
c 2014 NSP Natural Sciences Publishing Cor.
282 R. Singh, S.B. Gupta and S. Malik: An almost unbiased population parameter(s)
Table 5.1: PRE of different estimators of Y with respect to y
Choice of scalars
0w
1w
2w
1K
2K
3K
4K
5K
m
q
γ
Estimator
PRE
(POPI)
PRE
(POPII)
1
0
0
y
100
100
0
1
0
1
1
0
1
NGRt
11.13
8.85
1
1
0
-1
NGPt
7.48
12.15
0
0
1
1
0
1d(1,0)t
26.84
5.42
-1
0
1,0)1d(t
23.75
5.87
1
0
1
1
2d(1,1)t
82.55
1.23
1
0
1
-1
1)2d(1,t
8.56
8.46
1
0
0
1
2d(0,1)t
22.54
6.57
1
0
0
-1
1)2d(0,t
82.56
7.45
0w
1w
2w
1
1
1
1
1
1
1
1
tpd
optimum
112.55
106.89
c 2014 NSP Natural Sciences Publishing Cor.
J. Stat. Appl. Pro. 3, No. 2, 269-293 (2014) / www.naturalspublishing.com/Journals.asp 283
6. Conclusion
In this paper, we have proposed an unbiased estimator pt and pdt using information on the
auxiliary attribute(s) in case of single phase and double phase sampling respectively.
Expressions for bias and MSE’s of the proposed estimators are derived up to first degree of
approximation. From theoretical discussion and empirical study we conclude that the
proposed estimatorspt and pdt under optimum conditions perform better than other estimators
considered in the article.
c 2014 NSP Natural Sciences Publishing Cor.
284 R. Singh, S.B. Gupta and S. Malik: An almost unbiased population parameter(s) Appendix A.
Some members of the proposed family of estimators -
Some members (ratio-type) of the class 1t
When ,0w0 ,1w1 0w2 ,1
1K 3K Estimators ( 1K2 ) Estimators ( 1K2 PRE’S
1K2
PRE’S
1K2
1 pC
p
p
1a1Cp
CPyt
p
p
1b1Cp
CPyt
134.99 72.50
1 2
)(p
)(Pyt
2
22a1
)(p
)(Pyt
2
22b1
111.62 89.34
2 pC
p2
p2
3a1C)(p
C)(Pyt
p2
p2
3b1C)(p
C)(Pyt
226.28 12.99
pC 2
)(pC
)(PCyt
2p
2p
4a1
)(pC
)(PCyt
2p
2p
4b1
126.66 77.93
1 pb
pb
pb
5a1p
Pyt
pb
pb
5b1p
Pyt
207.46 39.13
NP S
SNPp
SNPyt
2
6a1
SNPp
SNPyt
2
6b1 18.14 6.86
NP F
fNPp
fNPyt
2
7a1
fNPp
fNPyt
2
7b1 13.79 10.85
c 2014 NSP Natural Sciences Publishing Cor.
2
pbK
pb2
pb2
8a1Kp
KPyt
pb2
pb2
8b1Kp
KPyt
24.15 5.40
NP pbK
pb
pb
2
9a1KNPp
KNPyt
pb
pb
2
9b1KNPp
KNPyt
18.62 7.78
N 1
1Np
1NPyt 10a1
1Np
1NPyt 10b1
15.93 9.26
N pC
p
p
11a1CNp
CNPyt
p
p
11b1CNp
CNPyt
19.79 6.86
N pb
pb
pb
12a1Np
NPyt
pb
pb
12b1Np
NPyt
15.18 9.78
N S
SNp
SNPyt 13a1
SNp
SNPyt 13b1
12.34 10.96
N F
fNp
fNPyt 14a1
fNp
fNPyt 14b1
12.99 11.54
N g=1-f
gNp
gNPyt 15a1
gNp
gNPyt 15b1
15.81 9.34
N pbK
pb
pb
16a1KNp
KNPyt
pb
pb
16b1KNp
KNPyt
13.52 11.10
N pb
pb
pb
17a1np
nPyt
pb
pb
17b1np
nPyt
25.13 4.86
N S
pb
pb
18a1np
nPyt
pb
pb
18b1np
nPyt
14.98 8.81
N F
fnp
fnPyt 19a1
fnp
fnPyt 19b1
13.38 11.20
c 2014 NSP Natural Sciences Publishing Cor.
J. Stat. Appl. Pro. 3, No. 2, 269-293 (2014) / www.naturalspublishing.com/Journals.asp 285
286 R. Singh, S.B. Gupta and S. Malik: An almost unbiased population parameter(s)
N g=1-f
gnp
gnPyt 20a1
gnp
gnPyt 20b1
29.13 3.68
N pbK
pb
pb
21a1Knp
KnPyt
pb
pb
21b1Knp
KnPyt
15.87 9.39
2 P
Pp
PPyt
2
222a1
Pp
PPyt
2
222b1
16.80 7.63
NP P
PNPp
PNPyt
2
23a1
PNPp
PNPyt
2
23b1 15.93 9.26
N P
PNp
PNPyt 24a1
PNp
PNPyt 24b1
13.23 11.32
N P
Pnp
PnPyt 25a1
Pnp
PnPyt 25b1
14.51 10.28
Appendix B.
Some members (product-type) of the class 1t
When ,0w0 ,1w1 0w2
,1
1K 3K Estimators ( 1K2 ) Estimators ( 1K2 ) PRE’S
1K2
PRE’S
1K2
1 pC
p
p
1c1CP
Cpyt
p
p
1d1CP
Cpyt
35.54 9.93
c 2014 NSP Natural Sciences Publishing Cor.
1 2
)(P
)(pyt
2
22c1
)(P
)(pyt
2
22d1
110.12 101.54
2 pC
p2
p2
3c1C)(P
C)(pyt
p2
p2
3d1C)(P
C)(pyt
6.09 0.127
pC 2
)(PC
)(pCyt
2p
2p
4c1
)(PC
)(pCyt
2p
2p
4d1
99.38 82.52
1 pb
pb
pb
5c1P
pyt
pb
pb
5d1P
pyt
0.00135 5.42
NP S
SNP
SNPpyt
26c1
SNP
SNPpyt
26d1 2.53 1.23
NP f
fNP
fNPpyt
27c1
fNP
fNPpyt
27d1 2.03 1.52
2 pbK
pb2
pb2
8c1KP
Kpyt
pb2
pb2
8d1Kp
KPyt
1.83 1.68
NP pbK
pb
2
pb
9c1KNP
KNPpyt
pb
2
pb
9d1KNP
KNPpyt
1.89 1.63
N 1
1NP
1Npyt 10c1
1NP
1Npyt 10d1
2.37 1.30
1 2
)(P
)(pyt
2
22c1
)(P
)(pyt
2
22d1
110.12 101.54
2 pC
p2
p2
3c1C)(P
C)(pyt
p2
p2
3d1C)(P
C)(pyt
6.09 0.127
pC 2
)(PC
)(pCyt
2p
2p
4c1
)(PC
)(pCyt
2p
2p
4d1
99.38 82.52
c 2014 NSP Natural Sciences Publishing Cor.
J. Stat. Appl. Pro. 3, No. 2, 269-293 (2014) / www.naturalspublishing.com/Journals.asp 287
288 R. Singh, S.B. Gupta and S. Malik: An almost unbiased population parameter(s)
N pC
p
p
11c1CNP
CNpyt
p
p
11d1CNP
CNpyt
2.50 1.23
N pb
pb
pb
12c1NP
Npyt
pb
pb
12d1NP
Npyt
1.79 1.70
N S
SNP
SNpyt 13c1
SNP
SNpyt 13d1
2.16 1.44
N f
fNP
fNpyt 14c1
fNP
fNpyt 14d1
1.98 1.56
N g=1-f
gNP
gNpyt 15c1
gNP
gNpyt 15d1
2.34 1.32
N pbK
pb
pb
16c1KNP
KNpyt
pb
pb
16d1KNP
KNpyt
1.93 1.60
N pb
pb
pb
17c1nP
npyt
pb
pb
17d1nP
npyt
1.49 1.96
N S
pb
pb
18c1nP
npyt
pb
pb
18d1nP
npyt
2.65 1.14
N f
fnP
fnpyt 19c1
fnP
fnpyt 19d1
2.06 1.51
N g=1-f
gnP
gnpyt 20c1
gnP
gnpyt 20d1
3.29 0.84
N pbK
pb
pb
21c1KnP
Knpyt
pb
pb
21d1KnP
Knpyt
1.88 1.63
c 2014 NSP Natural Sciences Publishing Cor.
J. Stat. Appl. Pro. 3, No. 2, 269-293 (2014) / www.naturalspublishing.com/Journals.asp 289
Appendix C.
Some members (product-type) of the class 2t
When ,0w0 ,0w1 1w2
4K 5K Estimators
( 1,1 )
PRE’S
1 pC
p
21C2Pp
Ppexp
P
p2yt
12.42
1 2
2
222Pp
Ppexp
P
p2yt
11.92
2 pC
p2
223
C2Pp
Ppexp
P
p2yt
16.29
pC 2
2p
p
242PpC
PpCexp
P
p2yt
12.53
c 2014 NSP Natural Sciences Publishing Cor.
2 P
PP
Ppyt
2
222c1
PP
Ppyt
2
222d1
2.99 0.97
NP P
PNP
PNPpyt
223c1
PNP
PNPpyt
223d1 2.37 1.30
N P
PNP
PNpyt 24c1
PNP
PNpyt 24d1
2.11 1.47
N P
PnP
Pnpyt 25c1
PnP
Pnpyt 25d1
2.49 1.23
290 R. Singh, S.B. Gupta and S. Malik: An almost unbiased population parameter(s)
1 pb
pb
252Pp
Ppexp
P
p2yt
13.86
NP S
S2PpNP
PpNPexp
P
p2yt 26
44.46
NP F
f2PpNP
PpNPexp
P
p2yt 27
61.84
2 pbK
pb2
228
K2Pp
Ppexp
P
p2yt
40.17
NP pbK
pb
29K2PpNP
PpNPexp
P
p2yt
48.09
N 1
2PpN
PpNexp
P
p2yt 210
54.10
N pC
p
211C2PpN
PpNexp
P
p2yt
44.84
N pb
pb
2122PpN
PpNexp
P
p2yt
56.48
N S
S2PpN
PpNexp
P
p2yt 213
62.40
N F
f2PpN
PpNexp
P
p2yt 214
65.67
c 2014 NSP Natural Sciences Publishing Cor.
c 2014 NSP Natural Sciences Publishing Cor.
J. Stat. Appl. Pro. 3, No. 2, 269-293 (2014) / www.naturalspublishing.com/Journals.asp 291
N g=1-f
g2PpN
PpNexp
P
p2yt 215
54.47
N pbK
pb
216K2PpN
PpNexp
P
p2yt
63.21
N pb
pb
2172Ppn
Ppnexp
P
p2yt
38.59
N S
S2Ppn
Ppnexp
P
p2yt 218
52.19
N F
f2Ppn
Ppnexp
P
p2yt 219
63.74
N g=1-f
g2Ppn
Ppnexp
P
p2yt 220
35.33
N pbK
pb
221K2Ppn
Ppnexp
P
p2yt
54.68
2 P
P2Pp
Ppexp
P
p2yt
2
2222
47.53
NP P
P2PpNP
PpNPexp
P
p2yt 223
54.10
N P
P2PpN
PpNexp
P
p2yt 224
64.43
N P
P2Ppn
Ppnexp
P
p2yt 225
58.91
292 R. Singh, S.B. Gupta and S. Malik: An almost unbiased population parameter(s)
In addition to above estimators a large number of estimators can also be generated from the
proposed estimators just by putting different values of constants s'w i , s'hi ,K1 ,K2 ,K3,K4
,K5, and .
References
[1] Naik, V.D., Gupta, P.C. (1996): A note on estimation of mean with known population
proportion of an auxiliary character. Journal of the Indian Society of
Agricultural Statistics 48(2) 151-158.
[2] Singh, H. P., Solanki, R.S. (2012): Improved estimation of population mean in simple
random sampling using information on auxiliary attribute.Applied Mathematics
and Computation 218 (2012) 7798–7812.
[3] Sahai A., S. K. Ray S. K. (1980) “An efficient estimator using auxiliary information,”
Metrika, vol. 27, no. 4, pp. 271–275.
[4] Singh, H.P. and Tailor, R. (2003): Use of known correlation coefficient in estimating the
finite population mean. Statistics in Transition, 6,4,555-560.
[5] Kadilar,C. and Cingi,H. (2003): Ratio Estimators in Straitified Random Sampling.
Biometrical Journal 45 (2003) 2, 218-225.
[6] Shabbir, J. and Gupta, S. (2007): A new estimator of population mean in stratified
sampling, Commun. Stat. Theo. Meth.35: 1201–1209.
[7] Gupta, S. and Shabbir, J. (2008) : On improvement in estimating the population mean in
simple random sampling. Jour.Of Applied Statistics, 35, 5, 559-566.
[8] Khoshnevisan, M. Singh, R., Chauhan, P., Sawan, N. and Smarandache, F. (2007): A
general family of estimators for estimating population mean using known value
of some population parameter(s). Far east journal of theoretical statistics, 22(2),
181-191.
[9] Singh, R.,Cauhan, P., Sawan, N. and Smarandache,F. (2007): Auxiliary information and
a priory values in construction of improved estimators. Renaissance High press.
c 2014 NSP Natural Sciences Publishing Cor.
J. Stat. Appl. Pro. 3, No. 2, 269-293 (2014) / www.naturalspublishing.com/Journals.asp 293
[10] Singh, R., Kumar, M. and Smarandache, F. (2008): Almost Unbiased Estimator for Estimating
Population Mean Using Known Value of Some Population Parameter(s). Pak. J. Stat.
Oper. Res., 4(2) pp63-76.
[11] Koyuncu, N. and Kadilar, C. (2009) : Family of estimators of population mean using two
auxiliary variables in stratified random sampling. Comm. In Stat. – Theory and Meth.,
38:14, 2398-2417.
[12] Diana G., Giordan M. Perri P.(2011): An improved class of estimators for the population mean.
Stat Methods Appl (2011) 20:123–140
[13] Upadhyaya LN, Singh HP, Chatterjee S, Yadav R (2011) Improved ratio and product exponential
type estimators. J Stat Theory Pract 5(2):285–302
[14] Jhajj, H. S., Sharma, M.K. and Grover, L.K. (2006): A family of estimators of population mean
using information on auxiliary attribute. Pakistan journal of statistics, 22(1), 43-50
(2006).
[15] Abd-Elfattah, A.M. El-Sherpieny, E.A. Mohamed, S.M. Abdou, O. F. (2010): Improvement in
estimating the population mean in simple random sampling using information on
auxiliary attribute. Appl. Mathe. and Compt. doi:10.1016/j.amc.2009.12.041.
[16] Grover L.,Kaur P.(2011): An improved exponential estimator of finite population mean in simple
random sampling using an auxiliary attribute.Applied Mathematics and Computation 218
(2011) 3093–3099.
[17] Malik, S. and Singh, R. (2013a) : An improved estimator using two auxiliary attributes. Appli.
Math. Compt., 219, 10983-10986.
[18] Malik, S. and Singh, R. (2013b) : A family of estimators of population mean using information
on point bi-serial and phi correlation coefficient. Int. Jour. Stat. Econ. 10(1), 75-89.
[19] Malik, S. and Singh, R. (2013c) : Dual to ratio cum product estimators of finite population mean
using auxiliary attribute(s) in stratified random sampling. WASJ, 28(9), 1193-1198
[20] Sukhatme, P.V. and Sukhatme, B.V.,(1970) : Sampling theory of surveys with applications. Iowa
State University Press, Ames, U.S.A.
[21] Mukhopadhyaya, P.(2000): Theory and methods of survey sampling. Prentice Hall of India, New
Delhi, India.
c 2014 NSP Natural Sciences Publishing Cor.