Research Article Estimation of Population Mean in Chain...
Transcript of Research Article Estimation of Population Mean in Chain...
Research ArticleEstimation of Population Mean in Chain Ratio-TypeEstimator under Systematic Sampling
Mursala Khan1 and Rajesh Singh2
1Department of Mathematics, COMSATS Institute of Information Technology, Abbottabad 22060, Pakistan2Department of Statistics, Banaras Hindu University, Varanasi 221005, India
Correspondence should be addressed to Mursala Khan; [email protected]
Received 7 September 2015; Accepted 11 October 2015
Academic Editor: Shein-chung Chow
Copyright Β© 2015 M. Khan and R. Singh. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.
A chain ratio-type estimator is proposed for the estimation of finite population mean under systematic sampling scheme usingtwo auxiliary variables. The mean square error of the proposed estimator is derived up to the first order of approximation and iscompared with other relevant existing estimators. To illustrate the performances of the different estimators in comparison with theusual simple estimator, we have taken a real data set from the literature of survey sampling.
1. Introduction and Literature Review
Incorporating the knowledge of the auxiliary variables is veryimportant for the construction of efficient estimators for theestimation of population parameters and increasing the effi-ciency of the estimators in different sampling design. Usingthe knowledge of the auxiliary variables, several authorshave proposed different estimation technique for the finitepopulation mean of the study variable; Cochran [1], Tripathy[2], Kadilar and Cingi [3, 4], Singh et al. [5], Khan andArunachalam [6], Lone and Tailor [7], Khan [8], Khan andHussain [9], and Khan et al. [10] have worked on the estima-tion of population parameters using auxiliary information.
In the present paper, we will work on the estimationof population mean using the knowledge of the auxiliaryvariables under systematic sampling. Various statisticianshave worked on the estimation of population mean insystematic sampling: Cochran [11], Hansen et al. [12], Robson[13], Swain [14], Singh [15], Shukla [16], Kushwaha and Singh[17], Banarasi et al. [18], R. Singh andH. P. Singh [19], Singh etal. [20], Singh and Solanki [21], Singh and Jatwa [22], Singhet al. [23, 24], Tailor et al. [25], Verma and Singh [26], andVerma et al. [27], and so forth.
Consider a finite population π = {π1, π2, . . . , π
π} of size
π units, numbered from 1 to π in some order. A sample of
size π units is taken at random from the first π units and everyπth subsequent unit; then,π = ππ where π and π are positiveintegers; thus, there will be π samples (clusters) each of sizeπ and observe the study variate π¦ and auxiliary variate π₯ foreach and every unit selected in the sample. Let (π¦
ππ, π₯ππ), for π =
1, 2, . . . , π and π = 1, 2, . . . , π: denote the value of πth unit inthe πth sample.Then, the systematic samplemeans are definedas follows: π¦
π π¦= π‘0= (1/π)β
π
π=1π¦ππand π₯
π π¦= (1/π)β
π
π=1π₯ππ
are the unbiased estimators of the population means π =
(1/π)βπ
π=1π¦ππand π = (1/π)β
π
π=1π₯ππof π¦ and π₯, respec-
tively. Let π2π¦= (1/(πβ 1))β
π
π=1βπ
π=1(π¦ππβπ)2, π2π₯= (1/(πβ
1))βπ
π=1βπ
π=1(π₯ππβ π)2, and π
2
π§= (1/(π β 1))β
π
π=1βπ
π=1(π§ππβ
π)2 be the population variances of the study variable and
the auxiliary variables, respectively, with the correspondingpopulation covarianceβs π
π¦π₯= (1/(π β 1))β
π
π=1βπ
π=1(π₯ππ
β
π)(π¦ππ
β π), ππ¦π§
= (1/(π β 1))βπ
π=1βπ
π=1(π¦ππ
β π)(π§ππ
β π),and ππ₯π§
= (1/(π β 1))βπ
π=1βπ
π=1(π₯ππβ π)(π§
ππβ π) among the
three variables π¦, π₯, and π§, respectively. Also πΆ2
π¦, πΆ2π₯, and πΆ
2
π§
are the knownpopulation coefficients of variation of the studyvariable and the auxiliary variables, respectively.
To obtain the properties of the estimators up to firstorder of approximation, we use the following errors terms:
Hindawi Publishing CorporationJournal of Probability and StatisticsVolume 2015, Article ID 248374, 5 pageshttp://dx.doi.org/10.1155/2015/248374
2 Journal of Probability and Statistics
π0= (π¦π π¦π
β π)/π, π1= (π₯π π¦π
β π)/π, and π2= (π§π π¦π
β π)/π,such that πΈ(π
π) = 0, for π = 0, 1, and 2.
The first order of approximation of the above errors termsis given by
πΈ (π2
0) = ππ
β
π¦πΆ2
π¦,
πΈ (π2
1) = ππ
β
π₯πΆ2
π₯,
πΈ (π2
2) = ππ
β
π§πΆ2
π§,
πΈ (π0π1) = πππΆ
2
π₯βπβπ¦πβπ₯,
πΈ (π0π2) = ππ
βπΆ2
π§βπβπ¦πβπ§,
πΈ (π1π2) = ππ
ββπΆ2
π§βπβπ₯πβπ§,
(1)
where
π = (π β 1
ππ) ,
ππ¦π₯
=
ππ¦π₯
ππ¦ππ₯
,
ππ¦π§
=
ππ¦π§
ππ¦ππ§
,
ππ₯π§
=ππ₯π§
ππ₯ππ§
,
π =
ππ¦π₯
πΆπ¦
πΆπ₯
,
πβ=
ππ¦π§
πΆπ¦
πΆπ§
,
πβ
π¦= {1 + (π β 1) π
π¦} ,
πβ
π₯= {1 + (π β 1) π
π₯} ,
πβ
π§= {1 + (π β 1) π
π§} ,
πββ
=
πβ
π¦
πβπ₯
,
πββ
2=
πβ
π¦
πβπ§
,
πββ
1=
πβ
π₯
πβπ§
,
(2)
where ππ¦, ππ₯, and π
π§are the intraclass correlation among the
pair of units for the variables π¦, π₯, and π§, respectively.The variance of the usual unbiased estimator for popula-
tion mean is
Var (π‘0) = ππ
2
πβ
π¦πΆ2
π¦. (3)
The classical ratio and product estimators for finite popula-tion mean suggested by Swain [14] and Shukla [16] are givenby
π‘1= π¦π π¦
(π
π₯π π¦
) ,
π‘2= π¦π π¦exp(
π§π π¦
π
) .
(4)
The mean square errors of the estimators, to the first order ofapproximation, are given as follows:
MSE (π‘1) = ππ
2
[πβ
π¦πΆ2
π¦+ πβ
π₯πΆ2
π₯(1 β 2πβπββ)] , (5)
MSE (π‘2) = ππ
2
[πβ
π¦πΆ2
π¦+ πβ
π§πΆ2
π§(1 + 2π
ββπββ
2)] . (6)
The usual regression estimator, using single auxiliary variableand its variance, is given as follows:
π‘3= π¦π π¦
+ π (π β π₯π π¦) , (7)
MSE (π‘3) = ππ
β
π¦π2
π¦[1 β π
2
π¦π₯] . (8)
Utilizing the known knowledge of the auxiliary variable,Singh et al. [20] suggested the following ratio and producttype exponential estimators:
π‘4= π¦π π¦exp(
π β π₯π π¦
π + π₯π π¦
) ,
π‘5= π¦π π¦exp(
π₯π π¦
β π
π₯π π¦
+ π) .
(9)
The mean square errors of the estimators up to first order ofapproximation are given by
MSE (π‘4) = ππ
2
[πβ
π¦πΆ2
π¦+
πβ
π₯πΆ2
π₯
4(1 β 4πβπββ)] , (10)
MSE (π‘5) = ππ
2
[πβ
π¦πΆ2
π¦+
πβ
π₯πΆ2
π₯
4(1 + 4πβπββ)] . (11)
After that, Tailor et al. [25] define the following ratio-cum-product estimator for the population mean π:
π‘6= π¦π π¦
(π
π₯π π¦
)(
π§π π¦
π
) . (12)
The mean square error of the estimator π‘6, up to first order of
approximation, is given by
MSE (π‘6) = ππ
2
[πβ
π¦πΆ2
π¦+ πβ
π₯πΆ2
π₯(1 β 2πβπββ)
+ πβ
π§πΆ2
π§(1 β 2π
βββπββ
1) + 2π
βπΆ2
π§βπβπ¦πβπ§] .
(13)
Journal of Probability and Statistics 3
2. Proposed Estimator
In this section, we have proposed the following regression inratio-cum-product type estimator for the unknown popula-tion mean under systematic sampling:
π‘π
= π¦π π¦
(π
π + ππ¦π₯
(π₯π π¦
β π)
)
πΏ1
β (
π + ππ¦π§
(π§π π¦
β π)
π
)
πΏ2
,
(14)
where πΏ1and πΏ2are the unknown constants, whose values are
to be found for the minimummean square error.The mean square error (MSE) of the estimator up to first
order of approximation is
MSE (π‘π) = ππ
2
[πβ
π¦πΆ2
π¦+ πΏ2
1π½2
π¦π₯πβ
π₯πΆ2
π₯+ πΏ2
2π½2
π¦π§πβ
π§πΆ2
π§
β 2πΏ1π½π¦π₯
ππΆ2
π₯βπβπ¦πβπ₯+ 2πΏ2π½π¦π§
πβπΆ2
π§βπβπ¦πβπ§
β 2πΏ1πΏ2π½π¦π₯
π½π¦π§
πββ
πΆ2
π§βπβπ₯πβπ§] .
(15)
On differentiating (15), with respect to πΏ1and πΏ
2, we obtain
the minimummean squared error of the estimator π‘π, which
is given by
MSE (π‘π)
= ππ2
πβ
π¦[
[
πΆ2
π¦β π2β
π₯πΆ2
π§β
(πβπββ
πΆ2
π§β ππΆ2
π₯)2
(πΆ2π₯β π2ββπΆ2
π§)
]
]
,
(16)
where the optimum values are πΏ1
= βπβπ¦(ππΆ2
π₯β πβπββ
πΆ2
π§)/
π½π¦π₯βπβπ₯(πΆ2
π₯β π2ββ
πΆ2
π§) and πΏ
2= (βπβ
π¦/π½π¦π§βπβπ§){πββ
(ππΆ2
π₯β
πβπββ
πΆ2
π§)/(πΆ2
π₯β π2ββ
πΆ2
π§) β πβ}.
3. Comparison
In this section, we have compared the MSE of the proposedestimator with the MSEs of simple estimator, Swain [14]estimator, Shukla [16] estimator, Singh et al. [20] estimators,and Tailor et al. [25] estimator and found some theoreticalconditions under which the proposed estimator will alwaysperform better:
(i) By (16) and (3), MSE(π‘π) β€ MSE(π‘
0) if
[
[
π2β
π₯πΆ2
π§+
(πβπββ
πΆ2
π§β ππΆ2
π₯)2
(πΆ2π₯β π2ββπΆ2
π§)
]
]
β₯ 0. (17)
(ii) By (16) and (5), MSE(π‘π) β€ MSE(π‘
1) if
[
[
πβ
π¦
{
{
{
π2β
π₯πΆ2
π§+
(πβπββ
πΆ2
π§β ππΆ2
π₯)2
(πΆ2π₯β π2ββπΆ2
π§)
}
}
}
+ πβ
π₯πΆ2
π₯(1 β 2πβπββ)]
]
β₯ 0.
(18)
(iii) By (16) and (6), MSE(π‘π) β€ MSE(π‘
2) if
[
[
πβ
π¦
{
{
{
π2β
π₯πΆ2
π§+
(πβπββ
πΆ2
π§β ππΆ2
π₯)2
(πΆ2π₯β π2ββπΆ2
π§)
}
}
}
+ πβ
π§πΆ2
π§(1 + 2π
ββπββ
2)]
]
β₯ 0.
(19)
(iv) By (16) and (8), MSE(π‘π) β€ MSE(π‘
3) if
[
[
π2β
π₯πΆ2
π§β πΆ2
π¦π2
π¦π₯+
(πβπββ
πΆ2
π§β ππΆ2
π₯)2
(πΆ2π₯β π2ββπΆ2
π§)
]
]
β₯ 0. (20)
(v) By (16) and (10), MSE(π‘π) β€ MSE(π‘
4) if
[
[
πβ
π¦
{
{
{
π2β
π₯πΆ2
π§+
(πβπββ
πΆ2
π§β ππΆ2
π₯)2
(πΆ2π₯β π2ββπΆ2
π§)
}
}
}
+πβ
π₯πΆ2
π₯
4(1 β 4πβπββ)]
]
β₯ 0.
(21)
(vi) By (16) and (11), MSE(π‘π) β€ MSE(π‘
5) if
[
[
πβ
π₯πΆ2
π₯
4(1 + 4πβπββ)
+ πβ
π¦
{
{
{
π2β
π₯πΆ2
π§+
(πβπββ
πΆ2
π§β ππΆ2
π₯)2
(πΆ2π₯β π2ββπΆ2
π§)
}
}
}
]
]
β₯ 0.
(22)
(vii) By (16) and (13), MSE(π‘π) β€ MSE(π‘
6) if
[
[
πβ
π₯πΆ2
π₯(1 β 2πβπββ) + π
β
π§πΆ2
π§(1 β 2π
βββπββ
1)
+ 2πβπΆ2
π§βπβπ¦πβπ§
+ πβ
π¦
{
{
{
π2β
π₯πΆ2
π§+
(πβπββ
πΆ2
π§β ππΆ2
π₯)2
(πΆ2π₯β π2ββπΆ2
π§)
}
}
}
]
]
β₯ 0.
(23)
4 Journal of Probability and Statistics
Table 1: The percent relative efficiency of different estimators withrespect to π‘
0.
PopulationEstimator MSE(π‘
πΌ) PRE(π‘
πΌ, π‘0)
π‘0
1455.08 100.00π‘1
373.32 389.62π‘2
768.06 189.45π‘3
45.52 3196.57π‘4
820.09 177.43π‘5
1044.42 139.32π‘6
187.08 777.79π‘π
22.74 6400.00
4. Numerical Comparison
For comparing the theoretical efficiency conditions of thedifferent estimators numerically, we have used the followingreal data set.
Population 1 (source: Tailor et al. [25]). Consider
π = 15,
π = 3,
π = 44.47,
π = 80,
π = 48.40,
πΆπ¦= 0.56,
πΆπ₯= 0.28,
πΆπ§= 0.43,
π2
π¦= 2000,
π2
π₯= 149.55,
π2
π§= 427.83,
ππ¦π₯
= 538.57,
ππ¦π§
= β902.86,
ππ₯π§
= β241.06,
ππ¦π₯
= 0.9848,
ππ¦π§
= β0.9760,
ππ₯π§
= β0.9530,
ππ¦= 0.6652,
ππ₯= 0.707,
ππ§= 0.5487.
(24)
For the percent relative efficiencies (PREs) of the estimator,we use the following formula and the results are shown inTable 1:
PRE(π‘πΌ, π‘0) = MSE(π‘
0)/MSE(π‘
πΌ) Γ 100, for πΌ = 0, 1, 2,
3, 4, 5, 6, and π.
5. Conclusion
A chain ratio-type estimator is proposed under double sam-pling scheme using two auxiliary variables, and the propertiesof the proposed estimator are derived up to first order ofapproximations. Both theoretically and empirically, it hasbeen shown that the recommended estimator performedbetter than the other competing estimators in terms of higherpercent relative efficiency. Hence, looking on the dominancenature of the proposed estimator may be suggested for itspractical applications.
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper.
Acknowledgment
The authors are thankful to the anonymous learned refereesfor their valuable suggestions regarding the improvement ofthe paper.
References
[1] W. G. Cochran, Sampling Techniques, John Wiley & Sons, NewYork, NY, USA, 1977.
[2] T. P. Tripathy, βA general class of estimators of population ratio,βSankhya Series C, vol. 42, pp. 45β63, 1980.
[3] C. Kadilar and H. Cingi, βRatio estimators in simple randomsampling,β Applied Mathematics and Computation, vol. 151, no.3, pp. 893β902, 2004.
[4] C. Kadilar andH. Cingi, βImprovement in estimating the popu-lation mean in simple random sampling,β Applied MathematicsLetters, vol. 19, no. 1, pp. 75β79, 2006.
[5] R. Singh, P. Chauhan, N. Sawan, and F. Smarandache,βImproved exponential estimator for population variance usingtwo auxiliary variables,β Italian Journal of Pure and AppliedMathematics, vol. 28, pp. 101β108, 2011.
[6] M. Khan and A. Arunachalam, βEstimation of populationvariance using the knowledge of kurtosis of an auxiliaryvariable under simple random sampling,β International Journalof Applied Science andMathematics, vol. 1, no. 2, pp. 63β67, 2014.
[7] H. A. Lone and R. Tailor, βDual to separate product typeexponential estimator in sample surveys,β Journal of StatisticsApplications & Probability Letters, vol. 2, no. 2, pp. 89β96, 2015.
[8] M. Khan, βImprovement in estimating the finite populationmean under maximum and minimum values in double sam-pling scheme,β Journal of Statistics Applications & ProbabilityLetters, vol. 2, no. 2, pp. 1β7, 2015.
[9] M. Khan and S. Hussain, βAn improved class of ratio-typeestimators for finite population mean under maximum andminimum values,β Science International, vol. 27, no. 2, pp. 1135β1138, 2015.
[10] M. Khan, S. Ullah, A. Y. Al-Hossain, and N. Bashir, βImprovedratio-type estimators using maximum and minimum valuesunder simple random sampling scheme,β Hacettepe Journal ofMathematics and Statistics, vol. 44, no. 4, pp. 923β931, 2015.
Journal of Probability and Statistics 5
[11] W. G. Cochran, βThe estimation of the yields of cereal experi-ments by sampling for the ratio of grain to total produce,β TheJournal of Agricultural Science, vol. 30, no. 2, pp. 262β275, 1940.
[12] M. H. Hansen, W. N. Hurwitz, and M. Gurney, βProblems andmethods of the sample survey of business,β The Journal of theAmerican Statistical Association, vol. 41, no. 234, pp. 173β189,1946.
[13] D. S. Robson, βApplication of multivariate polykays to thetheory of unbiased ratio-type estimation,β The Journal of theAmerican Statistical Association, vol. 59, pp. 1225β1226, 1957.
[14] A.K. P.C. Swain, βThe use of systematic sampling ratio estimate,βJournal of Indian Statistical Association, vol. 2, pp. 160β164, 1964.
[15] M. P. Singh, βRatio cum product method of estimation,βMetrika, vol. 12, no. 1, pp. 34β42, 1967.
[16] N. D. Shukla, βSystematic sampling and product method ofestimation,β in Proceeding of All India Seminar on Demographyand Statistics, BHU, Varanasi, India, 1971.
[17] K. S. Kushwaha and H. P. Singh, βClass of almost unbiased ratioand product estimators in systematic sampling,β Journal of theIndian Society of Agricultural Statistics, vol. 41, no. 2, pp. 193β205, 1989.
[18] Banarasi, S. N. S. Kushwaha, and K. S. Kushwaha, βA class ofratio, product and difference (RPD) estimators in systematicsampling,β Microelectronics Reliability, vol. 33, no. 4, pp. 455β457, 1993.
[19] R. Singh and H. P. Singh, βAlmost unbiased ratio and producttype estimators in systematic sampling,β Questiio, vol. 22, no. 3,pp. 403β416, 1998.
[20] H. P. Singh, R. Tailor, andN. K. Jatwa, βModified ratio and prod-uct estimators for population mean in systematic sampling,βJournal of Modern Applied Statistical Methods, vol. 10, no. 2, pp.424β435, 2011.
[21] H. P. Singh and R. S. Solanki, βAn efficient class of estimators forthe population mean using auxiliary information in systematicsampling,β Journal of Statistical Theory and Practice, vol. 6, no.2, pp. 274β285, 2012.
[22] H. P. Singh and N. K. Jatwa, βA class of exponential typeestimators in systematic sampling,β Economic Quality Control,vol. 27, no. 2, pp. 195β208, 2012.
[23] R. Singh, S. Malik, and V. K. Singh, βAn improved estimator insystematic sampling,β Journal of Scientific Research, vol. 56, pp.177β182, 2012.
[24] R. Singh, S. Malik, M. K. Chaudhary, H. Verma, and A. A. Ade-wara, βA general family of ratio-type estimators in systematicsampling,β Journal of Reliability and Statistical Studies, vol. 5, no.1, pp. 73β82, 2012.
[25] T. Tailor, N. K. Jatwa, and H. P. Singh, βA ratio-cum-productestimator of finite population mean in systematic sampling,βStatistics in Transition, vol. 14, no. 3, pp. 391β398, 2013.
[26] H. K. Verma and R. Singh, βA family of efficient estimator incircular systematic sampling,β Journal of Advanced Computing,vol. 3, no. 2, pp. 56β68, 2014.
[27] H. K. Verma, R. D. Singh, and R. Singh, βSome improved esti-mators in systematic sampling under non-response,β NationalAcademy Science Letters, vol. 37, no. 1, pp. 91β95, 2014.
Submit your manuscripts athttp://www.hindawi.com
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttp://www.hindawi.com
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
CombinatoricsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
International Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com
Volume 2014 Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Stochastic AnalysisInternational Journal of