Submit manuscript...
eISSN: 2378-315X

Biometrics & Biostatistics International Journal

Research Article Volume 9 Issue 1

A new three-parameter size-biased poisson-lindley distribution with properties and applications

Rama Shanker,1 Kamlesh Kumar Shukla2

1Department of Statistics, Assam University, India
2Department of Statistics, College of Science, Eritrea

Correspondence: Rama Shanker, Department of Statistics, Assam University, Silchar, India

Received: September 02, 2019 | Published: February 11, 2020

Citation: Shanker R, Shukla KK. A new three-parameter size-biased poisson-lindley distribution with properties and applications. Biom Biostat Int J. 2020;9(1):1-14. DOI: 10.15406/bbij.2020.09.00294

Download PDF

Abstract

A new three-parameter size-biased Poisson-Lindley distribution which includes several one parameter and two-parameter size-biased distributions including size-biased geometric distribution (SBGD), size-biased negative binomial distribution (SBNBD), size-biased Poisson-Lindley distribution (SBPLD), size-biased Poisson-Shanker distribution (SBPSD), size-biased two-parameter Poisson-Lindley distribution-1 (SBTPPLD-1), size-biased two-parameter Poisson-Lindley distribution-2(SBTPPLD-2), size-biased quasi Poisson-Lindley distribution (SBQPLD) and size-biased new quasi Poisson-Lindley distribution (SBNQPLD) for particular cases of parameters has been proposed. Its various statistical properties based on moments including coefficient of variation, skewness, kurtosis and index of dispersion have been studied. Maximum likelihood estimation has been discussed for estimating the parameters of the distribution. Goodness of fit of the proposed distribution has been discussed.

Keywords: three-parameter Lindley distribution, new three-parameter Poisson-Lindley distribution, size-biased distributions, maximum likelihood estimation, goodness of fit

Abbreviations

SBGD, size-biased geometric distribution; SBNBD, size-biased negative binomial distribution; SBPLD, size-biased Poisson-Lindley distribution; SBPSD, size-biased Poisson-Shanker distribution; SBTPPLD-1, size-biased two-parameter Poisson-Lindley distribution-1; SBTPPLD-2, size-biased two-parameter Poisson-Lindley distribution-2; SBQPLD, size-biased quasi Poisson-Lindley distribution; SBNQPLD, size-biased new quasi Poisson-Lindley distribution; ATPLD, A three- parameter Lindley distribution

Introduction

A three- parameter Lindley distribution (ATPLD) introduced by Shanker et al.,1 is defined by its probability density function (pdf) and cumulative distribution function (cdf)

f(x;θ,α,β)=θ2θα+β(α+βx)eθx;x>0,θ>0,β>0,θα+β>0f(x;θ,α,β)=θ2θα+β(α+βx)eθx;x>0,θ>0,β>0,θα+β>0     (1.1)

F(x;θ,α,β)=1[1+θβxθα+β]eθx;x>0,θ>,β>0,θα+β>0F(x;θ,α,β)=1[1+θβxθα+β]eθx;x>0,θ>,β>0,θα+β>0     (1.2)

It has been observed that ATPLD is a convex combination of exponential and gamma distributions with mixing proportionp=θαθα+βp=θαθα+β . Shanker et al.,1 discussed its statistical properties, estimation of parameters using maximum likelihood estimation and applications to model lifetime data. Further, ATPLD includes several one parameter and two-parameter lifetime distributions for particular values of parameters. The particular distributions of (1.2) are summarized in table 1 along with their pdf and introducers.

Although Lindley distribution was proposed by Lindley,2 but various statistical properties of Lindley distribution was studied by Ghitany et al.3 Statistical properties, estimation of parameters and applications of the particular distributions of ATPLD given in table 1 are available in the respective papers.

Parameter Values

   Probability density function

Name of the distribution

Introducers (years)

α=1,β=0α=1,β=0

f(x;θ)=θeθx;x>0f(x;θ)=θeθx;x>0

Exponential distribution

 

α=β=1α=β=1

f(x;θ)=θ2θ+1(1+x)eθx;x>0f(x;θ)=θ2θ+1(1+x)eθx;x>0

Lindley distribution

Lindley2

α=θ,β=1α=θ,β=1

f(x;θ)=θ2θ2+1(θ+x)eθx;x>0f(x;θ)=θ2θ2+1(θ+x)eθx;x>0

Shanker distribution

Shanker11

β=1β=1

f(x;θ,α)=θ2θα+1(α+x)eθx;x>0f(x;θ,α)=θ2θα+1(α+x)eθx;x>0

Two-parameter Lindley distribution-1 (TPLD-1)

Shanker and Mishra12

α=1α=1

f(x;θ,β)=θ2θ+β(1+βx)eθx;x>0f(x;θ,β)=θ2θ+β(1+βx)eθx;x>0

Two-parameter Lindley distribution-2 (TPLD-2)

Shanker et al.13

β=θβ=θ

f(x;θ,α)=θα+1(α+θx)eθx;x>0f(x;θ,α)=θα+1(α+θx)eθx;x>0

Quasi Lindley distribution (QLD)

   Shanker and Mishra14

α=θα=θ

f(x;θ,β)=θ2θ2+β(θ+βx)eθx;x>0f(x;θ,β)=θ2θ2+β(θ+βx)eθx;x>0

New Quasi Lindley distribution (NQLD)

Shanker and Amanuel15

Table 1 Particular continuous distributions for specific values of parameters of ATPLD with probability density function and its introducers (year)

Recently, Das et al.4 proposed a new three-parameter Poisson-Lindley distribution (NTPPLD) by mixing Poisson distribution with ATPLD introduced by Shanker et al.1 given in (1.1). The probability mass function of NTPPLD proposed by Das et al. 4 is given by

P0(x;θ,α,β)=θ2θα+ββx+(θα+α+β)(θ+1)x+2;x=0,1,2,...,θ>0,α>0,θα+β>0P0(x;θ,α,β)=θ2θα+ββx+(θα+α+β)(θ+1)x+2;x=0,1,2,...,θ>0,α>0,θα+β>0     (1.3)

Statistical properties including moments based measures, generating functions, estimation of parameters and applications of the distribution have been discussed by Das et al.4

It has been observed that NTPPLD includes several one parameter and two-parameter discrete distributions based on Poisson mixture of lifetime distributions given in table 1. The particular discrete distributions of (1.3) for particular values of parametersare summarized in table 2 along with their probability mass function (pmf) and introducers (year).

Parameter Values

Probability mass function

Name of the distribution

Introducers (years)

β=0,α=1β=0,α=1

P(X=x)=θθ+1(1θ+1)x;x=0,1,2,...P(X=x)=θθ+1(1θ+1)x;x=0,1,2,...

Geometric distribution

 

α=0,β=1α=0,β=1

P(X=x)=(x+1)(θθ+1)2(1θ+1)x;x=0,1,2,...P(X=x)=(x+1)(θθ+1)2(1θ+1)x;x=0,1,2,...

Negative Binomial distribution

Greenwood and Yule16

α=β=1α=β=1

P(X=x)=θ2(x+θ+2)(θ+1)x+3;x=0,1,2,...P(X=x)=θ2(x+θ+2)(θ+1)x+3;x=0,1,2,...

Poisson-Lindley distribution (PLD)

Sankaran17

α=θ,β=1α=θ,β=1

P(X=x)=θ2θ2+1x+(θ2+θ+1)(θ+1)x+2;x=0,1,2,...P(X=x)=θ2θ2+1x+(θ2+θ+1)(θ+1)x+2;x=0,1,2,...

Poisson-Shanker distribution (PSD)

Shanker6

α=1α=1

P(X=x)=θ2θ2+ββx+θ2+θ+β(θ+1)x+2;x=0,1,2,...P(X=x)=θ2θ2+ββx+θ2+θ+β(θ+1)x+2;x=0,1,2,...

Two-parameter Poisson-Lindley distribution-1 (TPPLD-1)

Shanker et al.18

β=1β=1

P(X=x)=θ2θα+1x+θα+α+1(θ+1)x+2;x=0,1,2,...P(X=x)=θ2θα+1x+θα+α+1(θ+1)x+2;x=0,1,2,...

Two-parameter Poisson-Lindley distribution-2 (TPPLD-2)

Shanker and Mishra18

β=θβ=θ

P(X=x)=θα+1θx+θα+α+θ(θ+1)x+2;x=0,1,2,...P(X=x)=θα+1θx+θα+α+θ(θ+1)x+2;x=0,1,2,...

Quasi Poisson-Lindley distribution (QPLD)

Shanker and Mishra19

α=θα=θ

P(X=x)=θ2θ2+ββx+θ2+θ+β(θ+1)x+2;x=0,1,2,...P(X=x)=θ2θ2+ββx+θ2+θ+β(θ+1)x+2;x=0,1,2,...

New Quasi Poisson-Lindley distribution (NQPLD)

Shanker and Tekie20

Table 2 Particular discrete distributions for specific values of parameters of NTPPLD with pmf and its introducers (year)

The first four moments about origin and the variance of NTPPLD, obtained by Das et al.,4 are given by 

μ1=θα+2βθ(θα+β)μ1=θα+2βθ(θα+β)

μ2=θ2α+2(α+β)θ+6βθ2(θα+β)μ2=θ2α+2(α+β)θ+6βθ2(θα+β)

μ3=θ3α+6(α+β)θ2+6(α+3β)θ+24βθ3(θα+β)μ3=θ3α+6(α+β)θ2+6(α+3β)θ+24βθ3(θα+β)

μ4=θ4α+2(7α+β)θ3+6(6α+7β)θ2+24(α+6β)θ+120βθ4(θα+β)μ4=θ4α+2(7α+β)θ3+6(6α+7β)θ2+24(α+6β)θ+120βθ4(θα+β)

μ2=σ2=θ3α2+(α+3β)θ2α+2(2α+β)θβ+2β2θ2(θα+β)2μ2=σ2=θ3α2+(α+3β)θ2α+2(2α+β)θβ+2β2θ2(θα+β)2

The main purpose of this paper is to propose a new three-parameter size-biased Poisson-Lindley distribution which includes several one parameter and two-parameter size-biased distributions for particular cases of parameters. Its moments have been derived and various statistical properties based on moments have been studied. Maximum likelihood estimation has been discussed. Goodness of fit of the distribution has been discussed with several count datasets.

A new three-parameter size-biased poisson-lindley distribution

Using the pmf (1.3) and the mean of NTPPLD, a new three-parameter size-biased Poisson-Lindley distribution (NTPSBPLD) can be obtained as

P1(x;θ,α,β)=xP0(x;θ,α,β)μ1=θ3θα+2βx(βx+θα+α+β)(θ+1)x+2P1(x;θ,α,β)=xP0(x;θ,α,β)μ1=θ3θα+2βx(βx+θα+α+β)(θ+1)x+2     (2.1)


;x=1,2,3,..,(θ,α,β)>0;x=1,2,3,..,(θ,α,β)>0

It can be easily verified that NTPSBPLD contains several one-parameter and two-parameter size-biased distributions including size-biased geometric distribution (SBGD), size-biased negative binomial distribution (SBNBD), size-biased Poisson-Lindley distribution (SBPLD) proposed by Ghitany and Mutairi,5 size-biased Poisson-Shanker distribution(SBPSD) proposed by Shanker,6 size-biased two-parameter Poisson-Lindley distribution-1 (SBTPPLD-1) introduced by Shanker,7 size-biased two-parameter Poisson-Lindley distribution-2 (SBTPPLD-2) suggested by Shanker and Mishra,8 size-biased quasi Poisson-Lindley distribution (SBQPLD) proposed by Shanker and Mishra9 and size-biased new quasi Poisson-Lindley distribution (SBNQPLD) introduced by Shanker et al.,10 respectively for(β=0,α=1)(β=0,α=1) , (α=0,β=1)(α=0,β=1) ,(α=β=1)(α=β=1) ,(α=θ,β=1)(α=θ,β=1) ,(α=1)(α=1) , (β=1)(β=1) ,(β=θ)(β=θ) ,(α=θ)(α=θ) respectively.

Various characteristics of a distribution are based on their moments and it not easy to derive the moments of NTPSBPLD directly. Therefore, to derive the moments of NTPSBPLD, the pmf of NTPSBPLD can also be obtained as follows:

Let the random variable follows the size-biased Poisson distribution (SBPD) with parameter and pmf

g(x|λ)=eλλx1(x1)!;x=1,2,3,...,;λ>0g(x|λ)=eλλx1(x1)!;x=1,2,3,...,;λ>0     (2.2)

Suppose the parameterof SBPD follows the size-biased three-parameter Lindley distribution with pdf

h(λ;θ)=θ3θα+2βλ(α+λβ)eθλ;λ>0,(θ,α,β)>0h(λ;θ)=θ3θα+2βλ(α+λβ)eθλ;λ>0,(θ,α,β)>0     (2.3)

Thus the pmf of NTPSBPLD can be obtained as

P(X=x)=0g(x|λ)h(λ;θ)dλP(X=x)=0g(x|λ)h(λ;θ)dλ

=0eλλx1(x1)!θ3θα+2βλ(α+λβ)eθλdλ=0eλλx1(x1)!θ3θα+2βλ(α+λβ)eθλdλ     (2.4)


=θ3(θα+2β)(x1)!0e(θ+1)λλx(α+λβ)dλ=θ3(θα+2β)(x1)!0e(θ+1)λλx(α+λβ)dλ

=θ3(θα+2β)(x1)![αΓ(x+1)(θ+1)x+1+βΓ(x+2)(θ+1)x+2]=θ3(θα+2β)(x1)![αΓ(x+1)(θ+1)x+1+βΓ(x+2)(θ+1)x+2]

=θ3θα+2ββx2+(θα+α+β)x(θ+1)x+2;x=1,2,3,..,(θ,α,β)>0=θ3θα+2ββx2+(θα+α+β)x(θ+1)x+2;x=1,2,3,..,(θ,α,β)>0

which is the pmf of NTPSBPLD obtained in (2.1).

The behavior of the pmf of NTPSBPLD for varying values of parameters (θ,α,β)(θ,α,β) has been shown in Figure 1.

Figure 1 Behavior of NTPSBPLD for (θ,α,β)(θ,α,β) .

Moments

Using (2.4), the th factorial moment about origin μ(r)μ(r) of the NTPSBPLD (2.1) can be obtained as

μ(r)=E[E(X(r)|λ)]μ(r)=E[E(X(r)λ)] , where X(r)=X(X1)(X2)...(Xr+1)X(r)=X(X1)(X2)...(Xr+1)


=0[x=1x(r)eλλx1(x1)!]θ3θα+2βλ(α+λβ)eθλdλ=0[x=1x(r)eλλx1(x1)!]θ3θα+2βλ(α+λβ)eθλdλ

=0[λr1{x=rxeλλxr(xr)!}]θ3θα+2βλ(α+λβ)eθλdλ=0[λr1{x=rxeλλxr(xr)!}]θ3θα+2βλ(α+λβ)eθλdλ

Taking y=xry=xr , we get

μ(r)=0[λr1{y=0(y+r)eλλyy!}]θ3θα+2βλ(α+λβ)eθλdλμ(r)=0[λr1{y=0(y+r)eλλyy!}]θ3θα+2βλ(α+λβ)eθλdλ

=θ3θα+2β0λr(λ+r)(α+λβ)eθλdλ=θ3θα+2β0λr(λ+r)(α+λβ)eθλdλ

=θ3θα+2β0{βλr+2+(α+βr)λr+1+rαλr}eθλdλ=θ3θα+2β0{βλr+2+(α+βr)λr+1+rαλr}eθλdλ

After a little tedious algebraic simplification, the th factorial moment about origin of NTPSBPLD (2.1) can be expressed as

μ(r)=r!{rαθ2+(r+1)(α+rβ)θ+(r+1)(r+2)β}θr(θα+2β);r=1,2,3,...μ(r)=r!{rαθ2+(r+1)(α+rβ)θ+(r+1)(r+2)β}θr(θα+2β);r=1,2,3,...     (3.1)

The first four factorial moments about origin can be obtained by taking r=1,2,3,and4r=1,2,3,and4 in (3.1). The first four moments about origin of the NTPSBPLD, using the relationship between moments about origin and factorial moments about origin, are obtained as  


μ1=αθ2+2(α+β)θ+6βθ(θα+2β)μ1=αθ2+2(α+β)θ+6βθ(θα+2β)

μ2=αθ3+(6α+2β)θ2+(6α+18β)θ+24βθ2(θα+2β)μ2=αθ3+(6α+2β)θ2+(6α+18β)θ+24βθ2(θα+2β)

μ3=αθ4+(14α+2β)θ3+(36α+42β)θ2+(24α+144β)θ+120βθ3(θα+2β)μ3=αθ4+(14α+2β)θ3+(36α+42β)θ2+(24α+144β)θ+120βθ3(θα+2β)

μ4=αθ5+(30α+2β)θ4+(150α+90β)θ3+(240α+600β)θ2+(120α+1200β)+720βθ4(θα+2β)μ4=αθ5+(30α+2β)θ4+(150α+90β)θ3+(240α+600β)θ2+(120α+1200β)+720βθ4(θα+2β)

Now, using the relationship between moments about mean and the moments about origin, the moments about mean of the NTPSBPLD (2.1) can be obtained as


μ2=2{α2θ3+(α2+5αβ)θ2+(6β2+6αβ)θ+6β2}θ2(θα+2β)2μ2=2{α2θ3+(α2+5αβ)θ2+(6β2+6αβ)θ+6β2}θ2(θα+2β)2

μ3=2{α3θ5+(7α2β+3α3)θ4+(16αβ2+24α2β+2α3)θ3+(54αβ2+12β3+18α2β)θ2+(36αβ2+36β3)θ+24β3}θ3(θα+2β)3

μ4=2{α4θ7+(13α4+9α3β)θ6+(30α2β2+130α3β+24α2)θ5+(460α2β2+44αβ3+264α3β+12α4)θ4+(936α2β2+24β4+144α3β+696αβ3)θ3+(384β4+1368αβ3+504α2β2)θ2+(720β4+720αβ3)θ+360β4}θ4(θα+2β)4

The coefficient of variation (C.V) , coefficient of Skewness (β1) , coefficient of Kurtosis (β2) and index of dispersion (γ) of the NTPSBPLD (2.1)) are thus obtained as 

C.V=σμ1=2{α2θ3+(α2+5αβ)θ2+(6β2+6αβ)θ+6β2}{αθ2+2(α+β)θ+6β}

β1=μ3μ23/2={α3θ5+(7α2β+3α3)θ4+(16αβ2+24α2β+2α3)θ3+(54αβ2+12β3+18α2β)θ2+(36αβ2+36β3)θ+24β3}2{α2θ3+(α2+5αβ)θ2+(6β2+6αβ)θ+6β2}3/2

β2=μ4μ22={α4θ7+(13α4+9α3β)θ6+(30α2β2+130α3β+24α2)θ5+(460α2β2+44αβ3+264α3β+12α4)θ4+(936α2β2+24β4+144α3β+696αβ3)θ3+(384β4+1368αβ3+504α2β2)θ2+(720β4+720αβ3)θ+360β4}2{α2θ3+(α2+5αβ)θ2+(6β2+6αβ)θ+6β2}2

γ=σ2μ1=2{α2θ3+(α2+5αβ)θ2+(6β2+6αβ)θ+6β2}θ(θα+2β){(αθ2+2(α+β)θ+6β)}

The graphs of coefficient of variation (C.V) , coefficient of Skewness (β1) , coefficient of Kurtosis (β2) and index of dispersion (γ) of the NTPSBPLD are shown in figures 2,3,4 and 5 respectively.

Figure 2 Graphs of coefficient of Variation of the NTPSBPLD for varying values of the parameters (θ,α,β) .

Figure 3 Graphs of coefficient of Skewness of the NTPSBPLD for varying values of the parameters (θ,α,β) .

Figure 4 Graphs of Coefficient of Kurtosis of the NTPSBPLD for varying values of the parameter (θ,α,β) .

Figure 5 Index of dispersion of the NTPSBPLD for varying values of the parameter (θ,α,β) .

Maximum likelihood estimation

Let us consider (x1,x2,x3,...,xn) as random sample from NTPSBPLD (θ,α,β) . Suppose fx be the observed frequency in the sample corresponding to X=x(x=1,2,3,...,k) such that kx=1fx=n , where k is the largest observed value having non-zero frequency. The likelihood function L of NTPSBPLD (θ,α,β) can be expressed as

L=(θ3θα+2β)n1(θ+1)kx=1fx(x+2)kΠx=1[βx2+(θα+α+β)x]fx .

The log likelihood function of NTPSBPLD (θ,α,β) is

    logL=n{3logθlog(θα+2β)}kx=1fx(x+2)log(θ+1)+kx=1fxlog{βx2+(θα+α+β)x}    

The maximum likelihood estimates, MLE’s (ˆθ,ˆα,ˆβ) , of parameters (θ,α,β) of NTPSBPLD (θ,α,β) is the solutions of the following log likelihood equations

     logLθ=3nθ3nαθα+2βn(ˉx+2)θ+1+kx=1αxfxβx2+(θα+α+β)x=0      logLα=nθθα+2β+kx=1(θ+1)xfxβx2+(θα+α+β)x=0      logLβ=2nθα+2β+kx=1(x2+x)fxβx2+(θα+α+β)x=0 ,     

where ˉx is the sample mean.

Since these log likelihood equations cannot be expressed in closed forms and hence do not seem to be solved directly, the (MLE’s) (ˆθ,ˆα,ˆβ) of parameters (θ,α,β) can be computed directly by solving the log likelihood equation using R-software till sufficiently close estimates of (ˆθ,ˆα,ˆβ) are attained.

Goodness of fit

The goodness of fit of NTPSBPLD has been discussed with several count data from various fields of knowledge. The expected frequencies according to the SBPLD, SBQPLD and SBNQPLD using maximum likelihood estimates of parameters have also been given in these tables for ready comparison with those obtained by the NTPSBPLD. Clearly the goodness of fit of NTPSBPLD provides better fit over SBPLD and competing well with SBQPLD and SBNQPLD in majority of datasets. In some of the tables the degree of freedom is zero, and hence p-values have not been given and thus in such tables comparisons can be done on the basis of values of and AIC (Akaike information criterion). The datasets considered for testing the goodness of fit of SBPLD, SBQPLD, SBNQPLD and NTPSBPLD as follows: (Tables i-x)

Group Size

Observed frequency

Expected frequency

SBPLD

SBQPLD

SBNQPLD

NTPSBPLD

1
2
3
4
5
6

1486
694
195
37
10
1

1532.5
630.6
191.9
51.3
12.8
3.9

1485.4
697.2
189.7
41.1
7.8
1.8

1505.5
656.8
202.5
49.2
9.0
0.0

1485.4
697.2
189.7
41.1
7.8

Total

2423

2423.0

2423

 

 

ML Estimate

 

ˆθ=4.5082

ˆθ= 7.14063
ˆα= -0.79104

ˆθ= 2.69606
ˆα= -1.39128

ˆθ= 7.1386
ˆα= -0.9318
ˆβ=8.4164

χ2

 

13.760

0.776

6.1

0.77

d.f.

 

3

2

2

1

p-value

 

0.003

0.6804

0.04735

0.3802

2logL

 

4622.36

4607.8

4610.0

4607.8

AIC

 

4624.36

4611.8

4614.0

4613.8

Table i Pedestrians-Eugene, Spring, Morning, available in Coleman and James21

Group Size

Observed frequency

Expected frequency

SBPLD

SBQPLD

SBNQPLD

NTPSBPLD

1
2
3
4
5

316
141
44
5
4

322.9
132.5
40.2
10.7
3.7

315.7
142.7
40.1
9.1
2.4

313.5
141.4
44.1
10.4
0.6

315.7
142.7
40.1
9.1
2.4

Total

510

510.0

510.0

 

 

ML Estimate

 

ˆθ= 4.5211

ˆθ= 6.5501
ˆα= -0.5069

ˆθ= 2.4693
ˆα= -1.2977

ˆθ= 6.5560
ˆα= -0.6029
ˆβ=7.7499

χ2

 

3.07

0.94

0.38

0.94

d.f.

 

2

1

1

0

p-value

 

0.2154

0.3322

0.5376

 

2logL

 

972.78

971.07

970.24

971.07

AIC

 

974.78

975.07

974.24

977.07

Table ii Play Groups-Eugene, spring, Public Playground A, available in Coleman and James21

Group Size

Observed frequency

Expected frequency

SBPLD

SBQPLD

SBNQPLD

NTPSBPLD

1
2
3
4
5

306
132
47
10
2

309.4
131.2
41.1
11.3
4.0

304.4
137.9
41.3
10.3
3.1

306.4
134.4
41.6
11.0
3.6

304.4
137.9
41.3
10.3
3.1

Total

497

497.0

 

 

 

ML Estimate

 

ˆθ=4.3548

ˆθ= 5.71547
ˆα= -0.06947

ˆθ= 4.9998
ˆα= 25.6948

ˆθ= 5.7156
ˆα= -0.0708
ˆβ=5.8180

χ2

 

0.932

1.19

1.2

1.19

d.f.

 

2

1

1

0

p-value

 

0.6281

0.2753

0.2733

 

2logL

 

971.86

970.96

971.25

970.9

AIC

 

973.86

974.96

975.25

976.9

Table iii Play Groups-Eugene, spring, Public Playground A, available in Coleman and James21

Group Size

Observed frequency

Expected frequency

SBPLD

SBQPLD

SBNQPLD

NTPSBPLD

1
2
3
4
5
6

305
144
50
5
2
1

314.4
134.4
42.5
11.8
3.1
0.8

304.3
148.2
42.3
9.6
1.9
0.7

310.1
138.8
43.1
11.3
2.7
1.0

304.3
148.2
42.3
9.6
1.9
0.7

Total

507

507.0

507.0

507.0

 

ML Estimate

ˆθ=4.3179

ˆθ= 6.70804
ˆα= -0.74907

ˆθ= 6.70804
ˆα= -0.74907

ˆθ= 5.1516
ˆα= 48.6067

ˆθ= 6.7082
ˆα= -0.8290
ˆβ=7.4234

χ2

 

6.415

2.96

4.64

2.96

d.f.

 

2

1

1

0

p-value

 

0.040

0.0853

0.0312

 

2logL

 

993.10

990.02

991.51

990.02

AIC

 

995.1

994.02

995.51

996.02

Table iv Play Groups-Eugene, Spring, Public Playground D, available in Coleman and James21

Group Size

Observed frequency

Expected frequency

SBPLD

SBQPLD

SBNQPLD

NTPSBPLD

1
2
3
4
5

276
229
61
12
3

319.6
166.5
63.8
21.4
9.7

276.0
228.3
61.9
12.2
2.6

313.7
173.1
65.2
20.7
8.3

276.0
228.3
61.9
12.2
2.6

Total

581

581.0

581.0

581.0

581.0

ML Estimate

ˆθ=4.3179

ˆθ= 3.4359
ˆα= -0.74907

ˆθ=

8.6724
ˆα= -1.4944

ˆθ= 4.1645
ˆα= 61.0287
ˆβ=7.4234

ˆθ= 8.6726
ˆα= -2.5854
ˆβ=15.0041

χ2

 

37.86

0.017

29.6

0.017

d.f.

 

2

1

1

0

p-value

 

0.00

0.8962

0.000

0.0000

2logL

 

1277.42

1238.11

1268.77

1238.11

AIC

 

1279.42

1242.11

1272.77

1244.11

Table v Play Groups-Eugene, Spring, Public Playground D, available in Coleman and James21

No. of sites with particles

Observed Frequency

Expected Frequency

SBPLD

SBQPLD

SBNQPLD

NTPSBPLD

1
2
3
4
5

122
50
18
4
4

119.0
53.8
18.0
5.31.9}

119.2
53.5
17.9
5.3
2.1

119.3
53.3
17.8
5.3
2.3

119.3
53.3
17.8
5.3
2.3

Total

198

198.0

198.0

198.0

198.0

ML estimate

 

ˆθ=4.050987

ˆθ= 3.7564
ˆα= 10.1281

ˆθ= 3.4795
ˆα= 0.0216

ˆθ= 3.4737
ˆα= 1.3965
ˆβ=0.0001

χ2

 

0.43

0.34

0.28

0.28

d.f.

 

2

1

1

0

p-value

 

0.8065

0.5598

0.5967

 

2logL

 

409.28

409.17

409.13

409.13

AIC

 

411.28

413.17

413.13

415.13

Table vi Distribution of number of counts of sites with particles from Immunogold data, available in Mathews and Appleton22

No. times hares caught

Observed Frequency

Expected Frequency

SBPLD

SBQPLD

SBNQPLD

NTPSBPLD

1
2
3
4
5

184
55
14
4
4

177.3
62.5
16.43.81.0}

177.4
62.3
16.3
3.8
1.2

177.5
62.2
16.3
3.8
1.2

177.5
62.2
16.3
3.8
1.2

Total

261

261.0

261

261.0

 

ML estimate

 

ˆθ=5.351256

ˆθ= 4.9800
ˆα= 14.9193

ˆθ= 4.6959
ˆα= -0.0302

ˆθ= 4.6994
ˆα= 12.0044
ˆβ=0.0390

χ2

 

1.18

3.2

3.19

3.19

d.f.

 

1

1

1

0

p-value

 

0.2773

0.0736

0.07409

 

2logL

 

457.10

456.86

456.80

456.80

AIC

 

459.10

460.86

460.80

462.80

Table vii Distribution of snowshoe hares captured over 7 days, available in Keith and Meslow23

Number of pairs of running shoes

Observed frequency

Expected Frequency

SBPLD

SBQPLD

SBNQPLD

NTPSBPLD

1

18

20.3

17.4

19.5

17.4

2

18

17.4

19.6

18.0

19.6

3

12

10.9

12.3

11.3

12.3

4
5

7
5

5.9
5.5

6.1
4.6

6.0
5.2

6.1
4.6

Total

60

60.0

60.0

60

60

ML Estimate

 

ˆθ=1.818978

ˆθ= 2.5858
ˆα= -0.7318

ˆθ= 2.08739
ˆα= 17.3228

ˆθ= 2.5870
ˆα= -0.4739
ˆβ=1.6732

χ2

 

0.64

0.31

0.33

0.31

d.f.

 

3

1

2

0

P-value

 

0.8872

0.5777

0.8478

 

2logL

 

187.08

185.55

186.33

185.55

AIC

 

189.08

189.55

190.33

191.55

Table viii Number of counts of pairs of running shoes owned by 60 members of an athletics club, reported by Simonoff24

Number of fly eggs

Observed Frequency

Expected Frequency

SBPLD

SBQPLD

SBNQPLD

NTPSBPLD

1
2
3
4
5
6
7
8
9

22
18
18
11
9
6
3
0
1

20.3
22.0
17.2
11.6
7.2
4.2
2.4
1.3
1.8

19.8
22.1
17.5
11.8
7.3
4.2
2.3
1.3
1.7

19.8
22.1
17.5
11.8
7.3
4.2
2.3
1.3
1.7

19.8
22.1
17.5
11.8
7.3
4.2
2.3
1.3
1.7

Total

88

 

88.0

88.0

88.0

ML estimate

 

ˆθ= 1.2822

ˆθ= 1.3483
ˆα= 0.6925

ˆθ= 1.3465
ˆα= 2.5654

ˆθ= 1.4477
ˆα= 0.4315
ˆβ=1.3594

χ2

 

1.39

1.49

1.49

1.49

d.f.

 

4

3

3

3

p-value

 

0.8459

0.6845

0.6845

0.6845

2logL

 

329.92

329.86

329.86

329.86

AIC

 

331.92

333.86

333.86

335.86

Table ix The numbers of counts of flower heads as per the number of fly eggs reported by Finney and Varley25

X

Observed frequency

Expected frequency

SBPLD

SBQPLD

SBNQPLD

NTPSBPLD

1

375

262.8

363.3

363.6

363.6

2

143

157.4

156.5

156.3

156.3

3

49

50.4

50.4

50.4

50.4

4
5
6
7
8

17
2
2
1
1

14.2
3.7
0.9
0.2
0.3

14.4
3.9
1.0
0.2
0.4

14.4
3.8
1.0
0.2
0.3

14.4
3.8
1.0
0.2
0.3

Total

590

590.0

590.0

590.0

590.0

ML Estimate

 

ˆθ= 4.24

ˆθ= 3.8386
ˆα= 17.2968

ˆθ= 3.6534
ˆα= 0.00067

ˆθ= 3.6504
ˆα= 12.9869
ˆβ=0.0377

χ2

 

2.48

2.11

2.08

2.08

d.f.

 

3

2

2

1

P-value

 

0.4789

0.3481

0.3534

0.1492

2logL

 

1190.4

1189.67

1189.57

1189.57

AIC

 

1192.4

1193.67

1193.57

1193.57

Table x Number of households having at least one migrant according to the number of observed migrants, reported by Singh and Yadav26

Conclusion

A new three-parameter size-biased Poisson-Lindley distribution which includes several size-biased distributions including size-biased geometric distribution (SBGD), size-biased negative binomial distribution (SBNBD), size-biased Poisson-Lindley distribution (SBPLD), size-biased Poisson-Shanker distribution (SBPSD), size-biased two-parameter Poisson-Lindley distribution-1 (SBTPPLD-1), size-biased two-parameter Poisson-Lindley distribution-2 (SBTPPLD-2), size-biased quasi Poisson-Lindley distribution (SBQPLD) and size-biased new quasi Poisson-Lindley distribution (SBNQPLD) for particular values of parameters has been proposed. Its coefficient of variation, skewness, kurtosis and index of dispersion has been studied. Estimation of parameters has been discussed using maximum likelihood. Goodness of fit of the proposed distribution has been discussed with several count datasets.

Acknowledgments

Authors are grateful to the Editor-In-Chief of the Journal and the anonymous reviewer for minor comments for the improvement in the paper.

Conflicts of interest

Authors declare that there is no conflict of interests.

References

  1. Shanker R, Shukla KK, Shanker R, et al. A three–parameter lindley distribution. American Journal of Mathematics and Statistics. 2017;7(1):15–26.
  2. Lindley DV. Fiducial distributions and Bayes’ theorem. Journal of the Royal Statistical Society, Series B. 1958;20:102–107.
  3. Ghitany ME, Atieh B, Nadarajah S. Lindley distribution and its application. Mathematics Computing and Simulation. 2008;78:493–506.
  4. Das KK, Ahmad J, Bhattacharjee S. A new three–parameter poisson–lindley distribution for modeling over dispersed count data, International journal of Applied Engineering Research. 2018;13(25):16468–16477.
  5. Ghitany ME, Al–Mutairi DK. Size–biased poisson–lindley distribution and its applications, Metron. 2008; LXVI(3):299–311.
  6. Shanker R. A Size–biased poisson–shanker distribution and its applications. International Journal of Probability and Statistics, 2017;6(3):33–44.
  7. Shanker R. Size–biased discrete two parameters poisson–lindley distribution and its applications, Global Journal of Pure and Applied Mathematics. 2013;9(4):325–334.
  8. Shanker R, Mishra A. On size–biased two–parameter poisson–lindley distribution and its applications. American Journal of Mathematics and Statistics. 2017;7(3):99–107.
  9. Shanker R, Mishra A. on a size–biased quasi poisson lindley distribution, International Journal of Probability and Statistics. 2013b;2(2):28–34.
  10. Shanker R, Shanker R, Tekie AL, et al. Size–biased new quasi poisson lindley distribution and its applications, International Journal of Statistics and Systems. 2014; 9(2):109–118.
  11. Shanker R. Shanker Distribution and Its Applications. International Journal of Statistics and Applications. 2015;5(6):338–348.
  12. Shanker R, Mishra A. A two–parameter Lindley distribution, Statistics in Transition–new series. 2013a;14(1):45–56.
  13. Shanker R, Sharma S, Shanker R. A two–parameter Lindley distribution for modeling waiting and survival times data. Applied Mathematics. 2013;4(2):363–368.
  14. Shanker R, Mishra A. A quasi lindley distribution. African Journal of Mathematics and Computer Science Research. 6(4):64–71.
  15. Shanker R, Amanuel AG. A new quasi lindley distribution, International Journal of Statistics and Systems. 2013;8(2):143–156.
  16. Greenwood M, Yule Udny G. An Inquiry into the nature of frequency distributions representative of multiple happenings with particular reference to the occurrence of multiple attacks of disease or of repeated accidents. Journal of the Royal Statistical Society. 1920;83(2):255–279.
  17. Sankaran M. The discrete Poisson–Lindley distribution, Biometrics. 1970;26(1):145–149.
  18. Shanker R, Mishra A. A two–parameter poisson–lindley distribution, International Journal of Statistics and Systems. 2014;9(1):79–85.
  19. Shanker R, Mishra A. A quasi Poisson– Lindley distribution. Journal of 2014. Indian Statistical Association. 2016;54(1&2):113–125.
  20. Shanker R, Tekie AL. A new quasi Poisson–Lindley distribution, International Journal of Statistics and Systems. 2014;9(1):79 –85.
  21. Coleman JS, James J. The equilibrium size distribution of freely forming groups. Sociometry, 1961;24:36–45.
  22. Mathews JNS, Appleton DR. An application of the truncated Poisson distribution to Immunogold assay. Biometrics. 1993;49(2):617–621.
  23. Keith LB, Meslow EC. Trap response by snowshoe hares. Journal of Wildlife Management. 1968;32(4):795–801.
  24. Simonoff JS. Analyzing Categorical Data, New York Springer. 2003.
  25. Finney DJ, Varley GC. An example of the truncated Poisson distribution, Biometrics, 1955;11(3):387–394.
  26. Singh SN, Yadav RC. Trends in rural out–migration at household level. Rural Demogr. 1971;8(1):53–61.
Creative Commons Attribution License

©2020 Shanker, et al. This is an open access article distributed under the terms of the, which permits unrestricted use, distribution, and build upon your work non-commercially.