• Communications for Statistical Applications and Methods
• /
• v.2 no.2
• /
• pp.155-165
• /
• 1995

반복측정자료의 분석을 위해 제안된 Liang and Zeger(1986)의 회귀모형은 일반화추정식(generalized estimationg equations, GEE)을 이용하여 모형의 모수를 추정한다. 이 모형은 반복측정된 반응변수와 설명변수들과의 관계를 추정하는 것이 주된 목적이기 때문에 회귀모수는 중요한 모수로 간주되나 산포모수는 중요하지 않은 장애모수(nuisance parameters)로 간주된다. 일반적으로 GEE 분석에서 회귀모수의 추정량은 산포모수에 상관없이 일치적(consistent)으로 얻어진다고 알려져 있다. 그러나 본 논문에서는 포아송분포를 따르는 반복측정자료에 대한 사례연구와 모의 실험을 통해서 일반적으로 믿어져왔던 것과는 달리 GEE 방법이 산포모수에 민감하게 영향을 받고 있음을 보였다. 특히 산포모수의 값이 일정하지 않은 경우에는 GEE 방법이 산포모수에 민감 하게 영향을 받고 있음을 보였다. 특히 산포모수의 값이 일정하지 않은 경우에는 GEE 방법에서 밝혀진 회귀모수 추정량의 일치성에도 문제가 발생할 수 있음을 보였다.

• Proceedings of the Safety Management and Science Conference
• /
• 2013.04a
• /
• pp.593-607
• /
• 2013

관리도를 사용하여 공정평균, 공정산포 등 여러 가지 공정모수를 관리할 수 있다. 그러나 공정모수의 미세 변동을 효과적으로 관리할 수 있는 기법체계는 아직 미완이다. 식스시그마 공정관리 등 정밀공정관리를 위해서는 미세 공정평균과 공정산포관리가 전제되어야한다. 특히 높은 수준의 공정능력을 유지하기위해서는 공정산포관리가 선결과제이다. 본 본문에서는 공정평균과 공정불량률, 공정산포의 미세변동을 효과적으로 관리할 수 있는 기술체계의 연구동향을 분석하고 미세공정산포관리를 위한 대안을 제시하고자 한다.

• The Korean Journal of Applied Statistics
• /
• v.30 no.1
• /
• pp.95-102
• /
• 2017

Gamma generalized linear models have received less attention than Poisson and binomial generalized linear models. Therefore, many old-established statistical techniques are still used in gamma generalized linear models. In particular, existing literature and textbooks still use approximate estimates for the dispersion parameter. In this paper we study the efficiency of various dispersion parameter estimators in gamma generalized linear models and perform numerical simulations. Numerical studies show that the maximum likelihood estimator and Cox-Reid adjusted maximum likelihood estimator are recommended and that approximate estimates should be avoided in practice.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.1
• /
• pp.153-161
• /
• 2017

Branching processes which is used for epidemic dispersion as stochastic process model have advantages to estimate parameters by real data. We have to estimate both mean and dispersion parameter in order to use the negative binomial distribution as an offspring distribution on branching processes. In existing studies on biology and epidemiology, it is estimated using maximum-likelihood methods. However, for most of epidemic data, it is hard to get the best precision of maximum-likelihood estimator. We suggest a Bayesian inference that have good properties of statistics for small-sample. After estimating dispersion parameter we modelled the posterior distribution for 2015 Korea MERS cases. As the result, we found that the estimated dispersion parameter is relatively stable no matter how we assume prior distribution. We also computed extinction probabilities on branching processes using estimated dispersion parameters.

• The Korean Journal of Applied Statistics
• /
• v.23 no.3
• /
• pp.585-594
• /
• 2010

This study, investigates the effects of dispersion parameters between two response variables in zero-truncated bivariate generalized Poisson distributions. A Monte Carlo study shows that the zero-truncated bivariate Poisson and negative binomial models fit poorly wherein the zero-truncated bivariate count data has heterogeneous dispersion parameters on dependent variables. In addition, we derive the score test for testing the equality of the dispersion parameters and compare its efficiency with the likelihood ratio test.

• The Korean Journal of Applied Statistics
• /
• v.30 no.4
• /
• pp.529-538
• /
• 2017

The Ghosh and Kim zero-altered distribution model is used to analyze count data that have too many or too few zeros. The dispersion type parameter ${\delta}$ in the zero-altered distribution model consists of mean, variance and zero probability and has two forms depending on the relation between ${\mu}$ and ${\sigma}^2$. We derived the influence function on ${\delta}$ when ${\sigma}^2{\geq}{\mu}$. To show the validity of the influence function, we used the Census data on the number of births of married women in Korea to compare the estimated changes in ${\delta}$ using this function with those obtained using the direct deletion method. The result proved that the obtained influence function is very accurate in estimating changes in ${\delta}$ when an observation is deleted.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.3
• /
• pp.493-504
• /
• 2017

We often observe count data that exhibit over-dispersion, originating from too many zeros, and under-dispersion, originating from too few zeros. To handle this types of problems, the zero-altered distribution model is designed by Ghosh and Kim in 2007. Their model can control both over-dispersion and under-dispersion with a single parameter, which had been impossible ever. The dispersion type depends on the sign of the parameter ${\delta}$ in zero-altered distribution. In this study, we demonstrate the role of the dispersion type parameter ${\delta}$ through the data of the number of births in Korea. Employing both the chi-square statistic and the Kolmogorov statistic for goodness-of-fit, we also explained any difference between the theoretical distribution and the observed one that exhibits either over-dispersion or under-dispersion. Finally this study shows whether the test statistics for goodness-of-fit show any similarity with the role of the dispersion type parameter ${\delta}$ or not.

• Journal of the Korean Data and Information Science Society
• /
• v.8 no.2
• /
• pp.195-209
• /
• 1997

The joint modelling of mean and dispersion in quasi-likelihood models which greatly extend the scope of generalized linear models, is required in case that the dispersion parameter, the variance component of response variables, is not constant but changes by depending on any covariates. In this paper, by using statistical package GENSTAT(release 5.3.2, 1996) which makes a easily analyze real data through this joint modelling, we mention necessities that must consider this joint modelling rather than existing mean models through model checking based on graphic methods for esterase assay data introduced by Carrol and Ruppert(1987, pp.46-47), and then study methods finding reasonable joint model of mean and dispersion for this data.

• The Korean Journal of Applied Statistics
• /
• v.31 no.6
• /
• pp.733-749
• /
• 2018

Sample selection arises as a result of the partial observability of the outcome of interest in a study. Heckman introduced a sample selection model to analyze such data and proposed a full maximum likelihood estimation method under the assumption of normality. Recently sample selection models for binomial and Poisson response variables have been proposed. Based on the theory of symmetry-modulated distribution, we extend these to a model for overdispersed count data. This type of data with no sample selection is often modeled using negative binomial distribution. Hence we propose a sample selection model for overdispersed count data using the negative binomial distribution. A real data application is employed. Simulation studies reveal that our estimation method based on profile log-likelihood is stable.

• The Korean Journal of Applied Statistics
• /
• v.6 no.1
• /
• pp.41-52
• /
• 1993

In this paper we study numerical behavior of the adjustments for the variances of the pearson and deviance type dispersion statistics in two overdispersed mixture models; negative binomial and beta-binomial distribution. They are important families of an extended quasi-likelihood model which is very useful for the joint modelling of mean and dispersion. Comparisons are done for two types of dispersion statistics for various mean and dispersion parameters by simulation studies.