Go to the main menu
Skip to content
Go to bottom
REFERENCE LINKING PLATFORM OF KOREA S&T JOURNALS
> Journal Vol & Issue
Korean Journal of Applied Statistics
Journal Basic Information
Journal DOI :
The Korean Statistical Society
Editor in Chief :
Volume & Issues
Volume 10, Issue 2 - Sep 1997
Volume 10, Issue 1 - Mar 1997
Selecting the target year
The current status and the improvable directions of statistical system in Korea
Korean Journal of Applied Statistics, volume 10, issue 1, 1997, Pages 5~14
The difficulty of statistical survey has been weighted according to the development and the democratization of society. Since the survey cost increases and the quantity and the quality of the needed information become more diversity it is very hard to produce accurate and appropriate statistics in time. In order to cope with these survey conditions properly, we reviewed the statistical system of Korea and major countries, and proposed the improvable directions of our statistical system.
Alternation to the Randomized Block Design for Agricultural Experiments in Korea
Korean Journal of Applied Statistics, volume 10, issue 1, 1997, Pages 15~27
Randomized block design (RBD) with three replication is very frequently adopted in agricultural experiments of the Rural Development Administration of Korea. Even though it works well in field trials of traditional crops, it may not accomodate trial site conditions and/or experimental environment. In this research report, we deal with two such cases. The first case is for a crop experiment in green houses. In house conditions, RBD may not be appropriate since it cannot reflect two directions of the yield gradient. So, a Latin square design is suggested as an alternative. The second case is for local field experiments of the newly-inbred rice. RBD with three replications is used without doubt for decades, even though the site layout is not appropriately shaped for the design. In this case, we suggest the RBD in two blocks with multiple replicates for control varieties as an alternative. To improve the quality of statistical experimental designs in over one-thousand agricultural trials performed annually in the Rural Development Administration, we need to re-train agricultural researchers on the design and analysis of experiments and call for concerns of Korean statisticians.
Comparison of statistical packages useful for statistical process control
Korean Journal of Applied Statistics, volume 10, issue 1, 1997, Pages 29~36
Four statistical packages useful for statistical process control are reviewed. MINITAB, RS, SAS, and SPSS are compared in terms of general capability, control chart, process capability, design of experiment and Taguchi method.
Survey results of the usage of KOSIS and its improvement
Korean Journal of Applied Statistics, volume 10, issue 1, 1997, Pages 37~46
KOSIS(Korean Statistical Information System) was developed and serviced by National Statistical Office in Korea. In this paper, we report the survey results of the usage of KOSIS, and its improvement. Finally, we propose some recommendations for improvement of KOSIS, and discuss the usage of KOSIS in behalf of end-users.
A Heuristic approach for efficient blood testing : The case of Unknown prevalence rate
Korean Journal of Applied Statistics, volume 10, issue 1, 1997, Pages 47~60
Based on group testing, we present efficient blood testing procedures for the case of unknown prevalence rate.
A study of (M,S)-optimal h(
)-concorrance most balanced design
Korean Journal of Applied Statistics, volume 10, issue 1, 1997, Pages 61~68
In this paper, we show that the proper equi-replicate binary designs with h(
)-concurrence most balanced designs are (M, S)-optimal. It is very difficult to construct with more than 3-concurrence designs, so we choose 3-concurrence most balanced designs out of the known PBIBD(3).
Performance of a Bayesian Design Compared to Some Optimal Designs for Linear Calibration
Korean Journal of Applied Statistics, volume 10, issue 1, 1997, Pages 69~84
We consider a linear calibration problem,
where we observe
's for the controlled calibration experiments and later we make inference about
from a new observation
. The objective of the calibration design problem is to find the optimal design
that gives the best estimates for
. We compare Kim(1989)'s Bayesian design which minimizes the expected value of the posterior variance of
and some optimal designs from literature. Kim suggested the Bayesian optimal design based on the analysis of the characteristics of the expected loss function and numerical must be equal to the prior mean and that the sum of squares be as large as possible. The designs to be compared are (1) Buonaccorsi(1986)'s AV optimal design that minimizes the average asymptotic variance of the classical estimators, (2) D-optimal and A-optimal design for the linear regression model that optimize some functions of
, and (3) Hunter & Lamboy (1981)'s reference design from their paper. In order to compare the designs which are optimal in some sense, we consider two criteria. First, we compare them by the expected posterior variance criterion and secondly, we perform the Monte Carlo simulation to obtain the HPD intervals and compare the lengths of them. If the prior mean of
is at the center of the finite design interval, then the Bayesian, AV optimal, D-optimal and A-optimal designs are indentical and they are equally weighted end-point design. However if the prior mean is not at the center, then they are not expected to be identical.In this case, we demonstrate that the almost Bayesian-optimal design was slightly better than the approximate AV optimal design. We also investigate the effects of the prior variance of the parameters and solution for the case when the number of experiments is odd.
A study on the goodness-of-fit tests for proportional hazards model
Korean Journal of Applied Statistics, volume 10, issue 1, 1997, Pages 85~104
Proportional hazards model has been widely used for analyzing survival data. This article reviews some well-known goodness-of-fit tests for proportional hazards model. Simulation studies also provide some insights into the properties of these test statistics across several types of survival distributions and degerees of censorship.
Analysis of counts in the one-way layout
Korean Journal of Applied Statistics, volume 10, issue 1, 1997, Pages 105~119
Barnwal and Paul(1988) derived the likelihood ratio statistic and
statistic for testing the equality of the means of several groups of count data in the presence of a common dispersion parameter. These tests are generalized to be applicable without the restriction of a common dispersion parameter. And the assumed model of data is also extended from negative binomial to double exponential Poisson model. Monte Carlo simulations show the superiority of
statistic based on the double exponential Poisson family which has a very simple form and requires estimates of the parameters only under the null hypothesis.
Model identification of spatial autoregressive data analysis
Korean Journal of Applied Statistics, volume 10, issue 1, 1997, Pages 121~136
Spatial data is collected on a regular Cartesian lattice. In this paper we consider the model indentification of spatial autoregressive(SAR) models using AIC, BIC, pattern method. The proposed methods are considered as an application of AIC, BIC, 3-patterns for SAR models through three directions; row, column and diagonal directions. Using the Monte Carlo simulation, we test the efficiency of the proposed methods for various SAR models.
Log-density estimation based on a Fourier expansion
Korean Journal of Applied Statistics, volume 10, issue 1, 1997, Pages 137~149
In this paper we propose a logdensity estimation based on a Fourier expansion. The basis functions consisting of trigonometric functions are determinded by stepwise addition and deletion and the Bayes Information Criterion, where the maximum likelihood method is used to estimate the parameters. Numericla examples using real data and simulated data are provided to show the performance of proposed method.
Bayesian analysis of cumulative logit models using the Monte Carlo Gibbs sampling
Korean Journal of Applied Statistics, volume 10, issue 1, 1997, Pages 151~161
An easy Monte Carlo Gibbs sampling approach is suggested for Bayesian analysis of cumulative logit models for ordinal polytomous data. Because in the cumulative logit model the posterior conditional distributions of parameters are not given in convenient forms for random sample generation, appropriate latent variables are introduced into the model so that in the new model all the conditional distributions are given in very convenient forms for implementation of the Gibbs sampler.
Generation of Linear Trend-free block designs
Korean Journal of Applied Statistics, volume 10, issue 1, 1997, Pages 163~175
Randomization of the run order within a block is a technique commonly employed by the experimenters of block designs to avoid biases in the estimates of the effects of interest. In practice, however the experimental responses are sometimes affected by the spatial or temporal position of the experimental units within a block. In such cases, it is preferable to use a systematic ordering of the treatments. It is often possible to find an ordering which will allow the estimation of treatment effects independently of any trend is known as a trend-free block designs. In many idustrial and agricultural experiments, treatments are applied to experimental units sequentially in time or space. This paper begins with a review of concepts and properties of trend-free designs. We, then devise algorithms to generate linear trend-free designs. We extend and modify the existing algorithm which is given by Bradley and Odeh(1988). Also, the algorithm which generate all possible linear trend-free designs in provided.
Power analysis of testing fixed effects with two way classification
Korean Journal of Applied Statistics, volume 10, issue 1, 1997, Pages 177~187
This article considers the power performance of the tests in unbalanced two way mixed linear models with one fixed factor. The generalized least squares (GLS) F statistic testing no differences among the effects of the levels of the fixed factor is estimated using Henderson's method III, minimum norm quadratic unbiased estimator (MINQUE) with prior guess 1, maximum likelihood (ML) and resticted maximum likelihood (REML). We investigate the power performance of these test statistics. It can be shown, through simulation, that the GLS F statistics using four estimators produce similar type I error rates and power performance.
A new measure of tracking in repeated measurement data
Korean Journal of Applied Statistics, volume 10, issue 1, 1997, Pages 189~201
The primary purpose of this study is to develop a measure of tracking by using a modified kappa statistic. Understanding tracking phenomena in epidemiologic studies is quite important, because precautionary measure can be made in the early stage of the outcome event. Several authors proposed measures of tracking. Among them we compared ours against McMahan's using a simulation study. Finally we applied our procedure and McMahan's to real data. We may conclude that our statistic is adequate in explaining and detecting the tracking phenomenon.