• 한국통계학회:학술대회논문집
• /
• 한국통계학회 2001년도 추계학술발표회 논문집
• /
• pp.143-148
• /
• 2001

A dataset having missing observations is often completed by using imputed values. In this paper the performances and accuracy of complete case methods and four imputation procedures are evaluated when missing values exist only on the response variables in the Weibull regression model. Our simulation results show that compared to other imputation procedures, in particular, hotdeck and Weibull regression imputation procedure can be well used to compensate for missing data. In addition an illustrative real data is given.

• 한국통계학회:학술대회논문집
• /
• 한국통계학회 2005년도 추계 학술발표회 논문집
• /
• pp.13-17
• /
• 2005

In recent years, many people use R as a statistics system. R is frequently updated by many R project teams. We are interested in the method of multiple outlier detection and know that R is not supplied the method of multiple outlier detection. In this talk, we review these procedures for detecting multiple outliers and provide more efficient procedures combined with direct methods and indirect methods using R.

• Journal of the Korean Statistical Society
• /
• 제14권1호
• /
• pp.18-28
• /
• 1985

Optimal screening procedures for improving outgoing product quality based on correlated normal variables are presented. The performance variable and the screening variables are assumed to be jointly normally distributed. These procedures do not require specialized tables, and closed-form solutions are obtained for the case of one-sided specification. Methods for finding optimal solutions for the case of two-sided specifications are also considered.

• Communications for Statistical Applications and Methods
• /
• 제4권3호
• /
• pp.775-785
• /
• 1997

Robustness rates for repeated measures pairwise multiple comparison procedures were investigated in a split plot design with one between- and one within-subjects factor using untrimmed and trimmed data. Five factors were manipulated in the study: distribution, sphericity, variance-covariance heteroscedasticity, total sample size, and sample size ratio. The Welsch test (W) and the Welsch test on trimmed data $(W_{RT})$ performed better than the other procedures, but had a liberal tendency. The trimmed difference score Bonferroni Procedure $(B_{DT})$ was a good choice in some conditions.

• Communications for Statistical Applications and Methods
• /
• 제1권1호
• /
• pp.41-51
• /
• 1994

We consider nonparametric procedures for the selection and ranking problems. Tukey's generalized lambda distribution is condidered as the distribution for the score function because the distribution can approximate many well-known contionuous distributions. Also we compare these procedures in terms of efficiency, defined by the ratio of a probability of a correct selection divided by the expected selected subset size.

• Journal of the Korean Statistical Society
• /
• 제11권1호
• /
• pp.45-58
• /
• 1982

The problem of comparing k normal populations with a control (or a standard) in terms of the absolute values of their means is considered. Under the framework of indifference-zone formulation a single-state and a two-stage procedures for selecting the best are proposed, according to their commom vairances known or unknown respectively. The procedures guarantee that the probability of correct selection is not less than some preassigned lower limit. Selected tables necessary to implement the procedures are provided.

• 호남수학학술지
• /
• 제29권1호
• /
• pp.83-100
• /
• 2007

Normality is one of the most common assumptions made in sampling and statistical inference procedures without suffering from lack of attention. Its results may lead to an invalid conclusion. We present several testing procedures that can be used to evaluate the effects of departure from normality using concrete examples by hand or with the aid of Minitab. The goal is to influence prospective teachers in order to learn statistical concepts and techniques for testing normality on the basis of the didactical theory.

• 한국데이터정보과학회:학술대회논문집
• /
• 한국데이터정보과학회 2006년도 추계 학술발표회 논문집
• /
• pp.43-48
• /
• 2006

In 2001 US FDA proposed a draft guidance for future in vivo bioequivalence studies. The guidance suggested specific criteria for new drug sponsors to show prescribability and switchability in bioequivalence testing for approval of generic drugs. However, there is less acceptance of the need to change statistical procedures and study designs from those currently used to assess the current criterion of average bioequivalence. The measures of population and individual bioequivalence testing are introduced and statistical procedures for them are discussed.

• Genomics & Informatics
• /
• 제14권4호
• /
• pp.187-195
• /
• 2016

In genetic association studies with high-dimensional genomic data, multiple group testing procedures are often required in order to identify disease/trait-related genes or genetic regions, where multiple genetic sites or variants are located within the same gene or genetic region. However, statistical testing procedures based on an individual test suffer from multiple testing issues such as the control of family-wise error rate and dependent tests. Moreover, detecting only a few of genes associated with a phenotype outcome among tens of thousands of genes is of main interest in genetic association studies. In this reason regularization procedures, where a phenotype outcome regresses on all genomic markers and then regression coefficients are estimated based on a penalized likelihood, have been considered as a good alternative approach to analysis of high-dimensional genomic data. But, selection performance of regularization procedures has been rarely compared with that of statistical group testing procedures. In this article, we performed extensive simulation studies where commonly used group testing procedures such as principal component analysis, Hotelling's $T^2$ test, and permutation test are compared with group lasso (least absolute selection and shrinkage operator) in terms of true positive selection. Also, we applied all methods considered in simulation studies to identify genes associated with ovarian cancer from over 20,000 genetic sites generated from Illumina Infinium HumanMethylation27K Beadchip. We found a big discrepancy of selected genes between multiple group testing procedures and group lasso.

• 산업공학
• /
• 제6권1호
• /
• pp.55-65
• /
• 1993

A nonstandard instrument used in the filed frequently becomes out-of-calibration due to environmental noise, misuse, aging, etc. A substantial amount of loss may result if such nonstandard instrument is used to check product quality and performance. Traditional periodic calibration at the calibration center is not capable of detecting out-of-calibration status while the instrument is in use, and therefore, statistical methods need to be developed to check the status of a nonstandard instrument in the field between calibrations. Developed in this paper is a unified measurement assurance model in which statistical calibration at the calibration center and measurement assurance test in the filed are combined. We developed statistical procedures to detect changes in precision and in the coefficients of the calibration equation. Futher, computational experiments are conducted to evaluate how the power of test varies with respect to the parameters involved. Based upon the computational results we suggest procedures for designing effective measurement assurance tests.