• Communications for Statistical Applications and Methods
• /
• v.11 no.2
• /
• pp.247-254
• /
• 2004

Tusher, Tibshirani and Chu (2001) suggested SAM (Significance Analysis of Microarrays) to compare two groups under different conditions for each gene, using microarray data. They used two sample t-statistic adding fudge factor in the denominator to prevent the value of statistic from being inflated by large sample variance, which might result in significant difference despite of a small value in the numerator. This paper aims at finding robust fudge factor and replacing it in two-sample t-statistic used in SAM, which we call Modified SAM (MSAM). Using the simulated data and data used in Dudoit et al.(2002), it is shown that MSAM find significant genes better and has less error rate than SAM.

• Proceedings of the Korean Association for Survey Research Conference
• /
• 2006.12a
• /
• pp.67-83
• /
• 2006

In successive sampling on two occasions the problem of estimating a finite population quantile has been considered. The theory developed aims at providing the optimum estimates by combining (i) three double sampling estimators viz. ratio-type, product-type and regression-type, from the matched portion of the sample and (ii) a simple quantile based on a random sample from the unmatched portion of the sample on the second occasion. The approximate variance formulae of the suggested estimators have been obtained. Optimal matching fraction is discussed. A simulation study is carried out in order to compare the three estimators and direct estimator. It is found that the performance of the regression-type estimator is the best among all the estimators discussed here.

• Journal of Integrative Natural Science
• /
• v.11 no.3
• /
• pp.154-160
• /
• 2018

Consider the problem of estimating a $p{\times}1$ mean vector ${\theta}(p-r{\geq}3)$, r = rank(K) with a projection matrix K under the quadratic loss, based on a sample $Y_1$, $Y_2$, ${\cdots}$, $Y_n$. In this paper a James-Stein type estimator with shrinkage form is given when it's variance distribution is specified and when the norm ${\parallel}{\theta}-K{\theta}{\parallel}$ is constrain, where K is an idempotent and symmetric matrix and rank(K) = r. It is characterized a minimal complete class of James-Stein type estimators in this case. And the subclass of James-Stein type estimators that dominate the sample mean is derived.

• Journal of Korean Society for Quality Management
• /
• v.22 no.2
• /
• pp.89-97
• /
• 1994

We consider the linear calibration model: $y_1={\alpha}+{\beta}x_i+{\sigma}{\varepsilon}_i$, i = 1, ${\cdots}$, n, $y={\alpha}+{\beta}x+{\sigma}{\varepsilon}$ where ($y_1$, ${\cdots}$, $y_n$, y) stands for an observation vector, {$x_i$} fixed design vector, (${\alpha}$, ${\beta}$) vector of regression parameters, x unknown true value of interest and {${\varepsilon}_i$}, ${\varepsilon}$ are mutually uncorrelated measurement errors with zero mean and unit variance but otherwise unknown distributions. On the basis of simple small-sample low-noise approximation, we introduce a new method of comparing the mean squared errors of the various competing estimators of the true value x for finite sample size n. Then we show that a class of estimators including the classical and the inverse estimators are consistent and first-order efficient within the class of all regular consistent estimators irrespective of type of measurement errors.

• Journal of the Korean Statistical Society
• /
• v.33 no.3
• /
• pp.339-351
• /
• 2004

In this paper we generalize and improve the refined Pickands estimator of Drees (1995) for the extreme value index. The finite-sample performance of the refined Pickands estimator is not good particularly when the sample size n is small. For each fixed k = 1,2,..., a new estimator is defined by a convex combination of k different generalized Pickands estimators and its asymptotic normality is established. Optimal weights defining the estimator are also determined to minimize the asymptotic variance of the estimator. Finally, letting k depend upon n, we see that the resulting estimator has a better finite-sample behavior as well as a better asymptotic efficiency than the refined Pickands estimator.

• Journal of the Korean Statistical Society
• /
• v.36 no.4
• /
• pp.543-556
• /
• 2007

In successive sampling on two occasions the problem of estimating a finite population quantile has been considered. The theory developed aims at providing the optimum estimates by combining (i) three double sampling estimators viz. ratio-type, product-type and regression-type, from the matched portion of the sample and (ii) a simple quantile based on a random sample from the unmatched portion of the sample on the second occasion. The approximate variance formulae of the suggested estimators have been obtained. Optimal matching fraction is discussed. A simulation study is carried out in order to compare the three estimators and direct estimator. It is found that the performance of the regression-type estimator is the best among all the estimators discussed here.

• Communications for Statistical Applications and Methods
• /
• v.13 no.2
• /
• pp.327-341
• /
• 2006

Winkler (1990, 2001), Sitter and Skinner (1994), Wilson and Sitter (2002) present a method which applies linear programing to designing surveys with multi-way stratification, primarily in situation where the desired sample size is less than or only slightly larger than the total number of stratification cells. A comparison is made with existing methods both by illustrating the sampling schemes generated for specific examples, by evaluating sample mean, variance estimation, and mean squared errors, and by simulating sample mean for all methods. The computations required can, however, increase rapidly as the number of cells in the multi-way classification increase. In this article their approach is applied to multi-way stratification using real data.

• Journal of the Korean Society of School Health
• /
• v.10 no.1
• /
• pp.51-63
• /
• 1997

Changes in disease patterns increase the importance of health-promoting lifestyles in a healthy society. Thus health education in elementary schools is very important because the knowledge of health and health habits in this period become a habitual part of life. The purpose of this study is to identify the performance in promoting healthy lifestyles of elementary school nurses. Such data helps to provide data to judge whether they are capable health educators or not. The subjects were 365 elementary school nurses selected by purposive sample. The data were collected by a self reporting questionnaire from August to October of 1996. The methods used in this study were the health-promoting lifestyle profile developed by Park In Sook (1995) and Licker's flour-point scale. A Cronbach's ${\alpha}$, percentage, mean, standard deviation, ANOVA, Pearson's correlation coffcient and stepwise multiple regression in the SAS package were used to analyze the data. The results of this study were as follows: 1. 37.7% of the sample were aged between 30 and 39 and 33.7% were between 40 and 49. 44.9% of the sample had attended college. 87.7% of the sample were married. 42.5% of the sample were people with a career of 11~20 years, and 26.0% of it had 21 years or more of working experience as an elementary school nurses. 2. The average scores of the family function and the social support were 2.97 and 2.98 respectively. 3. 86.5% of the sample were satisfied with their job. 4. 85.8% of the sample answered that they were healthy and 14.2% answered that they were unhealthy. The average scores of the internal health locus of control and self-esteem were 3.35 and 3.15 respectively. 5. The average score of performance in health-promoting lifestyle variables was 3.16; the average scores for harmonious relationships, regular diet professional health maintenance, sanitary life, self-control, emotional support, diet, rest and sleep, exercise and activity. self-achievement, and diet control were 3.47, 3.30, 2.52. 3.60, 2.92, 3.18, 3.14, 3.11, 2.96, 3.26 and 3.12. The variable with the highest degree of correlation was a sanitary life, whereas the one with the lowest degree was professional health maintenance. 6. A significant difference was found in self-esteem according to age. There was a significant difference in the average score of internal health locus of control according to age and career. 7. There were significant differences in the health-promoting lifestyle performance depending on age, career and monthly household income. The longer the career, the higher the health-promoting lifestyle performance. 8. A weak positive correlation was found between self-esteem and health-promoting lifestyles performance(r=0.417, p<0.001) Also, a weak positive correlation was found between the internal health locus of control and health-promoting lifestyles performance (r=0.386, p<0.001). 9. Heath-promoting lifestyle performance showed significant correlations with family function, social support and job satisfaction. 10. Stepwise multiple regression analysis revealed that the most powerful predictor was the variance of family function. Social support, age, self-esteem, internal health locus of control, perceived heath status and job satisfaction accounted for 55.9% of the variance in heath-promoting lifestyle performance.

• Journal of the Korean Data and Information Science Society
• /
• v.21 no.4
• /
• pp.669-679
• /
• 2010

A sample survey design is suggested for the service satisfaction evaluation of regional education offices based on the sample size of 2009 Gyeongnam regional education offices's customer satisfaction survey. The sample design is developed to fit the goal of evaluation of individual regional offices and allocate at least the minimum sample size to each city or county in Gyeongnam to achieve the goal of the survey. The population is stratified according to the regions and the types of schools, and the sample of schools is selected with proportional to the size of classes within each stratum. Finally, each sample student is selected according to two-stage cluster sampling within each sample school. Weighting averages, weighting totals and so on can be evaluated for analysis purposes. Their variance estimates can be evaluated using re-sampling methods like BBR, Jackknife, linearization-substitution methods, which are generally used for the data from a complex sample.

• Tribology and Lubricants
• /
• v.1 no.1
• /
• pp.38-45
• /
• 1985

Ferrograms prepared from off samples collected during testing in the transition region were originally diluted at 20:1. To obtain some information about the effect of dilution on the analysis procedures, a series of measurements were made on ferrograms prepared to different dilutions in the range 6 to 30:1 from oil samples collected after testing in the four ball machine at the 51 kg and 55 kg load, respectively, Fig. 1. The variations in area covered, perimeter, intercept and particle count were then plotted as a function of dilution level and appropriate mathematical expressions established such that the results obtained at any dilution level specified within the range can then be corrected back to an equivalent undiluted value. The effect of dilution on the variance of the particle size distribution was also investigated. The main results are tabulated, Tables 1-5 and also plotted as a function of dilution, level Figs. 2-9.