• The Korean Journal of Applied Statistics
• /
• v.11 no.2
• /
• pp.257-268
• /
• 1998

Many epidemiological studies investigate whether an association exists between a binary risk factor X and a binary response variable Y. They analyse whether an observed association between X and Y persists when the level of another factor Z that might influence the association is controlled. This involves testing conditional independence of X and Y controlling for Z. The Mantel-Haenszel test is most widely used to test conditional independence for sparse tables. But if the association between X and Y varies along the levels of Z, Mantel-Haenszel test has a low power problem. In this study, we propose an alternative test procedure which overcomes the low power problem in that case. We find out the null distribution of the alternative test statistic and compare its performance with the Mantel-Haenszel test by simulation.

• Journal of the Korean Data and Information Science Society
• /
• v.21 no.6
• /
• pp.1081-1090
• /
• 2010

Most nations including Korean government make a great endeavor to realize low-carbon and green-growth world. We also work hard to expand bicycle facilities and bicycle road in order to increase bicycle transportation rate. Nowadays number of cyclists is increasing but fortunately, bicycle accidents also increase rapidly. Most data of bicycle accidents published by National Police Agency annually are represented as frequencies in two dimensional contingency tables. In this work, risk rates and characteristics of bicycle accidents are analyzed by using concepts of the probability and conditional probability. Especially with numbers of estimated cyclists and registered cars, risk rates of various kinds of bicycle accidents are obtained. Under the assumption of the conditional independence, probability of bicycle accident occurred at realistic situations could be estimated. Furthermore we discuss to reduce bicycle accidents with these results obtained in this work.

• Radiation Oncology Journal
• /
• v.26 no.4
• /
• pp.289-294
• /
• 2008

Purpose: To improve the quality of the statistical analysis of papers published in the Journal of the Korean Society for Therapeutic Radiology and Oncology (JKOSTRO) by evaluating commonly encountered errors. Materials and Methods: Papers published in the JKOSTRO from January 2006 to December 2007 were reviewed for methodological and statistical validity using a modified version of Ahn's checklist. A statistician reviewed individual papers and evaluated the list items in the checklist for each paper. To avoid the potential assessment error by the statistician who lacks expertise in the field of radiation oncology; the editorial board of the JKOSTRO reviewed each checklist for individual articles. A frequency analysis of the list items was performed using SAS (version 9.0, SAS Institute, NC, USA) software. Results: A total of 73 papers including 5 case reports and 68 original articles were reviewed. Inferential statistics was used in 46 papers. The most commonly adopted statistical methodology was a survival analysis (58.7%). Only 19% of papers were free of statistical errors. Errors of omission were encountered in 34 (50.0%) papers. Errors of commission were encountered in 35 (51.5%) papers. Twenty-one papers (30.9%) had both errors of omission and commission. Conclusion: A variety of statistical errors were encountered in papers published in the JKOSTRO. The current study suggests that a more thorough review of the statistical analysis is needed for manuscripts submitted in the JKOSTRO.

• The Korean Journal of Applied Statistics
• /
• v.17 no.3
• /
• pp.499-505
• /
• 2004

The Pareto diagram analysis of the experimental data by the two level orthogonal arrays has been used widely in practice since it is a graphical, quick and easy method to analyze experimental results, which does not use the analysis of variance to screen significant effects. For the analysis of the experimental data by $L_{18}(2 \times 3^7)$ orthogonal array, Park(1996) proposed Pareto ANOVA in which the size of effects is defined by the mean squares of effects and the Pareto principle is used. In this paper, a new approach of the Pareto diagram analysis of the experimental data by $L_{18}(2 \times 3^7)$ orthogonal array is proposed. The main idea is to partition the size of three level effects by that of linear and quadratic orthogonal contrasts of those effects.

• The Korean Journal of Applied Statistics
• /
• v.14 no.1
• /
• pp.39-57
• /
• 2001

Categorical data has sometimes misclassification errors. If this data will be analyzed, then estimated cell probabilities could be biased and the standard Pearson X2 tests may have inflated true type I error rates. On the other hand, if we regard wellclassified data with misclassified one, then we might spend lots of cost and time on adjustment of misclassification. It is a necessary and important step to ask whether categorical data is misclassified before analyzing data. In this paper, when data is misclassified at one of two variables for two-dimensional contingency table and marginal sums of a well-classified variable are fixed. We explore to partition marginal sums into each cells via the concepts of Bound and Collapse of Sebastiani and Ramoni (1997). The double sampling scheme (Tenenbein 1970) is used to obtain informations of misclassification. We propose test statistics in order to solve misclassification problems and examine behaviors of the statistics by simulation studies.

• Journal of the Korean Data and Information Science Society
• /
• v.11 no.1
• /
• pp.83-90
• /
• 2000

The chi-squared test statistic is usually employed for testing independence of two categorical variables in a two-way contingency table. It is well known that, under independence, the test statistic has an asymptotic chi-squared distribution under multinomial or product-multinomial models. For the case where both margins fixed, the sampling model of the contingency table is a multiple hypergeometric distribution and the chi-squared test statistic follows the same limiting distribution. In this paper, we study the difference between the small sample and large sample distributions of the chi-squared test statistic for the case with fixed margins. For a few small sample cases, the exact small sample distribution of the test statistic is directly computed. For a few large sample sizes, the small sample distribution of the statistic is generated via a Monte Carlo algorithm, and then is compared with the large sample distribution via chi-squared probability plots and Kolmogorov-Smirnov tests.

• Proceedings of the Korean Statistical Society Conference
• /
• 2001.11a
• /
• pp.175-178
• /
• 2001

다원분할표에서 범주들의 대응관계를 그래프적으로 보여주는 다중대응분석(multiple correspondence analysis)은 주결여성(principal inertia)이 총결여성(total inertia)에서 차지하는 비율이 전반적으로 낮아 설명력(goodness-of-fit)이 낮은 2차원의 대응분석그림을 얻게 된다. 이를 극복하기 위해 Benzecri의 공식을 사용하면 낮은 주결여성을 높이고 새로운 2차원 대응분석그림을 얻을 수 있다. 그러나 이 새로운 대응분석그림도 범주들의 대응관계를 명확히 보여주지는 못한다(Greenacre and Blasius, 1994, chapter 10). 앤드류 플롯(Andrews plot)을 이용하여 범주들의 군집화(clustering)로 다중대응분석을 재해석 하고자 하나 범주의 수가 많은 경우 해석상 어려움이 따른다. 본 소고에서 이와 같은 경우 K-평균 군집분석을 활용하여 다중대응분석의 해석을 용이하게 하고자 한다.

• The Korean Journal of Applied Statistics
• /
• v.7 no.2
• /
• pp.145-157
• /
• 1994

For multivariate distributions satisfying stochastic ordering, we suggest maximum likelihood estimation with incomplete data via an EM algorithm. In this paper we restrict our attention to the contingency tables with partially cross-classified observations. We may use the existing isotonic regression program to implement EM algorithm, and we illustrate the estimation process through an example.

• Journal of Korea Multimedia Society
• /
• v.8 no.10
• /
• pp.1347-1359
• /
• 2005

We propose an efficient block classification of the document images using the second-order statistical texture features computed from spatial gray level dependence matrix (SGLDM). We studied on the techniques that will improve the block speed of the segmentation and feature extraction speed and the accuracy of the detailed classification. In order to speedup the block segmentation, we binarize the gray level image and then segmented by applying smoothing method instead of using texture features of gray level images. We extracted seven texture features from the SGLDM of the gray image blocks and we applied these normalized features to the BP (backpropagation) neural network, and classified the segmented blocks into the six detailed block categories of small font, medium font, large font, graphic, table, and photo blocks. Unlike the conventional texture classification of the gray level image in aerial terrain photos, we improve the classification speed by a single application of the texture discrimination mask, the size of which Is the same as that of each block already segmented in obtaining the SGLDM.

• Survey Research
• /
• v.4 no.2
• /
• pp.47-62
• /
• 2003

Based upon the Labor Demand Survey and the Survey on Wage Structure being conducted annually by the Ministry of Labor, we suggest a convenient statistical tool which can analyze and appraise the effect of introducing a new survey method. Since both surveys have in common very similar sampling frames and sampling schemes in structure, one can measure the variability of one survey on the basis of the other. The influence appraisal method adopted here is applied to ratio data between two independent estimates belonging to the identical category by year and has a statistical form of comparing ratio data before introducing a new survey with those after introducing a new survey.