• Communications for Statistical Applications and Methods
• /
• 제13권1호
• /
• pp.19-27
• /
• 2006

Outliers in survey are a perennial problem for applied survey statisticians to estimate the total or mean of population. The influence of outliers is more increasing as they have large weights in survey sampling. Many techniques have been studied to lower the impact of outliers on sample survey estimates. Outliers can be downweighted by winsorization or reducing the weight of outliers. The weight reduction is more reasonable than replacing one outlier by one value of non-outliers, because it has at least one unit. In this paper, we suggest the square root transformation of weight as the weight reduction method. We show this method is efficient with real data, and it's also easy to apply in practical affairs.

• 응용통계연구
• /
• 제21권3호
• /
• pp.399-411
• /
• 2008

현대 사회에서는 조직의 운영 및 의사 결정을 위하여 다양한 통계 조사가 실시되고 있으며, 연구의 목적에 따라 조사 문항을 다르게 하여 실시하고 있다. 현재 경상남도의 경우 3년 주기로 매년 설문 문항을 다르게 하여 사회 지표 조사를 실시하고 있어 유기적인 분석이 가능하지 못한 실정이다. 이에 본 장에서는 데이터 퓨전을 이용하여 다양한 통계 조사 자료를 결합하여 고부가적인 자료를 생성하고자 한다. 데이터 퓨전을 통해서 얻은 최종 결과에 대한 추가된 정보를 이용함으로써 통계 분석의 질을 향상시킬 수 있는 방법이므로, 데이터 퓨전에 의해서 얻어진 정보를 효율적으로 분석하는 것 또한 중요하다. 이에 본 논문에서는 통계 조사 자료에 대하여 데이터 퓨전을 실시하고, 데이터 퓨전에 의해 생성된 자료에 대하여 하이브리드 데이터마이닝 기법인 잠재변수를 이용한 신경망 분석을 적용하는 방안에 대하여 연구하고자 한다.

• Communications for Statistical Applications and Methods
• /
• 제8권2호
• /
• pp.375-382
• /
• 2001

The agricultural machine estimated sales survey is a survey to estimate annual sales quantities of eight major agricultural machines such as tracter, combine, etc. The purpose of this study is to design a multipurpose sample for the agricultural machine estimated sales survey. Main achievements of this study are to present an efficient stratification criterion and to suggest a reasonable estimation method by using the concept of post-stratification.

• Communications for Statistical Applications and Methods
• /
• 제10권3호
• /
• pp.707-718
• /
• 2003

The Korean Anthropometric Survey (Size Korea) is a sample survey which estimates on percentiles of several dimensional measurements of the human body and its component parts. The purpose of this study is to design a sample, which is designed on the base of 1997 survey database. Two different methods are considered to get the sample size for estimating the 5th and 95th percentile of body dimensions of Korean age range 0-80 years.

• The Plant Pathology Journal
• /
• 제38권3호
• /
• pp.175-181
• /
• 2022

Statistical analysis of data is an integral part of research projects in all scientific disciplines including the plant pathology. Appropriate design, application and interpretation of statistical analysis are also, therefore, at the center of publishing and properly evaluating studies in plant pathology. A survey of research works published in the Plant Pathology Journal, however, cast doubt on high standard of statistical analysis required for scientific rigor and reproducibility in the journal. Here I first describe, based on the survey of published works, what mistakes are commonly made and what components are often lacking during statistical analysis and interpretation of its results. Next, I provide possible remedies and suggestions to help guide researchers in preparing manuscript and reviewers in evaluating manuscripts submitted to the Plant Pathology Journal. This is not aiming at delineating technical and practical details of particular statistical methods or approaches.

• Nuclear Engineering and Technology
• /
• 제53권5호
• /
• pp.1664-1675
• /
• 2021

In order to verify the statistical performance of the nonparametric tests used in the MARSSIM approach, all plausible contamination distribution types that can be encountered in a survey area should be investigated. As the first of such investigations, this study aims to perform the verification for normal distribution of the contamination in a survey area by simulating the collection of random samples from it through the Monte Carlo simulation. The results of the simulations conducted for a total of 81 simulation cases showed that Sign test and WRS test both exhibited an excellent statistical performance: 100% for the former and 98.8% for the latter. Therefore, in final status surveys of the MARSSIM approach, a high statistical performance can be expected in applying the nonparametric hypothesis tests to survey areas whose net contamination can be assumed to be normally distributed.

• Journal of the Korean Data and Information Science Society
• /
• 제16권4호
• /
• pp.849-860
• /
• 2005

This Paper analyzed activities of the average time use survey in 1999 and 2004 by Korea National Statistical Office. The survey of National Statistical Office considered general analysis for activities like that personal care(eating), employment and study, leisure, travel by purpose. Change of time of living causes a lot of effects on society whole economy, culture etc., and is used basic data in policy proposal of the country and enterprise's marketing operation establishment. Analysis for average time could look at change in a role of gender, holiday of saturday's enforcement and increasing of temporary post. Using this analysis we can do more developed and analytical sociological study.

• Communications for Statistical Applications and Methods
• /
• 제10권2호
• /
• pp.319-324
• /
• 2003

In this paper, we review three cards method by Droitcour et al.(2001). And we compare indirect survey methods - randomized response technique, item count method, nominative technique.

• Journal of the Korean Statistical Society
• /
• 제31권4호
• /
• pp.469-482
• /
• 2002

This paper suggests a new method of compensating for wave nonresponse in panel survey, which combines weighting adjustment and imputation. By deleting less frequent nonresponse patterns, we can get simplicity. A new mean estimator under the new combining method is provided and a limited simulation study employing a real data is conducted.

• Communications for Statistical Applications and Methods
• /
• 제8권3호
• /
• pp.731-737
• /
• 2001

In this paper, we suggest and implement an internet survey system cooperating the randomized response technique(RRT). We compare it with the direct survey. RRT is an indirect survey method to get true information from the respondent who is conceived to have sensitive character without revealing his/her status. We implement our system(method) based a data-base system, so it common all kind of data obtained through internet surveys. This system also can be used in spot survey independently.