• The Korean Journal of Applied Statistics
• /
• v.11 no.2
• /
• pp.403-413
• /
• 1998

Sample size determination is a crucial part of sampling design. This paper gives a assessment for sample size determination using two-stage sampling scheme for estimating the population mean with a given precision.

• Survey Research
• /
• v.8 no.1
• /
• pp.55-79
• /
• 2007

This article attempts to provide an overview of the exit polls for the 2006 local elections in Korea. The sampling method, sampling error, non-response rate, and prediction error of the exit polls are reviewed. Also, we explore the fact that the propensity to vote varies according to age and gender of voters. In terms of age and gender, the representativeness of the sample is investigated by comparing to the data released by the National Election Commission. Through this empirical research, we show that the exit poll samples are unbalanced in terms of age and this unbalance may be one of the causes of bias occurred in the prediction of the 2006 local election results. The design effects of the sample design implemented for the exit polls are also examined.

• Survey Research
• /
• v.5 no.1
• /
• pp.103-130
• /
• 2004

Main monthly household surveys conducted by Korea National Statistical Office are economically active population survey(EAPS) and household income and expenditure survey(HIES). Samples of these two surveys are redesigned every 5 years based on Census. This paper is about sample redesign of household survey conducted in 2002 based on 2000 Census. Main improvements of 2002 sample redesign are the introduction of rotation sampling system, the expansion of HIES survey area from urban to whole country and the foundation of basement to make small area estimation for the unemployment statistics. Also the number of sample households within a enumeration district(ED) is reduced from 24 to 20. That makes it possible to select more ED samples which provides better precision for EAPS and HIES. To select representative samples for the population, different classification index is used for each metropolitan area and provinces.

• The Korean Journal of Applied Statistics
• /
• v.33 no.2
• /
• pp.215-228
• /
• 2020

Under Article 19, Paragraph 1 of the Food Industry Promotion Act, Rural Development Administration renews the Korean foods composition table every five years. Before the publication of the tenth revision of the Korean foods composition table in 2021, this paper suggests methods for collecting representative samples of 182 highly consumed foods in Korea. Food markets are categorized by their distribution channels, which are supermarkets and local markets. Eight samples are collected from each category by applying the National Food and Nutrient Analysis Program (NFNAP)'s stratified multi-stage sampling. The NFNAP was implemented in 1997 as a collaborative food composition research effort between the National Institute of Health (NIH) and the US Department of Agriculture (USDA) to secure reliable estimates for the nutrient content of food and beverages consumed by the US population. Selected supermarkets for selecting representative food samples are Emart Kayang, Homeplus Siheung, Lottemart Dongducheon, Emart Suwon, Lottemart Dunsan, Lottemart Yeosu, Emart Ulsan, and Hanaroclub Ulsan. Selected local markets also are Doksandongusijang in Geumcheon-gu and Pungnapsijang in Songpa-gu, Seoul, Ilsansijang in Ilsanseo-gu, Goyang, Unamsijang in Buk-gu, Gwangju, Beopdongsijang in Daedeok-gu, Daejeon, Bongnaesijang in Yeongdo-gu and Jwadongjaeraesijang in Haeundae-gu, Busan, and Jungangsijang in Jinhae-gu, Changwon.

• Survey Research
• /
• v.12 no.1
• /
• pp.153-174
• /
• 2011

This research aims to discuss the causes of inaccuracy in public opinion polls currently conducted in Korea. In particular, identifying the problems in sampling frame and sampling methods in political and social public opinion polls is an important question. Currently, most polling organizations operating in Korea are using phone number directories provided by Korea Telecom(KT) as its sampling frame for conducting most political polls. A critical problem of using a phone number directory as a sampling frame is that unlisted phone numbers can never be included in the sample. If a systematic difference in socio-demographic or politico-economic characteristics exists between the listed number using group and the unlisted group, using a phone number directory as a sampling frame cannot produce a sample that can represent the whole adult population in Korea. According to the poll result commissioned by the Asan Institute for Policy Studies in January 2011, there are statistically significant differences in socio-demographic and politico-economic characteristics between those two groups, and those differences led to the differences in the presidential job approval rating and party support. Our findings include that the listed number using group is more pro-Grand National Party and show stronger support for the president than the unlisted group.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.6 no.1
• /
• pp.25-34
• /
• 1988

Digital Terrain Model is composed of two basic elements, namely, sampling and interpolation. They represent the procurement of data and the geometric reconstruction of terrain relief respectively. For evaluation DTM data, following two paremeters appear to be the most important factor, that is, the accuracy and efficiency and, in particular, the break point information significantly affect to the accuracy of DTM data. The main objective of this study is to improve the accuracy and efficiency of DTM by applying Progressive Sampling with distinct morphologic information. In this study, the total of 240 individual numerical tests has been implemented and the appropriate computer program is also developed for the test. The result of investigation shows that the Progressive Sampling with break point information improves the accuracy of DTM by 30 percent approximately.

• Proceedings of the Korean Society for Information Management Conference
• /
• 1998.08a
• /
• pp.227-230
• /
• 1998

문헌정보학분야의 용어사전을 이용한 자동시소러스 구축을 위한 첫단계로$\ulcorner$문헌정보학 용어사전$\lrcorner$ MRD를 구성하고 이를 이용하여 상위어 자동 추출알고리즘을 개발하였다. MRD구성시 전처리과정을 통하여 상위어 추출에 불필요한 정보가 수록되는 것을 방지하였다. 상위어 추출을 위한 알고리즘 개발은 무작위 표본추출을 통하여 $\ulcorner$문헌정보학 용어사전$\lrcorner$에 기술된 문장의 구문적 특성을 분석한 후, 이 구문정보를 이용하여 수행하였다. 본 연구에서 제시된 알고리즘의 효율성 평가결과 89.4%의 정확도를 보였다.

• The Korean Journal of Applied Statistics
• /
• v.26 no.2
• /
• pp.209-222
• /
• 2013

We apply stratified cluster sampling to a replicated systematic unrelated question model for a large scale survey in which the population is comprised of several strata developed by several clusters and with sensitive parameters. We first present a replicated systematic unrelated question model using an unrelated question model to procure sensitive information from the population of clusters and then develop a suggested model to an unrelated question by a stratified cluster replicated systematic sampling that can be used in large population of strata. We cover the proportional and optimum allocation for the suggested model. Finally, we compare and analyze the efficiency of the suggested model with the replicated systematic unrelated question model.

• The Korean Journal of Applied Statistics
• /
• v.13 no.2
• /
• pp.207-224
• /
• 2000

Generally cluster size is predetermined when we use the stratified two-stage cluster sampling But in case that the sizes of clusters vary greatly one may want to make the sizes to be about equal. In this paper we study the optimal cluster size in stratified twostage cluster sampling. Also we find the optimal primary sampling unit sizes and optimal secondary sampling unit sizes under the given cost restriction.

• The Korean Journal of Applied Statistics
• /
• v.22 no.6
• /
• pp.1289-1300
• /
• 2009

The double sampling scheme is effective in reducing the sampling cost. However, the doubly sampled data is contaminated by two types of error, namely false-positive and false-negative errors. These would make the statistical analysis more difficult, and it would require more sophisticate analysis tools. For instance, the Wald method for the interval estimation of a proportion would not work well. In fact, it is well known that the Wald confidence interval behaves very poorly in many sampling schemes. In this note, the property of the Wald interval is investigated in terms of the coverage probability and the expected width. An alternative confidence interval based on the Agresti-Coull's approach is recommended.