• The Korean Journal of Applied Statistics
• /
• v.10 no.2
• /
• pp.367-374
• /
• 1997

In panel surveys it is necessary to manage both sampling frame and sample units across time. When sample is updated according to the change of its frame, it should be incorporated in the estimation procedure. This paper derives the bias of the conventional estimator caused by neglecting the change of sample, and provides a bias-adjusted estimator with its variance.

• Korea journal of population studies
• /
• v.29 no.3
• /
• pp.159-188
• /
• 2006

This study attempts to create a longitudinal dataset by linking tdata on the identical individuals across the monthly sample household management lists of the Economically Active Population Survey(EAPS) and the Household Income and Expenditure Survey(HIES). Using the data constructed through such process, the study also tryies to analyze the duration of longitudinal responses and the characteristics of nonrespondents. Between 1998 and 2002, longitudinal response rates had declined to 46% of total EAPS and 34% of total HIES. The fact that nonresponse was not a random phenomenon leads to concerns about the representativeness of the remaining sample. Using Cox's proportional hazard model the study revealed that the duration of longitudinal responses is affected by the ownership of house and the age of the respondent.

• The Korean Journal of Applied Statistics
• /
• v.22 no.5
• /
• pp.981-993
• /
• 2009

It would cause a serious problem in the panel data when panel attrition is concentrated on certain socioeconomic groups. Using the GOMS, this study investigates whether there exists non-random attrition bias in the data and seeks for feasible solutions to minimize the bias. The results of logit analyses show that panel attrition in the GOMS results mainly from surveying system but not from the surveyed. Therefore, the result suggests to develop well-organized management skill and systems as well as to construct weighting methods.

• The Korean Journal of Applied Statistics
• /
• v.24 no.1
• /
• pp.25-34
• /
• 2011

The transition probability can be used for the estimation of subpopulation total in panel data analysis. In this paper a real data analysis is performed and the sensitivity of the sample size allocated in the subpopulation is examined by small simulation studies.

• Survey Research
• /
• v.12 no.3
• /
• pp.173-186
• /
• 2011

We introduced the methodologies used to construct the longitudinal weights and cross-sectional weight that are required for the analysis of Korea Youth Panel Survey. To analyze the longitudinal dynamic change of the population, we derived the longitudinal weight through nonresponse adjustment based on logistic regression and post-stratification. Cross-sectional weights that are necessary to produce an asymptotically unbiased estimator of the population parameter were constructed through simple nonresponse adjustment based on overall response rate and post-stratification.

• Proceedings of the Korean Association for Survey Research Conference
• /
• 2011.10a
• /
• pp.273-285
• /
• 2011

본 논문에서는 패널조사와 같은 종단면 연구에서 시간의 흐름에 따라 패널의 노후화 등의 원인으로 각 조사주기별로 발생하는 무응답(결측)에 대해 특정한 패널집단을 대상으로 무응답 패턴을 통계모형을 이용하여 분석하였다. 이러한 무응답 패턴분석을 기반으로 결측자료가 존재하는 종단자료의 분석에서 적절한 방법을 선택하여 분석을 수행할수 있으며, 만일 무응답 대체가 필요한 경우 적절한 대체 방법을 결정할 수 있을 것이다. 횡단면 조사와는 달리 이용가능한 보조정보가 각 웨이브별로 다양하게 존재하며, 이와 같은 보조정보를 무응답 대체에 활용할수 있다면, 결측자료가 존재하는 패널 자료에 비해 전통적인 통계분석 방법을 적용하여 표준적인 결과를 산출할 수 있을 것으로 기대된다.

• Survey Research
• /
• v.11 no.1
• /
• pp.1-18
• /
• 2010

We compare various imputation methods for compensating wave nonresponse that are commonly adopted in many panel surveys. Unlike the cross-sectional survey, the panel survey is involved a time-effect in nonresponse in a sense that nonresponse may happen for some but not all waves. Thus, responses in neighboring waves can be used as powerful predictors for imputing wave nonresponse such as in longitudinal regression imputation, carry-over imputation, nearest neighborhood regression imputation and row-column imputation method. For comparison, we carry out a simulation study on a few income data from the Korean Welfare Panel Study based on two performance criteria: predictive accuracy and estimation accuracy. Our simulation shows that the ratio and row-column imputation methods are much more effective in terms of both criteria. Regression, longitudinal regression and carry-over imputation methods performed better in predictive accuracy, but less in estimation accuracy. On the other hand, nearest neighborhood, nearest neighbor regression and hot-deck imputation show higher performance in estimation accuracy but lower predictive accuracy. Finally, the mean imputation shows much lower performance in both criteria.

• Survey Research
• /
• v.6 no.2
• /
• pp.51-74
• /
• 2005

The internet survey methods have been more and more widely used as the coverage of the fixed-line telephone is being reduced due to the diffusion of mobile phone. So, there is a need to know the accuracy of this new survey method. This study aims to estimate the accuracy of the internet survey method and identify the factors affecting the accuracy of this method, For this purpose, we analyzed the election poll data during the 17th general election period. These data include fixed-line telephone survey data, internet survey data, mobile phone survey data and the election voting data. The analysis shows that the prediction errors of the internet survey were a little more than those of the telephone or mobile phone survey. But the differences are not significant. It follows from this result that we can use the internet survey method in social survey context. This study also found that the respondent's willingness to participate in the survey, the probability of being at home during survey and the respondent's educational level were affecting the accuracy of the internet survey. Further studies to develop weighting method with these factors are needed.

• Survey Research
• /
• v.11 no.1
• /
• pp.107-121
• /
• 2010

The purpose of this paper is to design and implement a Computer Assisted Personal Interviewing (CAPI) system for Panel Survey of Employment for the Disabled (PSED) using Blaise software developed by Statistics Netherlands which is a powerful and flexible system used for computer-assisted survey processing. CAPI system for PSED is consisted of three parts, which are survey management program, electronic questionnaire program, and postprocessing program. These programs are constructed with modular approach and are integrated into CAPI system. The developed software is highlighted in its reusability which results in the improved efficiency and reduced cost of software maintenance. CAPI system for PSED proved its effectiveness in carrying out the survey by guaranteeing correctness of survey data, improved timeliness of survey, easy reforming the questionnaires, easy control on the interviewers, and rapid data processing in spite of the complexity of PSED.

• Survey Research
• /
• v.10 no.2
• /
• pp.23-43
• /
• 2009

The purpose of this paper is to explain characteristics of panel data and display the academic cases using panel data. Panel data have merit to control various variables to influence respondents' attitude and estimate the effects of independent variables researchers are interested in by surveying the same respondents several times. However, panel data also have problems such as contaminating the respondents and reducing the number of the respondents as survey conducted several times. In spite of a few inherent problems of panel data, informations that panel data provide are valuable and some research themes could not be possible without panel data. This paper investigates the relationship between interest in elections and voting. Electoral stimulation such as watching TV debates influences the consistent voting intended. But changing nonvoters who had voting intention but not vote are not influenced by the electoral stimulation.