Journal of the Korean Data and Information Science Society

The Korean Data and Information Science Society (한국데이터정보과학회)
권호 표지
DOI QR코드

DOI QR Code

The unit-nonresponse status and use of weight in the KCYPS

한국아동·청소년패널조사자료에서 단위무응답의 실태 및 가중치 적용

  • Lee, Hwa-Jung (College of Ecology and Environmental Science, Kyungpook National University) ;
  • Kang, Suk-Bok (Department of Statistics, Yeungnam University)
  • Received : 2014.09.24
  • Accepted : 2014.10.27
  • Published : 2014.11.30
Download PDF
⟨ Previous Next ⟩

Abstract

Usually unit-nonresponse or item-nonresponse occurs in the survey. In case the rate of nonresponse is high, the analysis ignoring nonresponse may cause the wrong effect. The characterization of nonresponse is required. In a cross-sectional data, it is possible to study the characteristics of item-nonresponse but it is hard to study the characteristics of the unit-nonresponse. In order to identify the characteristics of the unit-nonresponse, this study used the first-year student of middle schools in the Korea children and youth panel survey (KCYPS) data. We investigated the handling situation of nonresponse and analyzed the characteristics of the unit-nonresponse. Most of the papers applied the way of getting rid of nonresponse, so that there was little paper using weights. In this paper, we compared the results of the analyses depending on whether the weight is used or not. The method of using weights showed statistically significant results much more than that of removing. More discussion will be needed.

설문조사에서는 대부분 단위무응답 또는 항목무응답이 발생한다. 무응답 발생비율이 높은 경우 무응답을 무시한 분석은 잘못된 결과를 초래할 수도 있으므로 무응답에 대한 특성 파악도 필요하다. 횡단자료에서는 항목무응답 특성에 대한 연구는 가능하나 단위무응답의 특성을 파악하기가 어렵다. 단위무응답의 특성을 파악하기 위하여 본 논문에서는 종단자료인 한국아동 청소년패널조사 중 중1패널자료를 이용하였다. 최근까지 공개된 한국아동 청소년패널조사자료를 이용한 논문들의 무응답 처리 실태를 파악하고, 그 중 단위무응답의 특성에 대하여 분석하였다. 이 자료를 이용한 논문에서는 대부분 단위무응답의 처리를 제거하는 방식을 취하고 있으며, 이 자료에서 제공하고 있는 가중치를 이용한 논문은 거의 없었다. 이에 단위무응답의 처리방법 중 가중치를 이용한 방법과 그 대상을 제외하는 두 가지 방법에 대해 여러 가지 특성들의 분석을 비교하였다. 많은 항목에서 두 가지 방법의 결과가 상이하게 나타났으며, 가중치를 이용한 경우가 제거법에 의한 결과에 비해 통계적으로 유의한 결과가 더 많이 나타나 무응답에 대해 많은 논의가 필요할 것으로 생각된다.

Keywords

References

  1. Chang, H. L. and Chung, I. J. (2013). The effects of parenting attitudes on adolescent prosocial behavior and aggression mediated by self-esteem. Korean Journal of Family Social Work, 40, 33-61.
  2. Chung, C. E. (2012). Complex sample design effects and inference for Korea National Health and Nutrition Examination Survey data. Journal of Nutrition and Health, 45, 600-612.
  3. Heo, S. and Chung, Y. (2012). Effect of complex sample design on Pearson test statistic for homogeneity. Journal of the Korean Data & Information Science Society, 23, 757-764. https://doi.org/10.7465/jkdi.2012.23.4.757
  4. Hirschi, T. and Gottfredson, M. R. (1993). Commentary: Testing the general theory of crime. Journal of Research in Crime and Delinquency, 30, 47-54. https://doi.org/10.1177/0022427893030001004
  5. Junger-Tag, J. and Marshall, I. H. (1999). The self-report methodology in crime research. In Crime and Justice; A Review of Research, edited by M. Torny, University of Chicago Press, Chicago.
  6. Kim, S. W. and Chung, L. J. (2012). Factors influencing item nonresponse and unit nonresponse in a youth panel survey. Studies on Korean Youth, 23, 27-50.
  7. Lee, H. G. (2008). Status analysis of Korea panel research (How to increase the accuracy and efficiency of the survey statistics: Based on the field survey and sampling), Statistical research institute, Daejeon.
  8. Lee, J. Y. (2005). Longitudinal study of household survey data (Statistical studies 05-03), Statistics Korea, Daejeon.
  9. National Youth Policy Institute (2013). 1st 3rd survey data user's guide in Korea childeren and youth panel survey(KCYPS), National Youth Policy Institute, Seoul.
  10. Park, C. and Choi H. S. (2014). A study of the factors influential on a health-related quality of life using complex sample design. Journal of the Korean Data & Information Science Society, 25, 829-846. https://doi.org/10.7465/jkdi.2014.25.4.829
  11. Shin, D. K. (1998). Status and implications of United States panel data - Focus on PSID, NLSY, KHPS, KLIPS, Korean Labor Institude, Seoul.
  12. Watkins, A. K. and Melde, C. (2007). The effec of self-control on unit and item nonresponse in an adolescent sample. Journal of Research in Crime and Delinquency, 44, 267-294. https://doi.org/10.1177/0022427807301675

Cited by

  1. The sample characteristics affecting longitudinal initial sample attrition in the Korea youth panel survey vol.28, pp.2, 2014, https://doi.org/10.14816/sky.2017.28.2.157