Korean Academic Society of Business Administration
[ Article ]
korean management review - Vol. 47, No. 1, pp.25-43
ISSN: 1226-1874 (Print)
Print publication date 28 Feb 2018
Received 31 Mar 2017 Revised 01 Dec 2017 Accepted 08 Dec 2017
DOI: https://doi.org/10.17287/kmr.2018.47.1.25

사례기반추론을 이용한 전통적 컨조인트 분석의 불완전한 응답 자료의 효용추정방법에 관한 연구

서상윤*
*(주저자) 경남대학교 경영학부 syseo@kyungnam.ac.kr
A Study on the Predicting Utility for Incomplete Response in the Traditional Conjoint Analysis Based on the Case Based Reasoning
Sang Yun Seo*
*School of Business Administration, Kyungnam University, First Author

초록

본 연구는 전통적 컨조인트 분석에서 서열척도로 상품 프로필을 평가할 때 발생하는 불완전한 응답 자료의 문제점을 사례기반추론을 이용하여 보완하고 그 응답자의 효용을 추정하는 방법에 대해 설명한다. 사례기반추론은 서열 변수들 간의 연관관계를 파악하는 켄달의 서열 상관계수와 스피어만의 서열 상관계수를 이용하여 추론에 활용할 자료를 검색하였다. 그리고 유사도가 높은 응답자의 속성 수준별 효용 값을 이용하여 불완전한 컨조인트 응답자료의 효용 값을 계산하고, 효용 값의 크기에 따라 불완전 응답 자료의 순위를 입력하여 효용을 추정하였다. 분석결과 불완전한 응답 자료와 추론에 사용된 자료 간의 유사성을 나타내는 상관계수 값이 클수록 자료의 복원을 통해 도출된 컨조인트 효용모형의 응답자 선택에 대한 예측정확도는 높아졌다. 그리고 유사성 검색에 사용된 방법 간에는 켄달의 서열 상관계수를 이용한 추론방법이 스피어만 서열 상관계수 보다 높은 예측정확도를 보였다. 끝으로 상관계수의 크기와 추론에 활용된 자료의 수는 상관계수의 크기와 상호작용하여 예측정확도를 더욱 높이는 것으로 나타났다.

Abstract

The conjoint analysis has been a representative methodology to find out consumer’s preference. When carrying on the conjoint analysis, the most challenge task for the subjects is to evaluate more than eight profiles. When the respondent evaluate the profiles based on a ranking scale, as the predictive validity was greater compared to rating, this ranking scale have been usually applied with traditional conjoint analysis. However, several mistakes by respondents occurred such as the omitted ranking and same ranking answers. These incomplete answers are exempted from estimating utilities step. However, this study tried to find out the utility of subject who incompletely responds in traditional conjoint analysis based on the case based reasoning. The Kendall’s and Spearman’s ranking correlation coefficient are applied to search the similar cases to complement the incomplete response, also the utilities are estimated based on the preference data evaluated in regular sequence. The result shows that the hit-ratio predicted by data with higher correlation coefficient is higher than it done by the data with lower correlation coefficient. Also, in terms of the reasoning methodology, the Kendall’s coefficient rather than the Spearman’s coefficient shows higher predictive accuracy. In addition, the interaction between correlation coefficient and the number of profiles has positive impact on the predictive accuracy.

Keywords:

case based reasoning, traditional conjoint, incomplete response, Kendall’s coefficient, Spearman coefficient

키워드:

사례기반추론, 전통적 컨조인트, 불완전 응답자료, 켄달의 서열 상관계수, 스피어만 서열 상관계수

Acknowledgments

이 논문 또는 저서는 2016년 대한민국 교육부와 한국연구재단의 지원을 받아 수행된 연구임(NRF-2016S1A5A8019608)

References

  • 서상윤·이훈영·남양호(2009), “상품 속성의 특성과 자료수집 방법이 컨조인트 모형의 효과에 미치는 영향에 관한 비교연구,” 마케팅관리연구, 14(1), 25-38.
  • 이재식·명훈식(2008), “사례기반추론을 이용한 인터넷 서점의 서적 추천시스템 개발,” 한국전자거래학회지, 13(4), 173-191.
  • 정영우·이은용(2008), “컨조인트 분석을 통한 푸드코트 선택 속성에 관한 탐색적 연구,” 한국조리학회지, 14(4), 106-118.
  • Aamodt, A., and E. Plaza(1996), “Case-Based Reasoning; Foundational Issues, Methodological Variations, and System Approaches,” Artificial Intelligence Communications, 4(3), 39-59. [https://doi.org/10.3233/AIC-1994-7104]
  • Chen, J., and J. Shao(2000), “Nearest Neighbor Imputation for Survey Data,” Journal of Official Statistics, 16, 113-131.
  • Curry, J.,(1997), “After the Basics : Keeping Key Issue in Mind Makes Conjoint Analysis Easier to Apply,” Marketing Research, 9(Spring), 6-11.
  • Eggers, F., and H. Sattler(2009), “Hybrid Individualized Two-level Choice-based Conjoint (HIT-CBC): A New Method for Measuring Preference Structures with Many Attribute Levels,” International Journal of Research in Marketing, 26(June), 108-118. [https://doi.org/10.1016/j.ijresmar.2009.01.002]
  • Green, P. E, S. M. Goldberg, and M. Montemayor (1981), “A Hybrid utility Estimation Model for Conjoint Analysis,” Journal of Marketing, 45(1), 33-41. [https://doi.org/10.1177/002224298104500104]
  • Green, P. E., and V. Srinivasan(1990), “Conjoint Analysis in Marketing : New Developments with Implications for Research and Practice,” Journal of Marketing, 54(Oct), 3-16. [https://doi.org/10.2307/1251756]
  • Green, P. E., and M. Krieger(1991), “Product Design Strategies for Target Marketing Positioning,” Journal of Product Innovation Management, 8(3), 189-202. [https://doi.org/10.1111/1540-5885.830189]
  • Green, P. E., and M. Krieger(1992), “An Application of a Product Positioning Model to Pharmaceutical Products,” Marketing Science, 11 (Feb), 117-132. [https://doi.org/10.1287/mksc.11.2.117]
  • Green, P. E., M. Krieger and Y. Wind(2001), “Thirty Years of Conjoint Analysis: Reflections and Prospects,” Marketing Engineering, 31(3), S56-73. [https://doi.org/10.1287/inte.31.3s.56.9676]
  • Green, P. E., and M. Krieger(2002), “What’s Right with Conjoint Analysis?” Marketing Research, Spring, 24-27.
  • Gustafsson, A., A. Herrmann, and F. Huber(2007), Conjoint Measurement: Methods and Applications 4th, Springer. 23. [https://doi.org/10.1007/978-3-540-71404-0]
  • Johnson, R. M.(1987), “Adaptive Conjoint Analysis,” in: Sawtooth Software Conference Proceedings, Ketchum, ID: Sawtooth Software, July, 253-65.
  • Johnson, R., J. Huber, and L. Bacon (2003), Adaptive Choice Based Conjoint Analysis, Sawtooth Software.
  • Kendall, M.(1938), “A New Measure of Rank Correlation,” Biometrika. 30(1–2), 81–89. [https://doi.org/10.2307/2332226]
  • Lenk, P. J., W. S. Sesarbo, P. E. Green, and M. R. Young (1996), “Hierarchical Bayes Conjoint Analysis: Recovery of Partworth Heterogeneity from Reduced Experimental Designs,” Marketing Science, 15(2), 173-191. [https://doi.org/10.1287/mksc.15.2.173]
  • Louviere, J. J., D. A. Hensher, and J. D. Swait(2000), Stated Choice Methods: Analysis and Applications in Marketing, Transportation and Environmental Valuation. Cambridge University Press, Cambridge. [https://doi.org/10.1017/CBO9780511753831]
  • Sande, G.(1979), “Numerical Edit and Imputation,” International Association for Statistical Computing, 42nd Session of the International Statistical Institute.
  • Spearman, C.(1904), “The Proof and Measurement of Association between Two Things,” American Journal of Psychology. 15, 72-101. [https://doi.org/10.2307/1412159]
  • Toubia, O., J. R. Hauser, and D. I. Simester(2004), “Polyhedral Methods for Adaptive Choice Based Conjoint Analysis,” Journal of Marketing Research, 41(1), 116-131. [https://doi.org/10.1509/jmkr.41.1.116.25082]
  • Vacek, P. M., and T. Ashikaga(1980), “An Examination of the Nearest Neighbor Rule for Imputing Missing Value,” Proceedings of the Statistical Computing Section, American Statistical Association, 326-331.

• 저자 서상윤은 현재 경남대학교 경영학부 마케팅 전공 조교수로 재직 중이다. 경희대학교에서 경영학 학사, 석사, 박사를 취득하였다. 박사 학위 취득 이후에는 San Diego State University 박사 후 연수과정과 경희대학교 경영연구원을 거쳐 경남대학교에 부임하였다. 주요연구분야는 Marketing Modeling, Marketing ROI, Segmentation, Conjoint 등이다.