The effect of range and frequency of attribute levels on the attribute importance in conjoint analysis: focused on traditional conjoint and self-explicated methods

컨조인트 속성수준의 분포 범위와 수가 응답자의 속성 중요도 평가에 미치는 영향: 전통적 컨조인트와 자체설명식 접근법 비교

  • Received : 2017.03.24
  • Accepted : 2017.06.30
  • Published : 2017.06.30

Abstract

The author introduce how the range and the number of attribute levels affect measures of attribute importance implemented by traditional conjoint analysis. We find that the relative importance of attribute are affected by the range manipulation and the number of attribute levels. The results show that the extension of both end-point attribute levels increases the importance of attribute. Also conjoint importance estimates increases by the insertion of intermediate attribute levels. However, the attribute importance estimated by self-explicated method does not vary with the variation of attribute level range and number of attribute levels. Therefore, these results imply that we need to select carefully the attribute levels of stimuli which the respondents are exposed to in purchasing situation.