地理学报 ›› 2011, Vol. 66 ›› Issue (2): 267-278.doi: 10.11821/xb201102012

• 旅游地理 • 上一篇    下一篇

条件价值法中的偏差分析及信度和效度检验——以九寨沟游憩价值评估为例

董雪旺1, 张捷1, 刘传华2, 李敏1, 钟士恩1   

  1. 1. 南京大学国土资源与旅游学系,南京210093;
    2. 浙江省农业科学院农业信息研究所,杭州310021
  • 收稿日期:2010-02-21 修回日期:2010-09-23 出版日期:2011-02-20 发布日期:2011-03-31
  • 作者简介:董雪旺(1974-), 男, 山西临汾人, 博士研究生, 讲师, 中国地理学会会员(S110008124M), 主要从事旅游地理与 旅游规划的研究。E-mail: dxuewang@163.com
  • 基金资助:

    英国British Academy 项目(SG-47266); 国家自然科学基金项目(40371030)

Bias Analysis and Reliability and Validity Test in Contingent Valuation Method:A Case Study of Assessment of Jiuzhaigou's Recreational Value

DONG Xuewang1, ZHANG Jie1, LIU Chuanhua2, LI Min1, ZHONG Shi'en1   

  1. 1. Department of Land Resources and Tourism Sciences, Nanjing University, Nanjing 210093, China;
    2. Institute of Agricultural Information, Zhejiang Academy of Agricultural Sciences, Hangzhou 310021, China
  • Received:2010-02-21 Revised:2010-09-23 Online:2011-02-20 Published:2011-03-31
  • Supported by:

    British Academy, No.SG-47266; National Natural Science Foundation of China, No.40371030

摘要: 条件价值法(CVM) 是一种基于虚拟市场的公共物品价值评估方法,其假想性使得人们对 其信度和效度缺乏信任,评估结果的精度成为学术界广泛争议的焦点。采用CVM价值评估方 法,对九寨沟世界自然遗产地的游憩价值进行了评估,并结合国内其他相关文献,对CVM在旅 游资源价值评估中存在的偏差进行了分析述评,最后对CVM的评估结果进行了再测信度、内容 效度和收敛效度检验。研究发现:① 九寨沟旅游资源2008、2009 年的游憩价值分别为3.33 亿 元、3.46 亿元。与其他方法相比,CVM的评估结果明显偏低。② CVM在旅游资源价值评估中存 在的偏差包括研究方法本身的偏差和研究实施过程中产生的偏差。前者包括假想偏差、信息偏 差、抗议性偏差以及策略性偏差等;后者包括CVM调查在总体和样本的界定、样本规模的确定、 抽样方式的选择、调查方式和回收率、数据的统计和分析等环节出现的偏差。③ 再测信度检验 表明本研究的信度较高,即评估结果具有较高的稳定性和可重复性;而内容效度和收敛效度检 验则表明本研究对九寨沟进行的CVM调查不具有良好的效度,CVM有低估旅游资源价值的倾 向。④ CVM今后的研究重点应从个案研究向信度和效度检验转向,为此需要尽快建立完善适 合发展中国家的CVM实施规范,提高CVM评估结果的信度和效度。

关键词: 条件价值法(CVM), 旅游资源价值, 偏差, 信度, 效度, 九寨沟

Abstract: Contingent valuation method (CVM) is a valuation instrument for public goods on the basis of simulated market. However, its hypothesis arouses suspicions and controversies on its reliability and validity among academics. Combined with the literature review, this paper employs CVM to assess the recreational value of Jiuzhaigou and analyzes the biases in the survey. And finally, we conduct tests on test-retest reliability, content validity and convergent validity. It is found that (1) the recreational values of Jiuzhaigou in 2008 and 2009 are 333 million RMB yuan and 346 million RMB yuan respectively. Compared with other valuation results, these values are significantly lower; (2) the biases in CVM lie both in CVM itself and in the procedures where CVM is carried out. The former problems include hypothetical bias, information bias, protest response bias and strategic bias. The latter ones refer to questionnaire design, definition of population and samples, sampling methodology, and statistical analysis process; (3) the test on test-retest reliability indicates that this study is of high reliability; i.e. the valuation results are of high stability and repeatability; but the tests on content validity and convergent validity suggest low validity of the results and the fact that CVM tends to underestimate the value of tourist resources; (4) the development of CVM should shift gradually from case studies to validity and reliability tests. Thus, we have to establish and perfect the CVM implementation rules, which are suitable for developing countries to ensure validity and reliability of the valuation results.

Key words: contingent valuation method (CVM), value of the tourist resources, bias, reliability, validity, Jiuzhaigou