The Anchoring Vignettes Website

This site supplements the information in

Gary King, Christopher J.L. Murray, Joshua A. Salomon, and Ajay Tandon. "Enhancing the Validity and Cross-cultural Comparability of Survey Research," American Political Science Review, 97, 4 (December, 2003); reprinted with printing errors corrected, February, 2004.

Gary King and Jonathan Wand. Comparing Incomparable Survey Responses: New Tools for Anchoring Vignettes, Political Analysis, 15, 1 (Winter, 2007): Pp. 46-66.

and more generally offers survey researchers, and others, help in addressing two long-standing questions:

How can we do survey research if different respondents (perhaps from different cultures, countries, or ethnic groups) understand questions in completely different ways, or if investigators mean one thing and respondents think they mean something else?
How can we develop accurate measures of complicated concepts which we can define only by example ("you know it when you see it"), and when attempts to produce more concrete questions tend to be more concrete but no more valid?

To at least partially ameliorate these problems, we introduce the idea of anchoring vignettes. These are (usually brief) descriptions of hypothetical people or situations that survey researchers can use to correct otherwise interpersonally incomparable survey responses. Survey questions are a function of the actual quantity being measured along with a dose of interpersonal incomparability (i.e., different for each respondent). The survey literature has focused for decades on asking more concrete questions intended to reduce the incomparable portion, but with only mixed success. The new idea underying anchoring vignettes is to measure directly, and then subtract off, the incomparable portion. To do this, we ask respondents for self-assessments of the concept being measured along with assessments, on the same scale, of each of several hypothetical individuals described in anchoring vignettes. Since the actual (but not necessarily reported) levels for the people in the vignettes are, by the design of the survey, invariant over respondents, the only reason answers to the vignettes will differ over respondents is interpersonal incomparability. This provides sufficient information for the statistical models we have designed to correct the self-assessments.

Anchoring vignettes are also being used by philosphers, lawyers, and others to help define (and not necessarily measure) concepts by example, or from the bottom up. Especially when agreeing on broad theoretical concepts is difficult or impossible, it is still often possible to agree on the specifics of individual cases. By defining sets of anchoring vignettes, it is often possible to arrive at a definition inductively using this same approach.

Academic Papers
Library of Anchoring Vignette Examples, by researchers in many fields
Understanding Complicated Concepts by Example: Defining "Privacy" with Anchoring Vignettes, by the National Research Council's Committee on Privacy in the Information Age
Frequently Asked Questions
Software
Conferences
Anchoring Vignettes Email Discussion list: [Un-]Subscribe or Search archives
Full length video of a lecture on anchoring vignettes (to the Boston Chapter of the American Statistical Association, filmed by Harvard@Home)
10 second video clip with everything you ever wanted to know about interpersonal incomparability

Please send email if you have example vignettes, suggestions for the FAQ, papers, or other information related to anchoring vignettes you would like us to list here.

This research is an outgrowth of an ongoing National Institute on Aging grant on The Global Burden of Disease 2000 in Aging Populations (P01 AG17625-01). We also received support from NSF (grants SES-0112072 and IIS-9874747), the World Health Organization, and the Weatherhead Center for International Affairs and the Institute for Quantitative Social Science at Harvard University.