Starting point anchoring effects in choice experiments

Jacob Ladenburg, Søren Bøye Olsen

    1264 Downloads (Pure)

    Abstract

    Anchoring is acknowledged as a potential source of considerable bias in Dichotomous Choice Contingent Valuation studies. Recently, another stated preference method known as Choice Experiments has gained in popularity as well as the number of applied studies. However, as the elicitation of preferences in Choice Experiments resembles the Dichotomous Choice format, there is reason to suspect that Choice Experiments are equally vulnerable to anchoring bias. Employing different sets of price levels in a so-called Instruction Choice Set presented prior to the actual choice sets, the present study finds that preferences elicited by Choice Experiments can be subject to starting point anchoring bias. Different price levels provoked significantly different distributions of choice in two otherwise identical choice set designs. On a more specific level, the results indicate that the anchoring subjectivity in the present study is gender dependent, pointing towards, that female respondents are prone to be affected by the price levels employed. Male respondents, on the other hand, are not sensitive towards these prices levels. Overall, this implicates that female respondents, when employing a low-priced Instruction Choice Set, tend to express lower willingness-to-pay than when higher prices are employed.
    Original languageEnglish
    Place of PublicationCopenhagen
    PublisherCenter for Skov, Landskab og Planlægning/Københavns Universitet
    Number of pages33
    Publication statusPublished - 2006
    SeriesIFRO Working Paper
    Number6
    Volume2006

    Keywords

    • Former LIFE faculty
    • Choice experiment
    • Starting point
    • Starting point anchoring bias
    • Non-market valuation
    • Gender-specific discrfepancy

    Fingerprint

    Dive into the research topics of 'Starting point anchoring effects in choice experiments'. Together they form a unique fingerprint.

    Cite this