|
on Discrete Choice Models |
| By: | Jens Leth Hougaard (Department of Economics, University of Copenhagen); Tue Tjur (Copenhagen Business School); Lars Peter Østerdal (Department of Economics, University of Copenhagen) |
| Abstract: | Recent studies have tested the preference axioms of completeness and transitivity, and have detected other preference phenomena such as unstability, learning- and tiredness effects, ordering effects and dominance, in stated preference discrete choice experiments. However, it has not been explicitly addressed in these studies which preference models are actually being tested, and the connection between the statistical tests performed and the relevant underlying models of respondent behavior has not been explored further. This paper tries to fill that gap. We specifically analyze the meaning and role of the preference axioms and other preference phenomena in the context of stated preference discrete choice experiments, and examine whether or how these can be subject to meaningful (statistical) tests. |
| Keywords: | stated preference discrete choice experiments; completeness; transitivity; random utility; statistical tests |
| JEL: | B41 C52 D01 |
| Date: | 2006–06 |
| URL: | http://d.repec.org/n?u=RePEc:kud:kuiedp:0611&r=dcm |
| By: | Ian J. Bateman (University of East Anglia); Michael P. Cameron (University of Waikato); Antreas Tsoumas (University of the Aegean) |
| Abstract: | This paper investigates the nature of stated preferences for reducing air pollution impacts. Specifically a contingent valuation (CV) experiment is designed to elicit individuals’ values for reducing these impacts and to examine how these may change when multiple schemes for reducing differing impacts are valued. The novel survey design allows simultaneous testing for the presence of several anomalies reported in the CV literature within the same context, including (i) scope sensitivity (ii) part-whole or substitution effects (iii) ordering effects and (iv) visible choice set effects. Results indicate some scope sensitivity and interaction between ordering effects and visible choice set effects, as well as substantial part-whole or substitution effects between two exclusive schemes. A practical consequence of these findings is that estimates of the value of combined programmes may not readily be obtained by summing the values of their constituent parts obtained using the CV method. |
| Keywords: | air pollution; contingent valuation; stated preferences; part-whole effect; experimental surveys |
| JEL: | C42 C90 Q51 Q53 |
| Date: | 2006–05–11 |
| URL: | http://d.repec.org/n?u=RePEc:wai:econwp:06/08&r=dcm |
| By: | Jeffrey Czajkowski (Department of Economics, Florida International University) |
| Abstract: | This paper demonstrates by way of data simulations that structural shift models designed to identify unbiased willingness-to-pay (WTP) from a double-bound dichotomous choice contingent valuation methodology fail to do so within a respondent Bayesian updating context. We show that failure of the existing structural shift models is due to identification issues, dummy variable misspecification, and incompatible priors of true WTP between researchers and respondents. We further show that much of the bias can be accounted for when the structural shift specification is extended to be a function of the presented bid amount. |
| Keywords: | Double-bound dichotomous choice, Structural shift models, True WTP, Bayesian updating |
| JEL: | Q50 Q51 |
| Date: | 2006–06 |
| URL: | http://d.repec.org/n?u=RePEc:fiu:wpaper:0606&r=dcm |
| By: | Yasemin Boztug; Lutz Hildebrandt |
| Abstract: | The following research is guided by the hypothesis that products chosen on a shopping trip in a supermarket can indicate the preference interdependencies between different products or brands. The bundle chosen on the trip can be regarded as the result of a global utility function. More specifically: the existence of such a function implies a cross-category dependence of brand choice behavior. It is hypothesized that the global utility function related to a product bundle results from the marketing-mix of the underlying brands. Several approaches exist to describe the choice of specific categories from a set of many alternatives. The models are discussed in brief; the multivariate logit approach is used to estimate a model with a German data set. |
| Keywords: | market basket analysis, multivariate logit model, brand choice behavior, marketing-mix |
| JEL: | C31 C33 M31 |
| Date: | 2005–05 |
| URL: | http://d.repec.org/n?u=RePEc:hum:wpaper:sfb649dp2005-028&r=dcm |
| By: | Alistair Munro (Department of Economics, Royal Holloway, University of London); Ian J. Bateman (School of Environmental Sciences, University of East Anglia, Norwich, UK) |
| Abstract: | Standard practice in stated preference typically blurs the distinction between household and individual responses, but without a clear theoretical or empirical justification for this approach. To date there have been no empirical tests of whether values for say a two adult household elicited by interviewing one randomly selected adult are the same as the values generated by interviewing both adults simultaneously. Using cohabiting couples, we conduct a choice experiment field study valuing reductions in dietary health risks. In one treatment a random individual is chosen from the couple and interviewed alone; in the other treatment, both partners are questioned jointly. We find significant differences in household values calculated from joint as opposed to individual responses, with further variation between the values elicited from men and women. Our results question the assumption, implicit in common practice, that differences between individually and jointly elicited estimates of household values can effectively be ignored. |
| Keywords: | Household values, choice experiment, contingent valuation, food and health risks |
| JEL: | C92 D13 D80 |
| Date: | 2006–02 |
| URL: | http://d.repec.org/n?u=RePEc:hol:holodi:0602&r=dcm |