Unstructured Direct Elicitation of Decision Rules

Abstract

We investigate the feasibility of unstructured direct-elicitation (UDE) of decision rules consumers use to form consideration sets. With incentives to think hard and answer truthfully, tested formats ask respondents to state non-compensatory, compensatory, or mixed rules for agents who will select a product for the respondents. In a mobile-phone study two validation tasks (one delayed 3 weeks) ask respondents to indicate which of 32 mobile phones they would consider from a fractional 4 5 x2 2 design of features and levels. UDE predicts consideration sets better, across profiles and across respondents, than a structured direct-elicitation method (SDE).It predicts comparably to established incentive-aligned compensatory, non-compensatory, and mixed decompositional methods. In a more-complex (20x7x5 2 x4x3 4 x22 ) automobile study, noncompensatory decomposition is not feasible and additive-utility decomposition is strained, but UDE scales well. Incentives are aligned for all methods using prize indemnity insurance to award a chance at $40,000 for an automobile plus cash. UDE predicts consideration sets better than either additive decomposition or an established SDE method (Casemap). We discuss the strengths and weaknesses of UDE relative to established methods.

Consumer Preference Elicitation of Complex Products Using Fuzzy Support Vector Machine Active Learning

“…Second, as a number of prior studies (e.g., Ding et al 2005;Ding 2007;Ding et al 2009;Dong et al 2010;Ding et al 2011;Hauser et al 2014) suggest that incentive alignment offers benefits such as greater respondent involvement, less boredom, and higher data quality, we incorporated incentive alignment in this application. At the beginning of the experiment, we told the participants that we would award a computer tablet device to one randomly selected participant from this study, plus cash representing difference between the price of the tablet device and $900.…”

“…We included both initial and delayed validation tasks as in Ding et al (2011) and Hauser et al (2014).…”

“…The KL divergence measure is an information-theory-based measure of divergence (Kullback and Leibler 1951;Chaloner and Verdinelli 1995). Dzyabura and Hauser (2011), Ding et al (2011), and Hauser et al (2014 demonstrate that, for consideration data, the KL divergence measure provides an evaluation of predictive ability that is rigorous and discriminates well. We followed the formulae provided in Dzyabura and Hauser (2011) to calculate the KL divergence measures.…”

Should Optimal Designers Worry About Consideration?

“…Shortcomings of this type of method have been reported and include inconsistencies between the stated screening criteria and observed choices from the same individual [39]. More recent research shows that the accuracy of direct elicitation approaches can be improved by designing experiments that are incentive-compatible: for example, by participating in a survey in which respondents describe their screening rules for new vehicles, they have a decent chance of actually winning a vehicle described [25]. Estimation tools that are widely applied in traditional discrete choice analysis, e.g.…”

“…Marketers have only shown the advantage of modeling consideration through improvement in model predictive accuracy, though this has been accomplished across a wide variety of product categories including cameras, batteries, automobiles, cellphones, and computers [23,24,18,25,26]; e.g., see Table 1 below. Simulation experiments have illustrated the limits of classical compensatory models including the multinomial and random coefficient (Mixed) Logit models when modeling non-compensatory choice behavior [27,28,29].…”