Advances in Consumer Research Volume 29, 2002 Pages 460-462
SPECIAL SESSION SUMMARY
GENERATING OPTIONS IN CONSUMER CHOICE
Alexander Chernev, Northwestern University
Research in consumer decision-making has traditionally focused on how consumers process the available information. For example, most choice research assumes that the decision problem has already been structured by the readily available attribute information. In reality, consumers often generate options in the process of evaluating choice alternatives. Papers presented in this session examine how consumers generate alternatives, the factors that drive this process, as well as its consequences for choice. The session offers an integrative view on factors influencing option generation and unifies the findings of the individual papers into a broader framework for investigating the role of generating options in consumer choice.
Specifically, the session addressed the following issues:
1. How do consumers make decisions about purchase intent? How does asking consumers to generate alternatives prior to purchase influences purchase intent and choice? Does purchase intent and choice differ systematically for hedonic and utilitarian products when consumers are forced to generate alternative uses for the purchase price?
2. What is the relationship between the mental availability of choice options and preference for those options? Might a highly available option be seen as attractive simply because of its availability? How and when might availability be used as a cue for preference? What factors moderate the preference-availability relationship?
3. How does generating attribute preferences prior to making a selection from a set of alternatives affect consumer decisions? How does a pre-choice preference articulation moderate the impact of the size of the choice set on consumer decisions? What is the role of a pre-choice articulation of attribute tradeoffs on the strength of consumer preferences and how does product assortment moderate this process.
The discussion, lead by Dipankar Chakravarti, focused on integrating the individual presentations into a more general framework. The discussion underscored that, in addition to providing insights on the theoretical implications of understanding how consumers generate alternatives, this session contributes to the understanding of several content areas such as information processing, behavioral decision theory, problem solving, consumer decision-making and choice.
"THE EFFECTS OF GENERATING OPTIONS ON PURCHASE INTENT FOR HEDONIC AND UTILITARIAN PRODUCTS"
Ravi Dhar, Yale University
Steve Nowlis, Arizona State University
How do consumers make decisions about purchase intent? The standard theory of choice assumes that preferences are completeBsuch that any expression of intent implicitly considers all possible alternatives that could be chosen. In reality, since information on all possible alternatives is either not readily accessible or too costly to process, the intent judgments may be susceptible to systematic errors. Our goal in this research is to examine how asking respondents to generate alternatives prior to purchase influences purchase intent and choice.
Specifically, we test whether purchase intent and choice differs systematically for hedonic and utilitarian products when consumers are forced to generate alternative uses for the amount of money corresponding to the purchase price. Our theoretical framework builds on prior research on context effects in judgment and choice by outlining how the generated options can systematically change the decision context within which options are evaluated. We show that asking respondents to generate unspecified alternatives or hedonic alternatives tends to hurt the purchase intent of utilitarian options more than hedonic options due to contrast and assimilation effects respectively.
Four studies investigated the effect of asking subjects to simulate alternative use for an amount of money on purchase intent for hedonic and utilitarian items. The items were pre-tested to be either primarily hedonic or primarily utilitarian and matched for the monetary amount. Experiment 1 used a 2 (hedonic or utilitarian item) X 4 (generation, no generation, generate hedonic alternatives, generate utilitarian options) between subjects design. The results show that purchase intent was more likely to decrease after generating unspecified or hedonic options when the good described was utilitarian than when the good described was hedonic. These results were obtained when the same item was framed as utilitarian (a modem to access internet for games) or hedonic (a modem to access the net to help with coursework).
A second experiment examined the effect of generating options on choice between hedonic and utilitarian options. A second objective of this study was to examine whether these results are caused by differences between the two types of goods or more generally, by how consumers think about them. The results show that generating options (unspecified and hedonic option generation condition) increases the relative choice for the hedonic option. Two other studies examined the mechanisms underlying the proposed choice and judgment effects to support our theory.
The results of these experiments support the proposition that asking respondents to generate unspecified alternatives or hedonic alternatives tends to hurt the purchase intent of utilitarian options more than hedonic options due to the corresponding contrast and assimilation effects.
"GENERATED AND SELECTED FAVORITES: ON AVAILABILITY AS A CUE FOR PREFERENCE"
Yuval Rottenstreich, University of Chicago
Sanjay Sood, UCLA
Lyle Brenner, University of Florida
What is the relationship between the cognitive availability of choice options and preference for those options? Some options are available precisely because they are highly liked. For example, when asked to name several local restaurants, a consumer is likely to recall restaurants she has recently enjoyed. Thus, preferences may drive option availability. In this research we examine the reciprocal relationship between availability and preference. We propose that a highly available option may be seen as attractive simply because it is easily called to mind. That is, consistent with work suggesting that availability is used to draw inferences about event likelihood (Tversky and Kahneman 1973), self-assessments (Schwarz et al. 1991), and brand evaluations (Wanke, Bohner, and Jurkowitsch 1997), we propose that availability may be used as a cue to preference.
We examine the effect of availability on preference in a series of experiments in which some consumers express preference when no options are explicitly provided ("What is your favorite fast-food restaurant?") and other consumers express preference given an explicitly provided list of options ("Which of the following is your favorite fast-food restaurant?"). We refer to preferences when options are not explicitly provided as generated favorites and preferences when options are explicitly provided as selected favorites. Across several tasks and contexts, we find that options higher in availability are relatively more likely to be generated favorites, whereas options lower in availability are relatively more likely to be selected favorites.
In Experiment 1, we document this pattern and show that the differences between generated and selected choice shares can be accounted for by independent measures of option availability. The results are consistent with the hypothesis that consumers use ease of option generation as a cue to preference, in effect confusing availability for preference. In Experiment 2, we replicate this pattern in categories where consumers are highly familiar with all the options in the category, and show that the relationship between availability and preference is not merely due to the inability to recall certain members of the category (which naturally cannot be generated if they are not recalledBcf. Nedungadi 1990). The discrepancy between generated and selected favorites occurs even in categories in which the set of all items is easily recalled (e.g., seasons of the year) and in categories where consumers have become familiar with all of the items based on exposure earlier in the experiment.
Two additional experiments examine moderators of the relationship between availability and preference. Experiment 3 investigates whether generated favorites in one category are sensitive to priming preferences in another category. Consumers first evaluated one of two types of pizza restaurants, either a national chain or a local upscale establishment. Later, consumers generated their favorite hamburger restaurant. Consumers who had been primed with a national pizza chain were more likely to generate a national hamburger chain as their favorite, while consumers who had been primed with a local upscale pizza restaurant were more likely to generate a local upscale hamburger restaurant as their favorite. This pattern suggests that option similarity may affect availability, which in turn serves as a cue to preference.
In Experiment 4 we examine the moderating effect of goals by manipulating whether consumers express their own preferences, or predict the preferences of another person (e.g., a close friend). We find that the relationship between availability and ones own generated favorites is higher than the correlation between availability and ones (generated) assessment of another persons preferences. This pattern supports the hypothesis that availability serves as a cue for ones own preferences, as the ease of option generation is presumably less likely to be viewed as a cue to somebody elses preference.
Taken together, the results of these experiments support the proposition that consumers use option availability as a cue to preference, and that factors such as priming and the goal of the consumer influence the relation between availability and preference.
"PREFERENCE ARTICULATION EFFECTS IN CONSUMER CHOICE"
Alexander Chernev, Northwestern University
How does generating attribute preferences prior to making a selection from a set of alternatives affect consumer decisions? How does the impact of pre-choice preference articulation on consumer preferences depend on the number of alternatives in the choice set? What factors account for the effect of pre-choice preference articulation on consumer decision processes? Building on the constructive view of preferences (Bettman, Luce, and Payne 1998; Simonson and Tversky 1992), this research argues that consumers often evaluate choice alternatives not only in search of the option that best matches their established preferences, but also in an attempt to construct the very preferences that are used to determine the attractiveness of the choice options. In this context, I propose that the nature of the decision process is a function of the degree to which consumers have a readily available ideal attribute combination. Consumers without a readily available ideal point will have to first construct their attribute preferences to evaluate the alternatives in the setBtask that can be more easily accomplished in the context of a smaller rather than larger assortment. Therefore, I predict that consumers without a readily available ideal attribute combination are likely to face easier decision, and hence have stronger preferences, in the context of a smaller rather than a larger assortment. In contrast, because individuals with a readily available ideal attribute combination are faced with a simpler decision task that does not necessarily imply preference construction, the difference in the impact of assortment is likely to be less pronounced.
The above predictions are empirically tested in a series of three experiments. The first experiment tests the key proposition that the pre-choice articulation of the ideal attribute combination moderates the impact of product assortment on consumer preferences. In this experiment, respondents were randomly assigned to the conditions of a 2 (preference articulation) X 2 (assortment) design and were asked to choose from either a small or a large assortment. In addition, some of the subjects were asked, prior to the choice task, to articulate their ideal point by stating their optimal attribute combination (i.e., the ideal value on each attribute). The key dependent variable was individuals strength of preferences measured by the likelihood of brand switching when subjects were given an option to substitute their initial selection with a default option (e.g., the most popular brand).
Consistent with the experimental predictions, the data show that that preference articulation moderates the impact of assortment on consumer decisions. For consumers without articulated preferences larger assortments were associated with higher probability of choosing the default option, whereas for consumers with an articulated ideal point, the effect was in the opposite direction. These data lend support for the proposition that large assortments could have an adverse impact on consumer preferences and increase the probability of a no-choice option and, more importantly, that the impact of assortment on consumer preferences is moderated by the availability of an articulated ideal point.
Experiment 2 tests the robustness of the findings reported in the first experiment in a scenario where the stimuli are presented in a more holistic manner, thus extending the validity of the Experiment 1 findings to a broader decision context. Finally, Experiment 3 explicitly investigates the role of tradeoffs in articulating a salient ideal attribute combination. It demonstrates that the effect observed in the first two studies is less pronouncedin a context where consumers articulate their preferences without explicitly articulating attribute tradeoffs. Specifically, the data show that consumers who were asked to articulate their attribute preferences in a format that explicitly required making tradeoffs were less likely to be adversely affected by increasing the assortment than consumers who articulated their preferences in an evaluation format. This finding lends further support to the notion that the articulation of an ideal point moderates the impact of assortment on consumer preferences.
The data across these three experiments show that the pre-choice articulation of attribute preferences moderates the impact of product assortment on consumer preferences, whereby consumers with an articulated ideal attribute combination are less likely to be susceptible to the potential adverse impacts of large assortments. This research further demonstrates that articulating attribute tradeoffs plays a key role in moderating the impact of assortment on consumer decision processes. The data obtained form these experiments offer a better understanding of the impact of pre-choice preference articulation on strength of preferences.
Bettman, James R., Mary Frances Luce, and John W. Payne (1998), "Constructive Consumer Choice Processes," Journal of Consumer Research, 25 (3), 187B217.
Nedungadi, Prakash (1990), "Recall and Consumer Consideration Sets: Influencing Choice without Altering Brand Evaluations," Journal of Consumer Research, 17 (3), 263-276.
Schwarz, Norbert, Herbert Bless, Fritz Strack, and Gisela Klumpp (1991), "Ease of Retrieval as Information: Another Look at the Availability Heuristic," Journal of Personality & Social Psychology, 61 (2), 195-202.
Simonson, Itamar and Amos Tversky (1992), "Choice in Context: Tradeoff Contrast and Extremeness Aversion," Journal of Marketing Research, 29 (3), 281-295.
Tversky, Amos and Daniel Kahneman (1973), "Availability: A Heuristic for Judging Frequency and Probability," Cognitive Psychology, 5 (2), 207-232.
Wanke, Michaela, Gerd Bohner, and Andreas Jurkowitsch (1997), "There Are Many Reasons to Drive a BMW: Does Imagined Ease of Argument Generation Influence Attitudes?," Journal of Consumer Research, 24 (2), 170-177.