Anova

STEPHEN M. NOWLIS, RAVI DHAR, and ITAMAR SIMONSON*

Consumers often resolve trade-offs in a particular order. For example, when making flavor and size decisions, consumers might first decide which flavors to choose and then decide which sizes of those flavors to choose. This research examines the effect of decision order on purchase quantity decisions. The authors build on prior work ondecision difficulty and conflict to show that consumers choose more overall, and more variety, when they consider a less replaceable attribute in an earlier, rather than a later, stage in the purchase decision. For example, consumers choose a greater quantity when flavor (or brand) decisions precede, rather than follow, size decisions. The authors find that the degree of attribute replaceability alsomoderates the effect of decision order on quantity chosen. Furthermore, marketers can influence the amount chosen by altering the organization of the shelf display. Finally, the authors find that when consumers explicitly consider the possibility of deferring their decisions, the effect of decision order declines. Keywords: consumer choice, purchase quantity decisions, decision order, brand choice,attribute replaceability

The Effect of Decision Order on Purchase Quantity Decisions

Imagine that a consumer is choosing between different types of yogurt and focusing on trade-offs involving size and flavors. This consumer might first decide which flavors and how many of those flavors to choose, or the consumer might first decide which sizes and how many of those sizes to choose. Althoughprior research on consumer decision making has commonly focused on choice and choice incidence, relatively little research has focused on another important aspect of consumer decision making—the quantity of the product chosen (e.g., Chernev 2008; Wansink, Kent, and Hoch 1998). A major finding from prior research on choice and choice incidence reveals that both aspects depend on the context, or theset of alternatives under consideration, and the task, or the procedure for making the decision (e.g., Huber, Payne, and Puto 1982; Simonson et al. 2001). In this research, we examine how the purchase

*Stephen M. Nowlis is Professor of Marketing, Olin School of Business, Washington University (e-mail: nowlis@wustl.edu). Ravi Dhar is George Rogers Clark Professor of Management and Marketing andDirector of the Center for Customer Insights, Yale School of Management, Yale University (e-mail: ravidhar@yale.edu). Itamar Simonson is Sebastian S. Kresge Professor of Marketing, Graduate School of Business, Stanford University (e-mail: itamars@stanford.edu). Dilip Soman served as associate editor for this article.

quantity decision may also be influenced by certain aspects of the context ortask—in this case, the order in which consumers make decisions, such as in the yogurt example. Prior work on consumer choice and choice incidence has shown that decision conflict is an important variable in determining what people choose and whether they make a choice. Our research builds on this work by suggesting another important consequence of decision conflict; namely, it can also affect theamount chosen. In particular, we propose that as the decision becomes more difficult, consumers resolve this difficulty by choosing a greater variety and, therefore, greater overall quantity. By distributing the choice of options across several different product configurations, consumers can resolve this decision conflict. If the purchase quantity decision is also systematically affected bydecision difficulty, it can have important consequences for decision making that occurs in stages. Because consumers often make decisions in a specified sequence, we posit that the order in which these decisions are considered (e.g., Tversky and Sattath 1979) will have a systematic effect on the overall quantity chosen. If a consumer confronts a more difficult attribute trade-off in the first stage of...