Comparing Apples to Apples or Apples to Oranges: The Role of Mental Representation in Choice Difficulty

Representation Concrete Representation Abstract Representation Concrete Representation Noncomparable • Comparable Noncomparable • Comparable Comparing Apples to Apples or Apples to Oranges 509 Moreover, we established that this mediation was indeed significant, because the bias-corrected confidence interval (CI) of the bootstrapping mediation test (Preacher and Hayes 2008) did not include zero (95% CI = -1.108, -.168; 5,000 resamples). Study 1 indicates support for our proposition that choice difficulty is not a stable property of the choice set configuration but depends on the mental level at which a choice is represented. Contrary to prior research, which has generally noted that noncomparable choices are intrinsically more difficult than comparable choices, we show that decision makers perceive noncomparable choices as more difficult than comparable choices when they adopt a concrete representation but view choices between the same comparable options as more difficult than the same noncomparable options when they form an abstract representation. One way to explain why prior research has found noncomparable choices to be more difficult than comparable choices is that a consumer's default mindset is likely to be more concrete than abstract (Khan, Zhu, and Kalra 2011). The decision criteria readily accessible in the default mindset should, therefore, be based on concrete, low-level considerations, making comparable choices easier than noncomparable ones. To test whether the default construal is indeed more concrete than abstract, we conducted a posttest and measured how participants in different construals identified certain actions. Previous research (e.g., Liberman and Trope 1998) has used this approach to measure shifts in construal, distinguishing a high-level identification in which the activity is linked to its superordinate purpose (the "why" of the activity) and a low-level identification in which the activity is linked to its subordinate means (the "how" of the activity). Participants from a national online pool (N = 149) completed the Behavioral Identification Form (BIF; Vallacher and Wegner 1987, 1989), a previously established measure of action identification. For each of the 25 activities listed in this measure (e.g., "locking a door"), participants chose between one of two identifications: an option related to the "why," or the abstract aspect of the activity (e.g., "securing the house"), and one related to the "how," or the concrete aspect of the activity (e.g., "putting a key in the lock"). Before beginning the BIF, participants were randomly assigned to a concrete, abstract, or control representation condition. As in Study 1, participants wrote what they imagined their life to be like tomorrow (in the concrete condition) and one year later (in the abstract condition). In the control condition, they proceeded directly to the BIF. We assigned 1 point if participants chose an abstract identification for the action and 0 if they chose a concrete identification. This resulted in a composite score ranging from 0 to 25, with higher scores corresponding to more abstract ratings. A one-way ANOVA on this composite score yielded a significant effect of representation condition (F(2, 146) 3.5,p < .05). Planned comparisons showed that although participants identified actions at a higher level in the abstract representation condition (M b̂stract = 17.43, SD b̂stract = 5.29) cornpared with the control condition (M -̂ô tpoi = 14.86, SDcontroi = 5-66; F(l, 98) 5.52,p < .05), there was no significant difference in the level of action identification between the control and the concrete representation conditions (Meoncrete 14.67, SD,„„„ete = 6.51; F(l, 96) = .02,/? = n.s.), suggesting that a consumer's default mindset is indeed likely to be more concrete than abstract. This finding explains why prior research has reported noncomparable choices to be more difficult than comparable choices. The purpose of the next study is threefold. First, we aim to test the robustness of the effects by using a fully balanced design to ensure that the findings in Study 1 cannot be attributed to specific choice option pairings. Second, in Study 2, we use a more naturalistic manipulation of the representation level to demonstrate that shifts in representation can occur naturally through everyday situational and contextual cues. Third, instead of using an indirect measure of difficulty, we directly assess participants' experienced level of difficulty. STUDY 2: EFFECT OF DELIVERY TIME ON CHOICE DIFFICULTY AND CHOICE SATISFACTION In Study 2, we use the length of the delivery time of an item to evoke an abstract or concrete mental representation naturally. Previous research has found that distant future occurrences are represented abstractly, whereas near future occurrences are represented concretely (Trope and Liberman 2003). Thus, we expect that choice options with a delivery date in the distant future will be represented more abstractly than those with a delivery date in the near future; because these shifts in representations can make different decision criteria more accessible, they will moderate the difficulty of choosing between comparable and noncomparable options as well as consumers' satisfaction with the decision.