Motivational Differences in Seeking Out Evaluative Categorization Information

Abstract

Previous research shows that people draw finer evaluative distinctions when rating liked versus disliked objects (e.g., wanting a 5-point scale to evaluate liked cuisines and a 3-point scale to rate disliked cuisines). Known as the preference-categorization effect, this pattern may exist not only in how individuals form evaluative distinctions but also in how individuals seek out evaluative information. The current research presents three experiments that examine motivational differences in evaluative information seeking (rating scales and attributes). Experiment 1 found that freedom of choice (the ability to avoid undesirable stimuli) and sensitivity to punishment (as measured by the Behavior Inhibition System/Behavioral Approach System [BIS/BAS] scale) influenced preferences for desirable and undesirable evaluative information in a health-related decision. Experiment 2 examined choice optimization, finding that maximizers prefer finer evaluative information for both liked and disliked options in a consumer task. Experiment 3 found that this pattern generalizes to another type of evaluative categorization, attributes.

Keywords

preference evaluation information seeking rating attribute

You’re planning a dinner, deciding on a movie, or searching for a mechanic. How do you make your decision? Although there are many sources of information, you may seek out evaluative information, such as rating scales, to help determine what items or services to purchase. These scales are usually bipolar, which some research suggests provides adequate reliability (Garner, 1960; Komorita & Graham, 1965; Weng, 2004). However, additional research shows that people not only prefer using evaluative scales with more distinctions for liked versus disliked objects but also find them more efficacious for conveying information. Known as the preference-categorization effect, previous research has shown that people see more nuanced distinctions for liked versus disliked items in both traditional categorization (e.g., sorting tasks; Smallman & Roese, 2008) and evaluative categorization (e.g., the number of rating scale points preferred to express an opinion; Smallman, Becker, & Roese, 2014). The current research examines whether this effect is evident in how individuals seek out evaluative information (in the form of rating scales and attributes).

The Preference-Categorization Effect

The preference-categorization effect states that people make finer distinctions for preferred objects (Smallman et al., 2014; Smallman & Roese, 2008). Evidence of this effect has been shown in both traditional and evaluative categorization. In traditional categorization tasks, people will create more groups for liked (vs. disliked) items (Smallman & Roese, 2008). Notably, this effect occurred independently of expertise, as preference was created using an affect-based associative conditioning paradigm (see Hofmann, De Houwer, Perugini, Baeyens, & Crombez, 2010). Evaluative categorization refers to the distinctions used to express an opinion or evaluation of a stimulus. Applied to this context, the preference-categorization effect demonstrates people’s preference for more evaluative distinctions for liked (vs. disliked) items. So far, research has focused only on the number of evaluative scale points preferred to express overall ratings of liked and disliked items (Smallman et al., 2014). For example, a jazz fan evaluating jazz pieces would prefer a rating scale with more distinctions (e.g., “Okay,” “Good,” “Really Good,” “Great,” and “Best”), whereas someone who dislikes jazz would prefer a rating scale with fewer distinctions (e.g., “Bad” and “Worst”).

One explanation for why the preference-categorization effect occurs relates to the level of elaboration of preferred objects. People spend a great deal of time thinking about the things they love, often keeping mementoes close by. This tendency to reengage relates to the Law of Effect (Thorndike, 1898), which states that people repeatedly engage in behaviors that are rewarding or result in feelings of positive affect (Carver, 2003; Gable & Harmon-Jones, 2008; Hoch & Deighton, 1989). This repetitive engagement, in turn, leads to greater elaboration (i.e., an effortful process that occurs when a person compares and connects attitude-relevant ideas; Petty & Cacioppo, 1986) of preferred objects, which may increase perceptions of nuances in preferred objects and activities. Conversely, disliking an object may limit the amount of elaboration and nuanced distinctions. This idea is supported by research on out-group homogeneity, which shows that, when people avoid disliked out-groups, they elaborate less on the disliked group’s distinctions and perceive less variability within that group (Linville & Fischer, 1993; Park & Rothbart, 1982; Rubin & Badea, 2012). When thinking about this in terms of preferences, people may go to great lengths to avoid the things they dislike, which may lead to decreased elaboration and perception of nuanced distinctions.

Both traditional and evaluative categorization research suggest that people make more nuanced distinctions for liked (vs. disliked) objects. However, the scope of evaluative categorization thus far has been fairly narrow in two important ways. First, research has focused mainly on how these distinctions are developed and on preferences for expressing one’s own opinion about liked and disliked items. A key question then is whether this pattern extends to how people seek out evaluative information about liked and disliked items. That is, when faced with a decision, will an individual be more likely to seek out finely differentiated evaluative information about a liked item, relative to a disliked item? Relatedly, how do certain situational contexts (e.g., the freedom to avoid negative stimuli) and individual differences (e.g., sensitivity to punishments, choice maximization) affect the likelihood of seeking out finely differentiated positive and negative evaluative information? Experiments 1 and 2 examine these questions.

Second, evaluative categorization thus far has only been operationalized as overall rating scales (i.e., how good/bad is this item). If the preference-categorization effect speaks broadly to the types of evaluative information individuals seek out, then it should be evident in other types of evaluative categorization, such as evaluative product attributes. Commonly used in consumer research, product attributes are features or characteristics associated with a consumer item (Lefkoff-Hagius & Mason, 1993; Mukherjee & Hoyer, 2001; Yeung & Wyer, 2004). In Experiment 3, evaluative attributes focus on the number of facets or dimensions a person would want included in a product’s evaluation. For example, when choosing a restaurant specializing in your favorite cuisine, you would not only want an overall evaluation of the restaurant but would also want evaluations on various attributes related to the restaurant (e.g., appearance, ingredients, price, taste, quantity). In contrast, when selecting a restaurant specializing in a disliked cuisine, evaluative information about fewer attributes (e.g., taste and price) may suffice. Experiment 3 examines this question.

Freedom of Choice and Sensitivity to Rewards and Punishments

Our perspective suggests that the ability to approach desirable and avoid undesirable stimuli plays a key role in the preference-categorization effect. However, not all undesirable experiences can be avoided. We theorize that the preference-categorization effect derives from approach behavior (i.e., the tendency for individuals to reindulge in that which they love; Carver, 2003) and avoidance of disliked objects. The freedom to select preference-consistent experiences is thus a key moderator. For preferred activities, freedom to repeatedly reindulge increases exposure. For disliked activities, the favored reaction is avoidance, which forecloses further opportunities to learn about disliked objects. Hence, the difference between preference and lack of preference confers approach and avoidance, which results in a differentiation in direct experience and experience-driven knowledge. When freedom of choice is removed and a person is “stuck” with a negative situation, a different result should emerge. Specifically, when choice is constrained, the road to easy avoidance is blocked. For example, normally a person who dislikes romantic comedies would not need a highly differentiated scale. They consider all romantic comedies “Bad” and will avoid them. However, if this person was forced to see a romantic comedy (e.g., complying with the wishes of a friend), they might prefer a more finely differentiated scale to select the best of the worst options. In this no-escape situation, the nuanced differences between undesirable movies are suddenly meaningful. When forced to interact with disliked items, the main goal becomes minimizing pain. Information distinguishing between bad, terrible, and worst becomes meaningful now that avoidance is impossible. Therefore, the effect of preference on evaluative categorization should be more evident under conditions of high versus low freedom of choice.

The effect of freedom of choice may also depend on a person’s approach or avoidance motivation, or sensitivity to rewards or punishments. Gray (1981, 1990) proposed an individual difference comprised of two systems (behavior inhibition system/behavioral approach system [BIS/BAS]), which affects people’s motivation to approach or avoid stimuli. The BIS focuses on avoidance of negative stimuli and is associated with negative affect (Bechara, Damasio, Tranel, & Damasio, 1997; Shen & Dillard, 2007), whereas the BAS focuses on rewards, pursuit of goals, and is associated with positive affect. Because the ability to avoid undesirable experiences may be an important factor underlying the preference-categorization effect, participants who are especially sensitive to punishment should be more likely to seek out finely differentiated evaluative information about liked (vs. disliked) items under high freedom of choice conditions. However, under low freedom of choice, they should seek out finely differentiated evaluative information for both liked and disliked items. In contrast, participants who are especially sensitive to reward may be more narrowly focused on repeated engagement with pleasurable stimuli, with little attention to the avoidance of unpleasurable stimuli. Therefore, people high in reward sensitivity may exhibit the preference-categorization effect, regardless of freedom of choice. However, this effect should be larger under high freedom of choice conditions. Experiment 1 examines whether freedom of choice moderates seeking out finely differentiated evaluative categorical information, and whether a person’s sensitivity to reward and punishment (as measured by BIS/BAS; Carver & White, 1994) influences this effect.

Choice Optimization

When freedom of choice is high, and an individual can freely approach and avoid different stimuli, a second motivational factor, choice optimization, may impact the preference-categorization effect. Choice optimization is the desire to ensure the best possible choice from available options. That is, when an individual has a goal of “getting the very best,” the motivation to seek out finely differentiated positive and negative evaluative distinctions may change. Referred to as maximization, research distinguishes between maximizers and satisficers (Parker, de Bruin, & Fischhoff, 2007; Polman, 2010; Schwartz et al., 2002). Maximizers are intrinsically motivated to seek out the “best” available option; they will naturally search for more information and examine more alternatives during the decision-making process. Conversely, satisficers are happy with options that meet an internal minimum requirement, unless extrinsically motivated. That is, satisficers are content with decisions that are “good enough” unless additional external motivation pushes them toward a more effortful decision-making process.

Although maximization can be both an individual difference (Schwartz et al., 2002) and a mind-set (Ma & Roese, 2014), the current research operationalizes individual differences in choice optimization via the Maximization Scale, which measures the degree to which a person is a maximizer versus satisficer (Schwartz et al., 2002). Relevant to the preference-categorization effect, previous research has found that maximizers prefer a larger selection of alternatives (Dar-Nimrod, Rawn, Lehman, & Schwartz, 2009; Iyengar, Wells, & Schwartz, 2006; Nenkov, Morrin, Ward, Schwartz, & Hulland, 2008; Schwartz et al., 2002), spend more time doing background research (Chowdhury, Ratneshwar, & Mohanty, 2009; Schwartz et al., 2002), and are more likely to engage in purchase-related social comparison (Schwartz et al., 2002). Taken together, this suggests that maximizers have a preference for more distinctions and are more likely to seek out finely differentiated choice-relevant information.

However, it is yet unknown whether this pattern extends to decisions focused on both liked and disliked items. Given their intrinsic motivation for information, we suggest that maximizers will find detailed information equally valuable for liked and disliked objects. They should be equally concerned with selecting the “least bad” from a disliked category as choosing the “best” from a liked category; therefore, they should seek out finely differentiated evaluative information for both liked and disliked objects. Conversely, satisficers are focused mainly on finding a good enough option, unless otherwise extrinsically motivated (e.g., by desire or liking). As a result, they should show the expected preference-categorization effect: seeking out finely differentiated evaluative information when selecting a desired item, yet preferring broadly differentiated information for an undesirable item. Experiment 2 examined how differences in maximization influence participants motivation to seek out finely differentiated evaluative information for liked (vs. disliked) objects.

The Present Investigation

Previous research has focused primarily on how participants make distinctions between items, showing that individuals prefer more differentiated scales for liked items (Smallman et al., 2014). In contrast, the current research focuses on differences in seeking out finely differentiated evaluative information (as measured by overall rating scales and attributes) for liked and disliked items. Experiment 1 used a health-related decision to test whether freedom of choice influences seeking out finely differentiated desirable and undesirable evaluative information. In addition, it examined whether this effect is contingent on an individual’s sensitivity to reward or punishment. Experiment 2 used a consumer choice task, examining whether individual differences in choice optimization (i.e., maximizers vs. satisficers; Schwartz et al., 2002) affect seeking out finely differentiated desirable and undesirable evaluative information. Finally, Experiment 3 tested whether this differential pattern of information seeking generalizes to another type of evaluative categorization, evaluative attributes (i.e., the number of product attributes evaluated). In addition, it also ruled out an alternative explanation for the preference-categorization effect based on differences in desirable and undesirable response labels.

Experiment 1

As suggested in the introduction, a potential explanation for the preference-categorization effect is having the freedom to pursue desirable and avoid undesirable stimuli. For liked objects, the main goal may be approach. Therefore, additional scale points and finely grained distinctions are necessary to capture the nuances perceived in these objects. For disliked objects, the main goal may be avoidance. Therefore, additional scale points and finely grained distinctions are unnecessary for these objects. However, not all negative experiences can be avoided. When the option to avoid is thwarted and a person is forced to interact with disliked stimuli, undesirable information may become more meaningful and greater distinctions more useful. Accordingly, when freedom of choice (i.e., the ability to avoid interacting with an item) is constrained, people should seek out finely differentiated evaluative information for both liked and disliked categories. Freedom of choice is thus a key moderator.

Experiment 1 used a 2 (freedom of choice: high vs. low) × 2 (valence: desirable outcomes vs. undesirable outcomes) between-subjects design. Participants focused on either the desirable or undesirable outcomes of an elective (high freedom of choice) or required (low freedom of choice) medical procedure. The dependent variable was the type of evaluative information (in the form of rating scales) they would want to see for each of their procedure’s desirable or undesirable outcomes (depending on condition). Participants also completed the BIS/BAS (Carver & White, 1994) scale to determine whether individual differences in approach/avoidance motivation would moderate this effect. We predicted an overall preference-categorization effect whereby people would seek out more finely differentiated evaluative information regarding desirable (vs. undesirable) outcomes. However, this effect may be driven by both the ability to avoid negative situations and the intensity of this approach/avoidance reaction. Therefore, the effect should be the strongest in high BIS (i.e., sensitivity to punishment) participants in the high freedom of choice condition. Under low freedom of choice conditions, those scoring high in BIS should seek out equally differentiated evaluative information for desirable and undesirable outcomes. Given their focus on pleasurable outcomes, high BAS (i.e., sensitivity to reward) participants should show the preference-categorization effect, regardless of freedom of choice.

Method

Participants were recruited using Amazon’s Mturk (N = 259; age M = 36.28 years, SD = 13.60; 45.9% male) in exchange for US$0.40. The data from seven participants were removed because they did not follow instructions (did not list a medical procedure or expected outcomes). Five participants did not complete the BIS/BAS, so their data are not included in those analyses. Sample size was based on a target minimum of 50 per cell. However, because of the complicated nature of the instructions, we ran an additional 59 participants to account for participant attrition. A power analysis (Gpower3.1) showed that a sample size of 259 provides 100% statistical power to detect a large effect size (d = 1.18), 99.98% statistical power to detect a medium effect size (d = 0.78), and 43.74% statistical power to detect a small effect size (d = 0.28). Participants were randomly assigned in a 2 (freedom of choice: high vs. low) × 2 (valence: desirable outcomes vs. undesirable outcomes) between-subjects design.

To manipulate freedom of choice, participants read instructions explaining elective (high freedom of choice) or required (low freedom of choice) medical procedures. Elective medical procedures were described as follows: “Medical procedures that we do not have to have to stay healthy. They are not medically required, and you have a choice in whether or not to have these procedures.” Required medical procedures were described as follows: “Medical procedures that we have to have to stay healthy. They are medically required, and you do not have a choice in whether or not to have these procedures.” To manipulate valence, participants focused on their procedure’s desirable or undesirable outcomes. Finally, they were asked to think about discussing these desirable or undesirable outcomes with their doctor. Imagining that their doctor was going to give them information about these outcomes, they were asked to determine which scale should be used to provide evaluative information about the desirable or undesirable outcomes (see supplementary materials). Evaluative scales ranged from 2 to 7 points, taken directly from Smallman et al. (2014; see the appendix).

Next, participants completed four manipulation check questions about their procedure (5-point Likert-type scales with strongly disagree and strongly agree as anchors; positive feelings, negative feelings, happiness, enjoyment; α = .92). Then, participants named their medical procedure and listed/described the desirable or undesirable outcomes they imagined. Elective procedures included LASIK eye surgery, face-lifts, implants, and gastric bypass. Required procedures included appendectomies, gallbladder removal, and cancer removal. Finally, participants completed the 24-item BIS/BAS scale (Carver & White, 1994) that included subscales for BAS Drive (e.g., “I go out of my way to get things I want,” α = .83), BAS Fun Seeking (e.g., “I often act on the spur of the moment,” α = .80), BAS Reward Responsiveness (e.g., “It would excite me to win a contest,” α = .74), and BIS (e.g., “If I think something unpleasant is going to happen, I usually get pretty worked up,” α = .85) in addition to demographic and debriefing questions.

Results

Manipulation check

The four manipulation check questions were averaged to form an overall preference rating (α = .92). A 2 (freedom of choice: high vs. low) × 2 (valence: desirable outcomes vs. undesirable outcomes) ANOVA revealed a significant main effect of choice, F(1, 248) = 5.53, p = .019, η² = .01, 95% CI = [−0.38, −0.03], and a significant main effect of valence, F(1, 248) = 658.87, p < .001, η² = .72, 95% CI = [−2.45, −2.10]). Importantly, the choice by valence interaction was not significant, F(1, 248) = 2.11, p = .15, η² = .002, 95% CI = [−0.09, 0.61]. Participants in the high freedom of choice conditions felt slightly more positive than those in the low freedom of choice conditions (M = 2.97, SD = 1.31 vs. M = 2.83, SD = 1.38), and participants in the desirable outcomes conditions felt more positive than those in the undesirable outcomes conditions (M = 4.05, SD = 0.68 vs. M = 1.77, SD = 0.74).

Choice and preference

First, we examined whether the preference-categorization effect would be impacted by freedom of choice. A 2 (freedom of choice: high vs. low) × 2(valence: positive outcomes vs. negative outcomes) ANOVA revealed a significant main effect of choice, F(1, 248) = 9.14, p = .003, η² = .03, 95% CI = [−0.92, −0.19]), a marginally significant main effect of valence, F(1, 248) = 3.74, p = .054, η² = .01, 95% CI = [−0.01, 0.72], and a nonsignificant choice by valence interaction, F(1, 248) = 2.12, p = .15, η² = .01, 95% CI = [−0.19, 1.26]. Participants sought out more finely differentiated evaluative information under high freedom of choice conditions (i.e., when thinking about elective medical procedures; M = 5.36, SD = 1.41) than under low freedom of choice conditions (i.e., when thinking about required medical procedures; M = 4.82, SD = 1.52). In addition, replicating previous results, participants sought out more finely differentiated evaluative information when thinking about positive outcomes (M = 5.26, SD = 1.52) than negative outcomes (M = 4.92, SD = 1.44).

Next, we examined our prediction that the preference-categorization effect may be driven by both the ability to avoid undesirable situations and the intensity of this avoidance reaction, using regression procedures recommended by Cohen and Cohen (1983). We entered valence (undesirable outcomes vs. desirable outcomes), freedom of choice (required procedure vs. elective procedure), and mean-centered BIS in the first step; the three two-way interaction terms in the second step; and the one three-way interaction term in the third step (Aiken & West, 1991). Unstandardized regression coefficients are from the third step (see Figure 1).

Figure 1.

Amount of evaluative information (in the form of preferred evaluative rating scales) sought out as a function of freedom of choice and valence, split by high/low sensitivity to aversive stimuli (BIS).

For the model predicting the number of evaluative scale points participants would want to see for medical procedures, there were no significant lower order effects of freedom of choice, b = 0.27, t(239) = 1.01, p = .31, 95% CI = [−0.25, 0.78]; valence, b = 0.07, t(239) = 0.27, p = .79, 95% CI = [−0.44, 0.59]; or BIS, b = 0.46, t(239) = 1.60, p = .11, 95% CI = [−0.11, 1.02]. There were also no significant two-way interactions between freedom of choice and valence, b = 0.46, t(239) = 1.24, p = .22, 95% CI = [−0.28, 1.20]; valence and BIS (b = −0.68, t(239) = −1.62, p = .11, 95% CI = [−1.50, 0.15]; or choice and BIS, b = −0.78, t(239) = −1.71, p = .09, 95% CI = [−1.67, 0.12]. However, there was a significant three-way interaction, b = 1.28, t(239) = 2.06, p = .04, 95% CI = [0.06, 2.49]. To further interpret this interaction, we conducted a simple slopes test as recommended by Dawson and Richter (2006). For high BIS participants (1 SD above mean-centered BIS), simple slopes analysis revealed that participants in the elective (high freedom of choice) condition preferred more scale points when thinking about desirable (vs. undesirable) outcomes, b = 0.90, t(239) = 2.38, p = .02. However, this effect did not occur for high BIS participants in the required (low freedom of choice) condition, b = −0.35, t(239) = −0.92, p = .36. Furthermore, a slope difference test showed that these two simple slopes were significantly different from each other, see Figure 1 first panel; t(239) = 2.33, p = .02. For low BIS participants (1 SD below mean-centered BIS), simple slopes test showed no significant difference between scale points preferred for desirable (vs. undesirable) outcomes for required, b = 0.49, t(239) = 1.38, p = .17, or elective, b = 0.17, t(239) = 0.42, p = .67, medical procedures. A slope difference test showed that there was not a significant difference between these two simple slopes, see Figure 1 second panel; t(239) = −0.60, p = .55.

Finally, we repeated the same analysis above using mean-centered BAS scores to examine whether individual differences in sensitivity to reward would impact the effect of freedom of choice on the preference-categorization effect. We did not find any significant lower order effects of freedom of choice, b = 0.32, t(239) = 1.23, p = .22, 95% CI = [−0.20, 0.84]; valence, b = 0.11, t(239) = 0.40, p = .69, 95% CI = [−0.41, 0.62]; or BAS, b = −0.16, t(239) = −0.37, p = .71, 95% CI = [−1.00, 0.69]. There were also no significant two-way interactions between freedom of choice and valence, b = 0.43, t(239) = 1.15, p = .25, 95% CI = [−0.31, 1.17]; freedom of choice and BAS, b = 0.67, t(239) = 1.04, p = .30, 95% CI = [−0.60, 1.93]; or valence and BAS, b = −0.25, t(239) = −0.42, p = .67, 95% CI = [−1.39, 0.90]; nor a significant three-way interaction, b = −0.39, t(239) = −0.45, p = .65, 95% CI = [−2.06, 1.29].¹

Discussion

The results of Experiment 1 support some of our hypotheses. Consistent with predictions, participants marginally preferred more finely differentiated evaluative information for desirable versus undesirable outcomes. Importantly, this finding goes beyond previous work (Smallman et al., 2014; Smallman & Roese, 2008) in two key ways. First, it demonstrates that this effect is evident in the types of evaluative information an individual will seek out. Second, it shows that the preference-categorization effect extends to the desirable and undesirable aspects of a single stimulus (i.e., the desirable and undesirable outcomes of the same medical procedure). In addition, we found partial support for the hypothesis that the preference-categorization effect partially stems from an ability to avoid undesirable stimuli. Supporting our hypothesis, participants who could freely avoid undesirable stimuli (i.e., those focusing on an elective medical procedure) sought out more precise evaluative information than those whose freedom to avoid was highly constrained (i.e., those focusing on a required medical procedure). Yet the nonsignificant choice by valence interaction is inconsistent with our hypotheses. So although freedom of choice affects preference for finely differentiated evaluative information, it is not dependent on valence.

Similarly, Experiment 1 partially supported predictions regarding impact of BIS/BAS on the preference-categorization effect. Consistent with predictions, the significant three-way interaction with BIS suggests that this effect is moderated by individual differences in sensitivity to negative stimuli. Whereas participants who are highly sensitive to punishment (high BIS) show the preference-categorization effect under high freedom of choice, the effect disappears when choice is constrained. That is, when participants who are highly sensitive to negative stimuli are forced to interact with stimuli, they will seek out evaluative information with similar differentiation for desirable and undesirable stimuli. In contrast, participants who are low in sensitivity to negative stimuli (low BIS) show little evidence of the preference-categorization effect regardless of choice constraint. Under both high and low freedom of choice conditions, they seek out evaluative information with similar differentiation for desirable and undesirable stimuli. In contrast, individual differences in reward sensitivity (BAS) did not influence scale selection. Inconsistent with predictions, both high and low BAS participants wanted similar differentiation regardless of valence or freedom of choice. Together, these results suggest that, although the ability to avoid undesirable stimuli may moderate the preference-categorization effect, this mechanism hinges on an individual’s reactions to negative (but not positive) stimuli more generally.

Experiment 2

Experiment 1 provided initial evidence that preference affects how individuals seek out desirable and undesirable evaluative information. In addition, it identified choice constraint and individual differences in sensitivity to punishment as important factors underlying this effect. Experiment 2 utilizes a new context (a consumer choice task) to test whether this motivation to seek out desirable and undesirable evaluative information is impacted by choice optimization (as measured by maximization, Schwartz et al., 2002). As discussed previously, maximizers are intrinsically motivated to seek out the best option when making decisions, whereas satisficers will accept any option that surpasses a minimum threshold, unless extrinsically motivated. Applied to the preference-categorization effect, maximizers should be motivated to make the best possible decision regardless of stimuli valence. As a result, they should seek out finely differentiated evaluative information for both liked and disliked items so that they can ensure they select the “best of the best” as well as the “best of the worst” alternatives. In contrast, satisficers will rely on their “good enough” minimum threshold unless extrinsically motivated (e.g., by desire or liking) to seek out finely differentiated information. Therefore, satisficers should only prefer finely differentiated evaluative information for liked items. Overall, we expected that satisficers would be more likely than maximizers to show the preference-categorization effect.

In Experiment 2, valence was manipulated on a between-subjects basis; participants focused on either a liked or disliked cuisine type. The dependent variable was the type of evaluative information (in the form of rating scales) they would want to see for their selected cuisine type. For exploratory purposes, an additional traditional categorization measure was also included. Participants also completed the Maximization Scale (Schwartz et al., 2002) to determine whether individual differences in maximization would moderate this effect.

Method

Participants were recruited using Amazon’s Mturk (N = 100; age M = 30.61 years, SD = 10.83; 62.00% male) in exchange for US$0.40. Five participants were excluded for not following instructions (did not focus only on the selected cuisine type). Sample size was based on a target minimum of 50 per cell. A power analysis (Gpower3.1) showed that a sample size of 100 provides 99.98% statistical power to detect a large effect size (d = 1.18), 93.71% statistical power to detect a medium effect size (d = 0.78), and 21.98% statistical power to detect a small effect size (d = 0.28). In a between-subjects design, participants were randomly assigned to select a liked or disliked cuisine from a list (American, Italian, Japanese, Mexican, seafood, or vegan). Participants then completed four manipulation check questions about this cuisine (5-point Likert-type scales with strongly disagree and strongly agree as anchors; positive feelings, negative feelings, happiness, enjoyment; α = .95).

Next, participants completed an evaluative categorization measure to examine the type of evaluative information (in the form of rating scales) they would want for their selected cuisine. Participants determined which scale should be used to provide evaluative information about their selected cuisine type. Participants selected their preferred rating scale from a selection of six unipolar rating scales (ranging from 2 to 7 scale points; see the appendix). These scales were identical to Experiment 1. Participants then completed an open-ended traditional categorization measure in which they imagined going out to dinner at a restaurant serving their selected cuisine and described what they would expect to see on the menu. Following previous categorization research (Medin, Lynch, Coley, & Atran, 1997; Rosch, Mervis, Gray, Johnson, & Boyes-Braem, 1976), each discrete food item was scored according to category level (superordinate, basic, subordinate, or uncodable). Category level served as a guide to coders, orienting them what to look for. We did not anticipate, nor did we find, variation as a function of category level, so a tabulation across all levels resulted in a single index of category volume. Two independent coders scored their responses; disagreement was resolved through discussion. Finally, participants completed the 13-item Maximization Scale (Schwartz et al., 2002). Sample items include, “I never settle for second best” and “Whenever I’m faced with a choice, I try to imagine what all the other possibilities are, even ones that aren’t present at the moment” (α = .72).

Results

Manipulation check

The four manipulation check questions were averaged to form an overall preference rating (α = .95). The valence manipulation was successful, in that liked cuisine was rated more favorably than disliked cuisine (M = 4.55, SD = 0.49 vs. M = 2.18, SD = 0.89), F(1, 93) = 258.18, p < .001, η² = .74, 95% CI = [−2.67, −2.08].

Categorization measures

To examine our prediction that maximization would moderate the effect of valence on the type of evaluative information sought out, we used regression procedures recommended by Cohen and Cohen (1983). We entered valence (disliked vs. liked cuisine) and maximization in the first step, and the valence by maximization interaction term in the second step (Aiken & West, 1991). Unstandardized regression coefficients are from the second step (see Figure 2).

Figure 2.

Amount of evaluative information (in the form of evaluative rating scales) sought out as a function of valence and maximization score (maximizers vs. satisficers) in Experiment 2.

For the model predicting the type of evaluative information (in the form of rating scales) participants would want for their selected cuisine, there was a significant effect of valence, b = 2.83, t(91) = 2.06, p = .043, 95% CI = [0.10, 5.57], and a significant effect of maximization, b = 0.60, t(91) = 2.72, p = .008, 95% CI = [0.16, 1.04]. This was qualified by a significant valence by maximization interaction, b = −0.64, t(91) = −1.98, p = .05, 95% CI = [−1.28, 0.00]. To further interpret this interaction, we compared differences in simple slopes. Both maximizers and satisficers sought out similarly differentiated evaluative information for their liked cuisine type, b = −0.04, t(91) = −0.16, p = .872. However, maximizers sought out significantly more differentiated evaluative information for their disliked cuisine type when compared with satisficers, b = 0.60, t(91) = 2.72, p = .008. To determine whether maximizers and satisfices preferred more evaluative scale points for liked (vs. disliked) cuisine, we used the Johnson–Neyman (1936) technique through PROCESS (Hayes, 2013) to determine the range of significance for the moderator. This allowed us to determine at what levels of the moderator (maximization) the two conditions (liked vs. disliked cuisine) significantly differed. This was used instead of the pick-a-point approach as we did not identify any nonarbitrary points to probe (see Hayes & Matthes, 2009). Results showed that the effect of valence on evaluative information seeking was significant when the maximization score was 2.82 or less (i.e., on the satisficing end of the scale), b = 1.03, 95% CI = [0.00, 2.06], t = 1.99, p = .05. Because the maximization index had a possible range of 1 to 7, only moderate to extreme satisficers preferred more evaluative scale points for liked (vs. disliked) categories.

A second model was run predicting the number of categories included in the traditional categorization task. There was a significant main effect of valence, b = 8.73, t(92) = 2.22, p = .029, 95% CI = [0.93, 16.54]; a nonsignificant main effect of maximization, b = .33, t(92) = .53, p = .60, 95% CI = [−0.92, 1.58]; and a nonsignificant valence by maximization interaction, b = −1.52, t(92) = −1.65, p = .10, 95% CI = [−3.35, 0.31]. Replicating previous findings (Smallman et al., 2014; Smallman & Roese, 2008), participants listed more categories in the liked (vs. disliked) condition (M = 5.87, SD = 4.44 vs. M = 3.43, SD = 3.16).

Discussion

Results from Experiment 2 support the idea that choice optimization affects the preference-categorization effect. Consistent with predictions, maximizers sought out finely differentiated evaluative information for both desirable and undesirable stimuli, whereas satisficers only sought out detailed evaluative information for desirable stimuli, preferring less nuanced evaluative information for undesirable stimuli. This supports our hypothesis that satisficers would be more likely to show the preference-categorization effect. Maximizers are intrinsically motivated to seek out the best option (regardless of valence) and use the finely differentiated evaluative information to aid in that decision. In contrast, satisficers will only move beyond their minimal threshold when externally motivated (in this case by desire for a liked item). This is consistent with prior research suggesting that maximizers are more likely to attend to and use more choice-relevant information (Chowdhury et al., 2009; Iyengar et al., 2006; Schwartz et al., 2002). Interestingly, this pattern did not replicate in the traditional categorization task. This may indicate that this moderating effect is specific to choice decision-making tasks (e.g., evaluative categorization) rather than the mental schematic of the category more generally.

Experiment 3

The preference-categorization effect suggests that people make more nuanced evaluative distinctions for liked (vs. disliked) items. So far, Experiments 1 and 2 have shown that this effect extends to the type of evaluative information people seek out, and that certain situational factors (i.e., choice constraint) and individual differences (i.e., punishment sensitivity and choice optimization) can impact this effect. However, the type of evaluative information examined thus far has been fairly narrow, focusing only on overall evaluative information (i.e., how good/bad is this item). If the preference-categorization effect speaks broadly to how individuals interact with evaluative information, then we would expect a similar pattern for other types of evaluative information. In addition, Experiments 1 and 2 used response labels for each scale point (see the appendix). Given that the scale point labels differ across valence, it is possible that previous results were driven not by a desire for more differentiated scale points for evaluating liked (vs. disliked) items but instead by the asymmetrical response labels provided for these evaluations. Experiment 3 addresses these questions by examining a different type of evaluative information, attributes. Evaluative attributes focus on the number of facets or dimensions a person would want in an item’s evaluation. In Experiment 3, valence was manipulated on a between-subjects basis; participants focused on either liked or disliked items for five different domains (food, clothing, music, university courses, television shows). The dependent variable was the amount of evaluative information (in the form of attributes) that they would want to have as part of the item’s evaluations. Importantly, as the evaluative attributes were identical for liked and disliked items, any evidence of the preference-categorization effect cannot be attributed to asymmetric response labels.

Method

Undergraduate students (N = 182; age M = 18.65 years, SD = 0.95; 52.0% female) participated for course credit. Sample size was based on a target minimum of 50 per cell. Given that lab sessions are scheduled on a weekly basis, data collection was terminated at the end of the week in which this minimum was exceeded. A power analysis (Gpower3.1) showed that a sample size of 182 provides 99.97% statistical power to detect a large effect size (d = 0.80), 91.84% statistical power to detect a medium effect size (d = 0.50), and 26.87% statistical power to detect a small effect size (d = 0.20). In a between-subjects design, participants were randomly assigned to think about liked or disliked items in five different domains (food, clothing, music, university courses, and television shows). Participants were instructed that the study focused on how people determine what attributes are important as part of an item’s evaluation. Next, they were given a domain (e.g., food) and asked to focus only on items within that domain that they either liked or disliked. Then, they were given a list of seven attributes for that domain and asked to select which attributes they would want to see as part of that item’s evaluation. This was repeated for all five domains. As in previous research on product attributes (e.g., Yeung & Wyer, 2004) attributes were selected from a pilot study (N = 157; age M = 18.53 years, SD = .90; 68.2% female) in which participants were asked open-ended questions about what attributes they would need to rate either liked or disliked items within each domain. The current study utilized seven commonly listed attributes for each domain (see the appendix).

Results

Across five domains, participants sought out more evaluative attributes for liked versus disliked items (M = 4.83, SD = 0.94 vs. M = 4.09, SD = 1.3), F(1, 180) = 19.79, p < .001, η² = .10, 95% CI = [−5.38, −2.08]. The same pattern was significant in each of the five domains (all Fs > 4.5, all ps < .04)

Discussion

Results from Experiment 3 were consistent with our hypothesis. This provides initial evidence that the preference-categorization effect affects how individuals seek out a new type of evaluative information, evaluative attributes. Importantly, both liked and disliked items conditions were provided identical evaluative information (i.e., both conditions received the same list of attributes to select from). Given that previous research has focused narrowly on overall evaluative information (i.e., general ratings of how good or bad this item is) that typically provided labels at each scale point, the current finding rules out the alternative that previous results were due to differences in liked/disliked scale point labels, instead suggesting that the preference-categorization effect speaks more broadly to how individuals interact with evaluative information regarding liked and disliked items.

General Discussion

The preference-categorization effect states that people make more nuanced distinctions for liked versus disliked items for both traditional and evaluative categorization (Smallman et al., 2014; Smallman & Roese, 2008). That is, not only do people draw finer categorical distinctions for liked than disliked objects, but they also prefer finer evaluative distinctions to convey their opinions. Underlying this effect is the idea that individuals have a differential motivation to seek out, process, and use information related to liked (vs. disliked) items. Across three studies, the current research extends previous findings by showing that this differential motivation influences how individuals seek out finely differentiated evaluative information about these items.

First, we found support for the hypothesis that freedom of choice, or the ability to avoid undesirable experiences, is an important factor underlying the preference-categorization effect. However, these effects were contingent on an individual’s sensitivity to negative stimuli. Second, choice optimization, or the desire to achieve the best possible outcome (i.e., maximize), moderated the preference-categorization effect. Finally, we showed that this differential pattern generalized to another type of evaluative categorization, evaluative attributes. As with rating scales, individuals sought out more attribute information for liked versus disliked items. Together, this research suggests that the preference-categorization effect influences not only the kinds of evaluative distinctions people prefer to express their own opinions but also the type of evaluative information they will seek out in making their own decisions.

A key aspect of the preference-categorization effect is the freedom to select experiences. For most situations, we willingly choose (or consciously avoid) interactions involving liked (or disliked) objects. This idea is consistent with work demonstrating that affective reactions are tightly coupled with motivations to approach or avoid stimuli (Damasio, 2003; Slovic & Peters, 2006). Positive affective reactions (such as preference) are associated with approach motivation; they urge us to draw nearer and focus on the desired object. In contrast, negative affective reactions (such as dislike) are associated with avoidance motivation. In this case, successful avoidance forecloses further opportunities for elaborative processing. In contrast, certain situations exist where we have little choice about whether we interact with particular objects. In these cases, preference does not drive behavior; regardless of our feelings toward these objects, we are “stuck with” them. This suggests choice constraint as a boundary condition; only when behavior is freely chosen can preference influence categorization. In Experiment 1, participants who could freely avoid undesirable outcomes and were especially sensitive to negative stimuli (i.e., high BIS) demonstrated the most pronounced preference-categorization effect. When freedom of choice was constrained, these same individuals sought out finely differentiated evaluative information for both desirable and undesirable outcomes. This finding points to a necessary condition for the preference-categorization effect. Preference invites categorization only when there is freedom to select experiences. However, this mechanism hinges on an individual’s reactions to negative stimuli more generally.

We tested an additional hypothesis that preference-categorization may result from pursuit of desirable experiences. Thus, participants with higher reward sensitivity (BAS) might have shown the preference-categorization effect regardless of freedom of choice. We did not find an effect for BAS, which may suggest that the preference-categorization effect results more from the avoidance of undesirable stimuli than the pursuit of desirable stimuli. However, it may also be the case that most individuals pursue desirable situations, so the additional motivation from high BAS does not further influence the preference-categorization effect. Alternatively, the null effect may be due to particular methodological choices in this study. For example, an item response theory (IRT) analysis of the BIS/BAS scale conducted by Gomez, Cooper, and Gomez (2005) suggests that BAS Drive and BAS Fun Seeking are “poor for measuring the appropriate traits at moderately high to high levels.” Therefore, BAS may influence preference-categorization but only at extremely high levels that were not effectively measured. Relatedly, the effect of BAS on preference-categorization may be a small effect size, which the current study did not have sufficient power to measure. Therefore, additional research is necessary to better examine the relationship between the preference-categorization effect and individual differences in approach motivation.

Whereas previous preference-categorization research has focused primarily on consumer products, Experiment 1 generalizes this effect by examining a new context, medical procedures. Consequently, our findings complement research identifying situational factors and individual differences that can impact health-relevant information seeking (Sweeny, Melnyk, Miller, & Shepperd, 2010). Most relevant to the current findings is research on individual differences in information coping styles (i.e., monitors and blunters; Miller, 1987, 1995). Accordingly, monitors typically seek out threat-relevant information about aversive events. In contrast, blunters typically cope with aversive events by distracting themselves from threat-relevant information. Applied to the preference-categorization effect, we might expect monitors to seek out finely differentiated evaluative information, regardless of valence. Yet, blunters might be more likely to show the expected preference-categorization effect. They could avoid threat-relevant information by selecting less differentiated negative evaluative information. Future research may provide insight into new ways to present evaluative health information.

Differences in choice motivation can be construed as either situational factors (Experiment 1) or as individual differences (Experiments 1 and 2). Based on Schwartz and colleagues (2002) work on maximization, Experiment 2 examined whether being a maximizer or a satisficer would impact the preference-categorization effect. Whereas maximizers have internal aspirations for high standards and tend to seek out “the best” alternative, satisficers focus mainly on finding a “good enough” solution that satisfies a minimum requirement unless externally motivated. In Experiment 2, satisficers showed the predicted preference-categorization effect, only seeking out finely differentiated evaluative information for liked items. In contrast, maximizers sought out finely differentiated evaluative information for both liked and disliked items, ensuring that they not only get the “best of the best” but also the “best of the worst.” Interestingly, this difference was only evident on the evaluative categorization measure; maximization did not influence the traditional categorization measure. This may indicate that this effect is specific to decision-making tasks (e.g., evaluative categorization) rather than more general mental representations of liked and disliked items.

This result is similar to previous work on need for cognition (NFC; Cacioppo & Petty, 1982) and the preference-categorization effect (Smallman et al., 2014). Accordingly, low NFC individuals (like satisficers) showed the preference-categorization effect whereas high NFC individuals (like maximizers) did not. However, there was one key difference. For NFC, this pattern was evident in both the evaluative and traditional categorization tasks. This may speak to the fact that NFC is a more general difference in thought processes, and is not just relevant to choice. To be sure, these measures are related, with research showing a small to medium correlation (Lai, 2010; Nenkov et al., 2008). In the context of the preference-categorization effect, both maximizers and high NFC are intrinsically motivated to process elaboratively and seek out additional choice-relevant information. However, the effect of maximization on categorization may be limited to choice-relevant tasks (e.g., evaluative categorization) whereas the effect of NFC may be seen more broadly in a variety of categorization tasks.

The scope of previous evaluative categorization research has been fairly narrow, focusing only on the preference for overall evaluative rating scales to convey one’s opinion about liked or disliked consumer products. The current findings broaden this research in a number of ways. First, it shows that the preference-categorization effect extends to the kinds of evaluative information individuals seek out. This makes new connections between the preference-categorization effect and research on information seeking, such as selective exposure and feedback preference (Afifi, Dillow, & Morse, 2004; Hart et al., 2009; Trope, Gervey, & Bolger, 2003). Second, Experiment 1 shows that the preference-categorization effect occurs when looking at the desirable and undesirable aspects of a single item. Given the potential impact of differences in overall knowledge of liked and disliked items, establishing that the preference-categorization effect occurs while holding the item constant and shifting only the focus on either desirable or undesirable aspects of the item is important. Third, Experiment 3 examined a new type of evaluative information, evaluative attributes. If the preference-categorization effect reflects differences in how we represent and evaluate positive and negative concepts, then we should expect to see this effect in evaluative information beyond overall evaluative rating scales. Future research should continue to examine new types of evaluative information.

Experiment 3 also provided evidence against an alternative explanation for the preference-categorization effect based on asymmetrical scale labels. Given research showing that participant responses are sensitive to response labels (Weijters, Cabooter, & Schillewaert, 2010; Weijters, Geuens, & Baumgartner, 2013), it is possible that the resulting difference in preferred scale points has less to do with a desire for more finely differentiated evaluative information and more to do with an affective or qualitative difference in positive and negative scale point labels. Although our scale labels were selected from the most frequent scale labels for each point provided by previous participants (Smallman et al., 2014), this does not necessarily mean that qualitative differences do not exist between the positive and negative scale points. However, additional evidence makes this alternative explanation seem unlikely. First, Experiment 3 replicated the preference-categorization effect using evaluative attributes. In this case, identical attributes were provided for liked and disliked items. Even with the evaluative information held constant, participants still preferred more finely differentiated evaluative information for liked items. In addition, in an earlier study (Experiment 1a, Smallman et al., 2014), participants provided the number of stars they would want to evaluate liked or disliked items (as opposed to selecting from preconstructed scales with labels). A similar result was obtained; participants wanted more stars for overall evaluations of liked items. Accordingly, we believe that our effect is driven more by a desire for finely differentiated information for liked (vs. disliked) items, rather than an asymmetry in scale point labels. However, future research should consider ways in which to avoid this concern, by either eliminating response labels, limiting them to qualitatively similar scale end-points (e.g., neutral and best; neutral and worst), or considering alternative operationalizations of evaluative categorization.

The current research examines people’s evaluative preferences when interacting with liked and disliked items, with a particular focus on consumer products or experiences (e.g., food, clothing, music). That people prefer more finely differentiated evaluative information for liked versus disliked items poses new questions for marketing and consumer research. Of particular interest is whether these preferences correlate with actual consumer behavior. That is, if a preference for finer evaluative distinctions is in part due to increased interactions with desired items, then measuring this preference for finer evaluative distinctions may help predict future consumer behavior. Given that consumer researchers often use these types of scales to determine customer satisfaction, it might also convey useful information for predicting future behavior. In addition, these results may be useful in considering how to frame evaluative information, such that desired evaluative information may be more appealing when presented in a more nuanced matter.

An additional next step would be to test people’s preferences when evaluating other individuals. Using a traditional categorization task, the preference-categorization effect shows that people sort liked (vs. disliked) objects into more categories (Smallman & Roese, 2008). This suggests that more nuanced distinctions are made for liked objects. Similarly, out-group homogeneity effect (OHE) research shows that in-group (vs. out-group) members are seen as more distinct from one another (Linville & Jones, 1980; Park & Judd, 1990; Park & Rothbart, 1982; Quattrone & Jones, 1980; Rubin & Badea, 2012). Both disliked objects, and out-group members seem to be characterized as being “all the same” leading to fewer nuanced distinctions attributed to individual members of those groups. Conversely, in-group members, similar to liked objects, are seen as more variable. OHE may, in part, be explained by preference in that in-group members are generally liked more than out-group members (Brewer, 1979; Nesdale & Flesser, 2001). If evaluative categorization extends to how people rate other individuals, there are potential implications for situations in which decisions are based on the evaluations of others.

Overall, we have shown that the preference-categorization effect extends to the type of evaluative information people seek out, as measured with overall rating scales and evaluative attributes. However, certain situational factors (i.e., choice constraint) and individual differences (i.e., punishment sensitivity and choice optimization) can impact this effect. Accordingly, future replications of the basic effect should be mindful that their stimuli and instructions do not introduce unwanted situational factors that might moderate the preference-categorization effect (e.g., having participants evaluate liked/disliked necessity items). In addition, existing preference-categorization research has been conducted primarily on consumer goods (e.g., food, clothes, music), with the current research providing initial findings using nonconsumer items (e.g., medical treatments, university courses). Although we believe that the effect should extend beyond consumer items, we are cautious about generalizability until future research is conducted. Finally, our samples included a broad range of participants, by collecting data from both university students and Mturk participants. Therefore, we believe that results can be reproduced in samples of various ages, gender distributions, and education levels. However, our participants were restricted to individuals currently living in the United States. Therefore, these findings may not generalize to non-Western populations.

Footnotes

Appendix

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Supplemental Material

The supplemental material is on the PSPB website.

Notes

References

Afifi

W. A.

Dillow

M. R.

Morse

(2004). Examining predictors and consequences of information seeking in close relationships. Personal Relationships, 11, 429-449.

Aiken

L. S.

West

S. G.

(1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: SAGE.

Bechara

Damasio

Tranel

Damasio

A. R.

(1997). Deciding advantageously before knowing the advantageous strategy. Science, 275, 1293-1295.

Brewer

M. B.

(1979). In-group bias in the minimal intergroup situation: A cognitive-motivational analysis. Psychological Bulletin, 86, 307-324.

Cacioppo

J. T.

Petty

R. E.

(1982). The need for cognition. Journal of Personality and Social Psychology, 42, 116-131.

Carver

C. S.

(2003). Pleasure as a sign you can attend to something else: Placing positive feelings within a general model of affect. Cognition and Emotion, 17, 241-261.

Carver

C. S.

White

T. L.

(1994). Behavioral inhibition, behavioral activation, and affective responses to impending reward and punishment: The BIS/BAS scales. Journal of Personality and Social Psychology, 67, 319-333.

Chowdhury

T. G.

Ratneshwar

Mohanty

(2009). The time-harried shopper: Exploring the differences between maximizers and satisficers. Marketing Letters, 20, 155-167.

Cohen

(1983). Applied multiple regression/correlation analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

10.

Damasio

(2003). Looking for Spinoza: Joy, sorrow, and the feeling brain. Orlando, FL: Harcourt.

11.

Dar-Nimrod

Rawn

C. D.

Lehman

D. R.

Schwartz

(2009). The maximization paradox: The costs of seeking alternatives. Personality and Individual Differences, 46, 631-635.

12.

Dawson

J. F.

Richter

A. W.

(2006). Probing three-way interactions in moderated multiple regression: Development and application of a slope difference test. Journal of Applied Psychology, 91, 917-926.

13.

Gable

P. A.

Harmon-Jones

(2008). Approach-motivated positive affect reduces breadth of attention. Psychological Science, 19, 476-482.

14.

Garner

W. R.

(1960). Rating scales, discriminability, and information transmission. Psychological Review, 67, 343-352.

15.

Gomez

Cooper

Gomez

(2005). An item response theory analysis of the Carver and White (1994) BIS/BAS Scales. Personality and Individual Differences, 39, 1093-1103.

16.

Gray

J. A.

(1981). A critique of Eysenck’s theory of personality. In Eysenck

H. J.

(Ed.), A model for personality (pp. 246-276). Berlin, Germany: Springer-Verlag.

17.

Gray

J. A.

(1990). Brain systems that mediate both emotion and cognition. Cognition and Emotion, 4, 269-288.

18.

Hart

Albarracin

Eagly

A. H.

Brechan

Lindberg

Merrill

(2009). Feeling validated versus being correct: A meta-analysis of selective exposure to information. Psychological Bulletin, 135, 555-588.

19.

Hayes

A.F.

(2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York: Guilford Press

20.

Hayes

A. F.

Matthes

(2009). Computational procedures for probing interactions in OLS and logistic regression: SPSS and SAS implementations. Behavior Research Methods, 41, 924-936.

21.

Hoch

S. J.

Deighton

(1989). Managing what consumers learn from experience. Journal of Marketing, 52, 1-20.

22.

Hofmann

De Houwer

Perugini

Baeyens

Crombez

(2010). Evaluative conditioning in humans: A meta-analysis. Psychological Bulletin, 136, 390-421.

23.

Iyengar

S. S.

Wells

R. E.

Schwartz

(2006). Doing better but feeling worse: Looking for the “best” job undermines satisfaction. Psychological Science, 17, 143-150.

24.

Johnson

P. O.

Neyman

(1936). Tests of certain linear hypotheses and their applications to some educational problems. Statistical Research Memoirs, 1, 57–93.

25.

Komorita

S. S.

Graham

W. K.

(1965). Number of scale points and the reliability of scales. Educational and Psychological Measurement, 25, 987-995.

26.

Lai

(2010). Maximizing without difficulty: A modified maximizing scale and its correlates. Judgment and Decision Making, 5, 164-175.

27.

Lefkoff-Hagius

Mason

C. H.

(1993). Characteristic, beneficial, and image attributes in consumer judgments of similarity and preference. Journal of Consumer Research, 20, 100-110.

28.

Linville

P. W.

Fischer

G. W.

(1993). Exemplar and abstraction models of perceived group variability and stereotypicality. Social Cognition, 11, 92-125.

29.

Linville

P. W.

Jones

E. E.

(1980). Polarized appraisals of out-group members. Journal of Personality and Social Psychology, 38, 689-703.

30.

Roese

N. J.

(2014). The maximizing mind-set. Journal of Consumer Research, 41, 71-92.

31.

Medin

D. L.

Lynch

E. B.

Coley

J. D.

Atran

(1997). Categorization and reasoning among tree experts: Do all roads lead to Rome? Cognitive Psychology, 32, 49-96.

32.

Miller

S. M.

(1987). Monitoring and blunting: Validation of a questionnaire to assess styles of information seeking under threat. Journal of Personality and Social Psychology, 52, 345-353.

33.

Miller

S. M.

(1995). Monitoring versus blunting styles of coping with cancer influence the information patients want and need about their disease: Implications for cancer screening and management. Cancer, 76, 167-177.

34.

Mukherjee

Hoyer

W. D.

(2001). The effect of novel attributes on product evaluation. Journal of Consumer Research, 28, 462-472.

35.

Nenkov

G. Y.

Morrin

Ward

Schwartz

Hulland

(2008). A short form of the maximization scale: Factor structure, reliability and validity studies. Judgment and Decision Making, 3, 371-388.

36.

Nesdale

Flesser

(2001). Social identity and the development of children’s group attitudes. Child Development, 72, 506-517.

37.

Park

Judd

C. M.

(1990). Measures and models of perceived group variability. Journal of Personality and Social Psychology, 59, 173-191.

38.

Park

Rothbart

(1982). Perception of out-group homogeneity and levels of social categorization: Memory for the subordinate attributes of in-group and out-group members. Journal of Personality and Social Psychology, 42, 1051-1068.

39.

Parker

A. M.

de Bruin

Fischhoff

(2007). Maximizers versus satisficers: Decision-making styles, competence, and outcomes. Judgment and Decision Making, 2, 342-350.

40.

Petty

R. E.

Cacioppo

J. T.

(1986). The elaboration likelihood model of persuasion. In Berkowitz

(Ed.), Advances in experimental social psychology (Vol. 19, pp. 123-205). New York, NY: Academic Press.

41.

Polman

(2010). Why are maximizers less happy than satisficers? Because they maximize positive and negative outcomes. Journal of Behavioral Decision Making, 23, 179-190.

42.

Quattrone

G. A.

Jones

E. E.

(1980). The perception of variability within in-groups and out-groups: Implications for the law of small numbers. Journal of Personality and Social Psychology, 38, 141-152.

43.

Rosch

Mervis

C. B.

Gray

W. D.

Johnson

D. M.

Boyes-Braem

(1976). Basic objects in natural categories. Cognitive Psychology, 8, 382-439.

44.

Rubin

Badea

(2012). They’re all the same!. . . but for several different reasons: A review of the multicausal nature of perceived group variability. Current Directions in Psychological Science, 21, 367-372.

45.

Schwartz

Ward

Monterosso

Lyubomirsky

White

Lehman

D. R.

(2002). Maximizing versus satisficing: Happiness is a matter of choice. Journal of Personality and Social Psychology, 83, 1178-1197.

46.

Shen

Dillard

J. P.

(2007). The influence of behavioral inhibition/approach systems and message framing on the processing of persuasive health messages. Communication Research, 34, 433-467.

47.

Slovic

Peters

(2006). Risk perception and affect. Current Directions in Psychological Science, 15, 322-325.

48.

Smallman

Becker

Roese

N. J.

(2014). Preferences for expressing preferences: People prefer finer evaluative distinctions for liked than disliked objects. Journal of Experimental Social Psychology, 52, 25-31.

49.

Smallman

Roese

N. J.

(2008). Preference invites categorization. Psychological Science, 19, 1228-1232.

50.

Sweeny

Melnyk

Miller

Shepperd

J. A.

(2010). Information avoidance: Who, what, when, and why. Review of General Psychology, 14, 340-353.

51.

Thorndike

E. L.

(1898). Animal intelligence: An experimental study of the associative processes in animals. Psychological Review Monograph Supplement, 2, 1-109.

52.

Trope

Gervey

Bolger

(2003). The role of perceived control in overcoming defensive self-evaluation. Journal of Experimental Social Psychology, 39, 407-419.

53.

Weijters

Cabooter

Schillewaert

(2010). The effect of rating scale format on response styles: The number of response categories and response category labels. International Journal of Research in Marketing, 27, 236-247.

54.

Weijters

Geuens

Baumgartner

(2013). The effect of familiarity with the response category labels on item response to Likert scales. Journal of Consumer Research, 40, 368-381.

55.

Weng

(2004). Impact of the number of response categories and anchor labels on coefficient alpha and test-retest reliability. Educational and Psychological Measurement, 64, 956-972.

56.

Yeung

C. W.

Wyer

R. S.

(2004). Affect, appraisal, and consumer judgment. Journal of Consumer Research, 31, 412-424.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.46 MB