U10_dq

Positioning Rationality and Emotion: Rationality Is Up and Emotion Is Down

LUCA CIAN ARADHNA KRISHNA NORBERT SCHWARZ

Emotion and rationality are fundamental elements of human life. They are abstract concepts, often difficult to define and grasp. Thus throughout the history of Western society, the head and the heart, concrete and identifiable elements, have been used as symbols of rationality and emotion. Drawing on the conceptual met- aphor framework, we propose that people understand the abstract concepts of ra- tionality and emotion using knowledge of a more concrete concept—the vertical difference between the head and heart. In six studies, we show a deep-seated conceptual metaphorical relationship linking rationality with “up” or “higher” and emotion with “down” or “lower.” We show that the association between verticality and rationality/emotion affects how consumers perceive information and thereby has downstream consequences on attitudes and preferences. We find the associ- ation to be most influential when consumers are unaware of it and when it applies to an unfamiliar stimulus. Because all visual formats—from the printed page to screens on a television, computer, or smartphone—entail a vertical placement, this association has important managerial implications. Our studies implement multiple methodologies and technologies and use manipulations of logos, web- sites, food advertisements, and political slogans.

Keywords: rationality, emotion, vertical placement, conceptual metaphor, market-

ing communication, visual persuasion, object positioning

E motion and rationality are fundamental elements ofhuman life. Their definition, their functioning, and their relation are ageless questions that have preoccupied humans for thousands of years (Kirman, Livet, and Teschl

2010). Emotions greatly affect the quality and meaning of

our existence, whereas rationality helps us organize our be-

liefs and optimize the outcomes of our actions (de Sousa

2014). And yet rationality and emotions are not indepen-

dent of each other (Cosmides and Tooby 2000; De Sousa

1990) because emotions are, in fact, often rational.

However, scientific evidence concerning the interplay

between rationality and emotions, on the one hand, and

people’s perception and naive understanding of rationality

and emotions, on the other, are different issues. How do we

understand rationality and emotion—concepts that we can

neither touch nor see? The conceptual metaphor frame-

work (Lakoff and Johnson 1999; Landau, Meier, and

Keefer 2010) asserts that people understand abstract do-

mains via more concrete domains of knowledge. That is,

according to a metaphor-enriched perspective, people con-

ceptualize abstract concepts in terms of more concrete

concepts. Applying the conceptual metaphor framework and idea

of scaffolding to rationality and emotion, we note that from

Luca Cian is assistant professor of marketing at the Darden School of

Business, University of Virginia, Charlottesville, VA 22903,

[email protected]. Aradhna Krishna is the Dwight F. Benton professor

of marketing at the Ross School of Business, University of Michigan, 701

Tappan Avenue, Ann Arbor, MI 48109, [email protected]. Norbert

Schwarz is Provost Professor of psychology and marketing, Department

of Psychology & Marshall School of Business, University of Southern

California, 3620 South McClintock Ave., Los Angeles, CA 90089-1061,

[email protected]. Address correspondence to Luca Cian,

Aradhna Krishna, and Norbert Schwarz. Authors contributed equally to

the research; author names are in alphabetical order. The authors acknowl-

edge the valuable input of the editor, associate editor, and reviewers.

Laura Peracchio and Gita Johar served as editors, and Andrea Morales

served as associate editor for this article.

Advance Access publication September 18, 2015

VC The Author 2015. Published by Oxford University Press on behalf of Journal of Consumer Research, Inc.

All rights reserved. For permissions, please e-mail: [email protected] � Vol. 42 � 2015 DOI: 10.1093/jcr/ucv046

632

childhood, humans tend to associate two “concrete” body parts—the head and the heart—with the more “abstract” concepts of rationality and emotion, respectively. For in- stance, in The Wizard of Oz, the Tin Woodman desires a heart because he is without emotions, and Scarecrow de- sires a brain because he lacks intelligence. In many chil- dren’s books and movies, the same connection is made, and a character often points to his head when thinking and to his heart when expressing love. Thus over time we de- velop the conceptual link of rationality with “up” or “higher” and emotion with “down” or “lower.” This argu- mentation also has historical support from the ancient times of Plato’s and Virgil’s teachings, to Aquinas’s medi- eval time’s philosophy to the more recent writings of Shakespeare and Descartes. Although people structure ab- stract concepts metaphorically, these metaphorical links are not necessarily expressed linguistically (Casasanto 2009). In the case of the association of rationality with up and emotion with down, some linguistic examples follow the conceptual metaphor. Indeed, in normal conversation, figures of speech depicting rational versus emotional be- haviors often invoke verticality (e.g., “the discussion fell to the emotional level, but I raised it back up to the rational plane”).

Although rationality and emotion as drivers of human behavior have been intensively researched in consumer be- havior, decision making, economics, neuroscience, philos- ophy, and psychology (e.g., Damasio 1994; Evans and Cruse 2004; Fetterman and Robinson 2013; Greenspan 1988; Pham 2007; Shiv and Fedorikhin 1999; Zajonc 2000), previous research has not examined their relation- ship with physical verticality. A metaphorical link between verticality and rationality/emotion would cause the same information to be perceived as more rational or more emo- tional depending on where it was placed along the vertical dimension. The link may also create expectations for where on a plane people would presume to see more rational or more emotional stimuli. The metaphorical association would have implications in many domains because all vi- sual presentation formats—from the printed page to screens on a television, computer, or smartphone—entail a vertical placement. We test for the proposed metaphorical association in six studies. These studies test (1) the exis- tence of verticality’s association with rationality/emotion, (2) the properties of this association, (3) the boundary con- ditions of the association, and (4) the association’s down- stream effects.

We first test the existence of a proposed conceptual met- aphor and show that people do, indeed, associate up with rational and down with emotional. Second, we explore the association’s properties. Specifically, we find the meta- phorical link to be bidirectional such that spatial placement can influence perceptions of rationality versus emotion, much as the rational versus emotional quality of material influences its preferred spatial placement. Third, we show

the association is most influential when consumers are un- aware of it and when they are less familiar with the stimu- lus. Finally, we demonstrate that the manipulation of this association affects how consumers perceive information and has downstream consequences on attitudes and prefer- ences. We begin by discussing literature pertinent to our research. We then build the conceptual framework for our hypothesized effects. We next describe six studies. We conclude by addressing specific contributions of the re- search and future directions.

THEORETICAL BACKGROUND AND CONCEPTUAL FRAMEWORK

Literature relevant to our research falls in the areas of con- ceptual metaphors and scaffolding, prior metaphorical as- sociations connected to verticality (valence, power, and morality), verticality’s association with rationality/emo- tion, effects of the metaphorical match on attitudes, and possible moderating effects for verticality’s association with rationality/ emotion. We briefly discuss each in turn.

Conceptual Metaphors and Scaffolding

Conceptual Metaphors.Cognitive linguists Lakoff and Johnson (1980, 1999) argue that a metaphor is a cognitive tool that allows people to communicate what they cannot see, taste, hear, smell, or touch. In their view, a conceptual metaphor can be understood as a mapping (a precise set of correspondences) from a source domain to a target domain (Lakoff 1993). The source domain represents the concrete knowledge that an individual gleans from experiences with the physical world (e.g., love is a journey), whereas the tar- get domain represents the conceptual and abstract domain that is more difficult to grasp (e.g., love is a journey; Lakoff and Johnson 1980). These mental associations al- low people to use pieces of knowledge coming from a con- crete and commonplace source domain (e.g., journey) to interpret, understand, and manipulate information associ- ated with the abstract target domain (e.g., love). For exam- ple, the metaphor might lead one to think of a lover as a fellow traveler, or of a shared life goal as a destination. Such metaphors guide how people interpret relationship events and the same event can take on different meaning, with different downstream consequences, depending on which metaphor is brought to bear on it (Lee and Schwarz 2014b).

Literature in psychology and consumer behavior is rich in works on conceptual metaphors, showing how subtle in- cidental physical and sensory experiences can uncon- sciously affect metaphorically related targets. Physical height is one type of concrete experience that has come to be linked metaphorically to a number of abstract concepts. Position in vertical space (i.e., that some stimuli are physi- cally higher than others) is omnipresent. Thus it is coopted

CIAN ET AL. 633

for multiple metaphorical associations (Lakoff and Johnson 1999; Meier et al. 2007), such as valence, power, and morality. Testing for the link between verticality with valence, Meier and Robinson (2004) had participants eval- uate words presented on a computer screen. Results showed that participants evaluated positive words faster when they were on the top of the screen versus the bottom. For negative words, the effect was the opposite. In terms of power, Schubert (2005) found that people were quicker to recognize stimuli representing power when they appeared at the top versus the bottom of the page. Similarly, in a se- ries of studies, Giessner and Schubert (2007) changed the length of a vertical line that separated the leader from the rest of the employees in an organization chart, and then they asked participants to estimate the leader’s power. The authors found that as the length of the line increased, so did participants’ estimation of the leader’s power. With re- spect to morality, Meier, Sellbom, and Wygant (2007) found that people more quickly recognize words with a moral meaning (e.g., caring, charity, nurture, truthful, and trustworthy) when those words appear at the top of the screen. Conversely, people more quickly recognize words with an immoral meaning (e.g., adultery, corrupt, dishon- est, evil, and molest) when those words appear at the bot- tom of the screen. Similarly, Meier et al. (2007) found that participants more quickly recognized God-related words when they were associated with up-related words and that the reverse occurred for devil-related words. One needs to be cognizant of these associations when testing for a new metaphorical link with verticality.

Scaffolding. Consistent with the conceptual metaphor framework is the concept of scaffolding (Williams, Huang, and Bargh 2009). Scaffolding, a natural process through which people integrate new concepts with extant knowl- edge structures (Bargh 2006), proposes that abstract con- cepts arise from an infant’s understanding of its physical environment (Clark 1973; Mandler 1992). Specifically, people acquire physical concepts during their infancy and childhood, and as they become older, they develop more complex and abstract knowledge structures around these early direct experiences, such that the meanings of the older, more basic concept and the newer, higher level con- cept are linked (Williams et al. 2009; Williams and Bargh 2008a). For example, before children can comprehend the abstract meaning of psychological distance (i.e., interper- sonal, emotional), they must perceptually experience these concepts, which they do through physical spatial distance (Williams and Bargh 2008b). Over time, with an incidental activation of the more basic physical concept (which may even occur outside conscious experience), the higher level concept is likely to be activated automatically (Bargh 2006; Williams et al. 2009). For example, priming people with spatial distance (plotting two points on a Cartesian plane) affects how they judge the emotional distance with

their family, such that the farther people plotted these points from each other, the weaker were the familial bonds they reported (Williams and Bargh 2008b).

Applying the conceptual metaphor framework and idea of scaffolding to verticality and its associations, we suggest that people have developed the link of “up is rational, and down is emotion” from their childhood experiences and from continuous cultural references, as we next discuss.

Perceptual Representation of Rationality and Emotion

Throughout childhood (at least in Western cultures), we learn that the head and brain are the “location” of rational- ity, whereas the heart is the “abode” of emotion and feel- ings (Berendt and Tanita 2011; Fetterman and Robinson 2013; Swan 2009). The association of the heart as a symbol of emotion and the head as a symbol of rationality is con- tinuously reinforced via common language and through media. Hundreds of songs, books, and movies use these metaphors. For example, in the children’s movie Finding Nemo, Nemo has to solve a problem, and he puts his “fins” to both of his temples to represent the act of thinking; in Dr. Seuss’ How the Grinch Stole Christmas, when the Grinch is depicted as mean spirited (at the beginning of the movie), his heart is two sizes too small, but when he be- comes more caring (at the end of the movie), his heart grows to three times its earlier size. Similarly, in many children’s movies, a character points to his head when thinking or to his heart when expressing a strong emotion (see online appendix for further examples).

Everyday physical phenomena tend to corroborate these symbolic associations. For example, we perceive that our heart rate increases when we are experiencing strong emo- tions (Appelhans and Luecken 2006). In contrast, we com- monly attribute mental fatigue and headaches to having “thought too much” about something (Helvig and Minick 2013). Only later in our development do we learn that ra- tionality and emotion both originate in the brain. However, at that point, through repeated experiences and continuous cultural references, evaluations of rationality and emotion may have already become grounded in perceptions of physical verticality, whereby the head is above the heart, and thus rationality is above emotion.

There is also much historical context for such a belief. Notions that thinking is situated on a physically higher level than feeling can indeed be traced throughout the de- velopment of Western culture.

Antiquity (7th century BC to 4th century AD). In written text, Plato (5th century BC) is considered the first person to define the head as the portion of the body in which rational- ity is located while claiming that all emotions are physically located below the neck, and specifically inside our heart (passions) and gut (lower instincts; Finger 2001). In Roman

634 JOURNAL OF CONSUMER RESEARCH

times, Virgil (1st century BC) in his masterpiece The Aeneid guides readers through an analysis of human rationality and emotion that corresponds to Plato’s structure (Alvis 1995).

Medieval Period (5th century AD to 15th century AD). Between Antiquity and medieval times, Neoplatonic philosophy was based on the axiom that the human soul con- sists of a lower irrational soul and a higher rational soul (O’Meara 1978). Aquinas (13th century AD), one of the great- est medieval thinkers, supported the interpretation of reason as something higher than emotions. For Aquinas, emotions could be identified with certain bodily changes, particularly with alterations in the movements of the heart. Aquinas be- lieved that rationality, the element that distinguishes humans from animals, has a corporeal nature, being situated in the middle part of the head (Aquinas, trans. 1981).

Modern Period (16th century AD to 20th century AD). In the early modern period (16th century), the Galenist medi- cal tradition still located rationality in the brain, emotion (particularly anger) in the heart, and desire in the liver (Ballester 2002). Shakespeare’s work (16th century) is re- plete with the heart-versus-head metaphor, attributing ra- tional wisdom to the head and passion to the heart (Swan 2009). Descartes (17th century) continued with this dualis- tic way of thinking, distinguishing between res cogitans (mental substance) and res extensa (corporeal substance). Throughout the modern era, the Cartesian separation be- tween res cogitans and res extensa strengthened the con- trast between “reason” and the “heart” (symbol of sentiments, impulses, emotions, and passion that reason could not understand; see discussions by Schmitter 2014 and Damasio 1994). A similar contrast can be found in Freud (19th/20th century), who thought of the human psy- che as an iceberg. In his model, the ego (the more rational part of the psyche) is represented by the highest part of the iceberg, whereas the lower part of the iceberg represents the id (the irrational and emotional part of the psyche; Elliott 2002; Josephs 1997). Finally, in 20th-century soci- ology, we can find the opposition between “highbrow” and “lowbrow” culture, in which the first term describes intel- lectual culture and the latter labels popular culture, repre- sented by things like gossip magazines and emotional novels (Levine 1988).

This historical synopsis helps illustrate why the head and heart are still used as symbols of intellect and emotion, even though we know scientifically now that both rational- ity and emotion originate in the brain. Thinkers such as Plato, Aquinas, Descartes, Shakespeare, and Freud, and doctrines such as Neoplatonism have had an immense im- pact on contemporary Western culture, also affecting the development of today’s vocabulary (Dodds 1951; O’Meara 1978; Schmitter 2010; Whitehead 1978). In this regard, al- though a linguistic counterpart is not vital to show that a conceptual metaphor exists (as previously discussed), some expressions in everyday discourse consistently link

rationality with up and emotion with down. For instance, the expressions “I raised it back up to the rational plane,” “put your thinking cap on,” and “we had a high-level intel- lectual discussion” seem to associate rationality with up. Similarly, “falling in love,” “from the bottom of my heart,” and “feel something deep down” link emotion with down.

Building on the conceptual metaphor framework and the scaffolding theories, and observing the recurrent cultural presence of beliefs in which thinking is on a higher plane than feelings, we formally hypothesize that:

H1: Rationality is associated with a higher placement than

emotion along the vertical dimension.

Metaphorical Match and Evaluative Judgment

Prior research has documented the influence of process- ing fluency on consumers’ evaluations and preferences. Processing fluency is the ease or difficulty with which one processes external information, and it is a largely auto- matic effect (Clore 1992; Reber, Schwarz, and Winkielman 2004; Schwarz 2004).

Higher (vs. lower) fluency yields a more positive meta- cognitive experience and sensation (due to the ease of pro- cessing), and this positive sensation is often misattributed to the stimulus itself (Schwarz 2004). Consistent with this assumption, prior work has found that fluent processing in- creases liking (Reber et al. 2004) and consumer preference (Lee and Labroo 2004). Lee and Labroo (2004), for in- stance, show that participants develop more favorable atti- tudes toward ketchup when they have been previously presented either with an ad for ketchup or an ad for mayon- naise. This finding presumably reflects that exposure to the preceding stimulus facilitates the subsequent processing of the target stimulus, resulting in a more fluent experience. Note that because fluency is automatic, effortless, and does not require conscious or strategic processing of the stimuli (Bornstein and D’Agostino 1992), it is not measured in these studies but is inferred.

In our context, a match between physical verticality of a stimulus presentation and its rational/emotional nature should result in more fluent processing and, accordingly, more positive evaluations. Thus verticality should not only affect the perception of information in terms of its rationality and emotion, but it should also have down- stream consequences on people’s evaluations of rational/ emotional stimuli. In the marketing literature, we can find further support of how a match (mismatch) between phys- ical position and product information increases (de- creases) product evaluation. Deng and Kahn (2009) found that product images placed at the bottom (top) of a pack- age appear to be heavier (lighter). Consequently, the match between position (bottom/top) and products for which heaviness is considered a positive (negative) attrib- ute influenced people’s preference toward the product.

CIAN ET AL. 635

Chae and Hoegg (2013) showed that consumers from cul- tures that read from left to right tend to visualize the past on the left and the future on the right. Then ads containing a match (mismatch) between time and product spatial rep- resentation produce more (less) favorable attitudes toward the product.

Building on the preceding literature, we hypothesize the following:

H2: A match (vs. mismatch) of physical verticality and the

rational/emotional nature of stimuli will generate more (vs.

less) positive evaluations.

Potential Moderators

Although we have found theoretical support for the asso- ciation between rationality/ emotion and verticality, we be- lieve that some factors should mitigate this conceptual link. Next, we discuss the moderating role of awareness and familiarity.

Association Automaticity and Awareness. Much litera- ture shows how conceptual metaphors derive from an auto- matic, subconscious, and effortless process (Crawford 2009; Evans and Pourcel 2009; Landau, Meier, and Keefer 2010). To reach such a conclusion, most of the studies use implicit association tests or reaction-time tests (Meier et al. 2007; Meier, Sellbom, and Wygant 2007; Schubert 2005), which are psychological measures aimed at detecting peo- ple’s automatic association between mental representations of concepts (Robinson 2007). However, a test for automa- ticity does not provide information relative to what could happen if one were aware of the effect (Bargh 1984). Awareness of the metaphorical association could under- mine its automatic nature and hence attenuate its effect, in a similar manner to how awareness has been shown to de- crease mood effects (Schwarz 1990, 2012). Schwarz and Clore (1983) demonstrated that subjects reported more overall happiness and satisfaction with their lives when in a good mood (sunny days) than when in a bad mood (rainy days). However, the negative impact of bad moods was eliminated when subjects became aware of the connection between weather and mood. This parallels the observation that accessible knowledge can influence judgment without the person becoming aware of the influence, as observed in numerous priming studies (Higgins 1996). Once the person does become aware of a possible influence, for example, because their attention is drawn to the priming episode (Strack et al. 1993), the otherwise observed effect is attenu- ated and sometimes reversed (for a review, see Schwarz 2015). Based on the preceding literature, we hypothesize the following:

H3: The metaphorical association between verticality and

rationality/emotion will be attenuated when people become

aware of the metaphorical association.

Familiarity. Although a match (vs. mismatch) between verticality and rational/emotional connotations can lead to more positive attitudes toward the stimulus, it does not al- ways do so. In some situations, such as when the stimulus is familiar and well known, verticality may not have as large an impact on stimulus interpretation and consequent attitudes toward the stimulus. Brand familiarity indicates the extent of a consumer’s experience with a brand, either direct or indirect (Alba and Hutchinson 1987; Kent and Allen 1994), and it captures the consumers’ brand knowl- edge. Although consumers are familiar with many adver- tised products, they are unfamiliar with many others, either because those brands are new to the marketplace or be- cause consumers have not yet had experience with them (Stewart 1992). When consumers view an ad for an unfa- miliar brand, they are more likely to form their opinions about the brand based on the ad’s cues (Campbell and Keller 2003; Hilton and Darley 1991). More specifically, in our case, people should rely on metaphors in order to un- derstand seemingly unfamiliar information (Landau, Meier, and Keefer 2010). Consumers viewing an ad for a familiar brand, however, are more likely to use their exist- ing brand knowledge (Campbell and Keller 2003) in evalu- ating the ad. Accordingly, metaphor effects should be more pronounced when the metaphor guides the interpretation of novel information about an unfamiliar brand than when consumers can draw on existing knowledge about a famil- iar brand. Prior knowledge and evaluation may therefore attenuate the effect of a metaphorical match on product evaluation. Thus we hypothesize that:

H4: Brand familiarity (i.e., greater familiarity with the

brand) will attenuate the effect of the metaphorical match

(mismatch) on preferences toward the stimuli.

To test for our four hypotheses, we implement multiple methodologies and technologies: reaction times measured with an Implicit Association Test (IAT, study 1), interpre- tation of ambiguous stimuli using custom-programmed software (study 2), placement of elements captured through click-data maps (studies 3 and 4), and behavioral intentions utilizing self-reported measures (studies 5 and 6). Next, we discuss our studies.

PILOT STUDY: ADULTS’ PERCEPTION OF CHILDHOOD BELIEFS CONCERNING

RATIONALITY AND EMOTION LOCATION

Overview and Method

As discussed, we assume that throughout childhood peo- ple learn that the head is a symbol of rationality, whereas the heart is the “abode” of emotion. The pilot study tests this assumption using a between-subjects design with two

636 JOURNAL OF CONSUMER RESEARCH

conditions (emotion vs. rationality). Sixty-two people from an online pool completed the study in exchange for mone- tary compensation. Participants were shown a 500�500- pixel human silhouette and were asked to recall where, when they were children, they believed emotion (rational- ity) was “located” in the human body. They used the com- puter mouse to click on that location in the human silhouette.

The placement of the click, as measured in pixel coordi- nates, represented our dependent variable of interest. Consistent with our theoretical framework, we predicted that respondents in the rationality condition would click on the head section of the silhouette, whereas participants in the emotion condition would click below, around the heart section.

Results and Discussion

An analysis of variance (ANOVA) with recall (emo- tion vs. rationality) as the independent variable and pixel coordinates along the vertical axis as the dependent variable revealed that participants placed rationality significantly higher than emotion along the vertical axis (Mrationality¼426.69 pixels, Memotion¼348.75 pixels; F(1, 60)¼186.77, p < .01). As the heat maps show (Figure 1), participants located rationality in the head area of the silhouette, whereas they located emotion in the heart area.

In this pilot test, we show how in childhood, we be- lieve rationality is “located” in the head and that the heart is the “abode” of emotions. Even if we learn later on that rationality and emotion both originate in the brain, our

evaluations of rationality and emotion by that time may have already become grounded in perceptions of physical verticality (hypothesis 1). We next discuss our main studies.

STUDY 1: IMPLICIT ASSOCIATION TEST: UP AND RATIONAL, DOWN AND

EMOTIONAL

Overview

The IAT (Greenwald, McGhee, and Schwartz 1998) is one of the most commonly used procedures for testing the automaticity of a metaphorical connection (Kawakami at al. 2007; Meier and Robinson 2004; Steidle, Hanke, and Werth 2013). In study 1, we used the IAT to determine whether people implicitly associate up with rational and down with emotional (hypothesis 1). The IAT is a com- puter-based procedure built on the assumption that tasks requiring responses to unassociated categories take more time to do than identical tasks requiring responses to asso- ciated categories. In the IAT, participants see words or stimuli flashed one at a time on a computer screen and are asked to categorize them as belonging to one of two cate- gories that are presented on the same screen.

The IAT. The standard version of the IAT presents sub- jects with two binary categorization tasks and uses these to infer congruent or incongruent associations (Gawronski 2009; Lane et al. 2007). For example, in research on im- plicit racial prejudices (“race IAT”), the procedure requires people to categorize faces of European and African origin

FIGURE 1

HEAT MAPS FROM THE PILOT STUDY

CIAN ET AL. 637

in terms of their ethnicity (white/black—first binary cate- gorization task) and to categorize positive and negative words, in terms of their valence (good/bad—second binary categorization task). The IAT presents some “blocks” of stimuli with a single binary categorization task (in which people are asked to group black/white faces or good/bad words) and other blocks of stimuli with both binary catego- rization tasks (e.g., white/black and good/bad). The blocks of interest are the latter and are divided into (hypothesized) congruent and incongruent blocks. Specifically, in the prejudice-congruent block participants are asked to group black faces and negative words with one response key, and white faces and positive words with another response key. The opposite combination (black: positive, white: nega- tive) with the same procedure is presented in the prejudice- incongruent block. The assumption underlying the IAT is that the categorization task is easier (and thus quicker) when the two categories that share a response key are con- gruent rather than when they are incongruent (Nosek, Greenwald, and Banaji 2007). Thus the main dependent variable in this test is the response time (measured in milli- seconds) for congruent versus incongruent blocks.

Method

A total of 104 undergraduates from a large university in the midwestern United States completed the study in exchange for course credit. We used four words for each of four categories: rational (intelligent, logic, reason, thinking), emotional (feeling, mood, sentiment, sympathy), up (above, over, top, upper), and down (below, under, bottom, lower).

Controls. Word length and frequency: To rule out the possibility that our results were influenced by the words’ length or frequency, we compared the words in the rational category with the words in the emotional category, and the words in the up category with those in the down category. Word frequencies were calculated using Kucera-Francis norms (Francis and Kucera 1967). We did not find any dif- ferences regarding the number of letters or word frequency (p’s > .1).

Valence of Words. To ensure that our results are not af- fected by a vertical valence, we also controlled for the va- lence of the words. For this, we used the NRC Word- Emotion Association Lexicon database (Mohammad and Turney 2010). A Fisher exact test revealed no significant difference in the valence of words in the rational category versus those in the emotional category (p’s > .5).

Procedure. We followed the IAT standard procedure (Greenwald et al. 1998 for a detailed description).

Category names (“rational”–“emotional” and/or “up”– “down”) were placed on the upper left and upper right of the screen. In each block, words included in the two cate- gories appeared in random order in the center of a 14-inch computer screen (in black 18-point Courier New font on a

silver background). Figure 2 shows a screenshot taken dur- ing the study.

Per the standard, our IAT procedure consisted of five practice blocks and two main blocks (Messner and Vosgerau 2010). Participants had two response key op- tions: “q” if the word appearing in the center of the screen was related to the category on the left, and “p” if the word was related to the category on the right. In the first practice block, the rational category was on the left and the emo- tional category was on the right. Rational and emotional words appeared in random order on the middle of the screen. Participants pressed the left key (q) when they saw a rational word and the right key (p) when they saw an emotional word. Participants were also asked to answer as quickly and accurately as possible. Incorrect categoriza- tions were signaled using the word INCORRECT. A 150- millisecond blank screen followed correct categorizations. Each word remained on the screen until participants made a selection. The second practice block presented the up and down categories. Participants were instructed to press the left response key if the word was related to something that is up, and the right response key if the word was related to something that is down. The third practice block presented the two word lists together. When a rational or up word ap- peared, participants pressed the left response key; when an emotional or down word appeared, they pressed the right response key.

The task in the fourth block—the first main IAT block— was the same as in the third practice block. The response latencies in this fourth block were recorded and labeled as the congruent block. The fifth block—another practice block—switched the combination between response keys/ categories, such that participants pressed the left response key for emotional words and the right key for rational words. The sixth practice block grouped both word lists to- gether. If participants saw an up or emotional word, they pressed the left key, and if they saw a down or rational word, they pressed the right key. The seventh block—the second main IAT block (incongruent block)—was the same as that in the sixth practice block. As in the congruent block, we recorded the response latencies and used them for the analysis.

We counterbalanced the order of the two main blocks and their corresponding practice blocks between subjects (Brunel, Tietje, and Greenwald 2004), such that we assigned half the respondents to the described sequence (1-2-3-4-5-6- 7) and the other half to a 1-5-6-7-2-3-4 sequence. Furthermore, the position of the categories along the horizon- tal dimension (left/right) was randomized between subjects.

Results and Discussion

We followed the data reduction procedure of Greenwald et al. (1998). The response times, measured in millisec- onds, were log-transformed for the congruent and the

638 JOURNAL OF CONSUMER RESEARCH

incongruent blocks; response times for inaccurate, ex- tremely fast (less than 300 milliseconds) and extremely slow (more than 300 milliseconds) responses were deleted (these limits are the standards generally implemented when analyzing data from an IAT; Greenwald et al. 1998; Meier et al. 2007). Analyses were performed on these transformed response times, but the means are reported in raw millisec- onds to facilitate understanding of the response times.

We conducted a mixed-model ANOVA, with the con- gruent versus incongruent block as a within-subjects factor (“IAT block”) and the counterbalanced order of the blocks as a between-subject factor (“IAT order”). Transformed re- sponse times were the dependent variable.

As hypothesized, the main effect of IAT block was sig- nificant (F(1, 102)¼8.48, p < .01). Subjects were quicker at categorizing words in the congruent condition, where ra- tional-related words were paired with up-related words and emotional-related words were paired with down-related words (M¼827.90 milliseconds, standard deviation [SD]¼252.29 milliseconds); than in the incongruent

condition, where these pairings were reversed (M¼891.95 milliseconds, SD¼272.57 milliseconds). The main effect of the IAT order and its interaction with IAT block were not significant (p’s > .17).

Study 1 thus supports our primary hypothesis (hypothe- sis 1), indicating an implicit association of rational with up and emotional with down.

STUDY 2: CHINESE CHARACTERS’ INTERPRETATION AS A FUNCTION OF

VERTICALITY

Overview

Building on the association documented in study 1, study 2 tests whether this association affects how people categorize an ambiguous stimulus. We manipulated the vertical position of five Chinese characters on a computer screen (see figure 3), asking respondents if each character was more likely to be a Chinese brand logo for a statistical

FIGURE 2

A SCREENSHOT OF STUDY 1 SHOWING AN EXAMPLE OF CATEGORY NAMES AND A TARGET WORD

CIAN ET AL. 639

website (a target with rational connotations) or a dating website (a target with emotional connotations). A pretest with 40 undergraduates from a large university in the mid- western United States showed that a statistical website was considered more rational (M¼7.38) than a dating website (M¼3.05, F(1, 38)¼136.83, p < .01), whereas a dating website was considered more emotional (M¼7.53) than a statistical website (M¼1.71; F(1, 38)¼272.30, p < .01; 9 point scales anchored at Not at all rational/emotional and Extremely rational/emotional).

Per hypothesis 1, when a Chinese character is shown in a higher (versus lower) part of the screen, people should be more likely to interpret it as a logo for a statistical (versus dating) website; conversely, when it is on a lower (versus higher) part of the screen, it should more likely be inter- preted as a logo for a dating (versus statistical) website.

Method

Thirty undergraduates from a large university in the midwestern United States completed the study in exchange for course credit. Only participants who did not have any knowledge about the Chinese language were included in the study.

As a cover story, participants were told that this experi- ment investigates judgments of novel images. They were informed that they would see an image of a Mandarin Chinese character in different locations on the screen; their task was to decide whether the Chinese character fits better as a logo for a statistical website or a dating website. We told them that the location of the image would vary from character to character to increase participants’ attention and interest.

The images were presented on a 14-inch computer screen with a white background. Each subject was asked to evaluate five Chinese characters, one by one, in a random order (“order”). Before each character appeared on the screen, subjects saw a “fixation clue (þ)” in the middle of the screen for 1000 milliseconds. This was done to center fixation, and thereby ensured that every character had the same centered attentional focus. We programmed the task such that each participant randomly saw each image in one of two positions on the screen. These positions were equi- distant from the middle in verticality but either at the far top or far bottom of the computer screen—we call these

the higher and lower position (“position”). Each image was shown for 1000 milliseconds.

After seeing each image, subjects indicated as quickly as possible (by pressing the p or the q key on the keyboard) whether the image was more likely to be a brand logo for a Chinese statistical website or a Chinese dating website. Whether p or q was used for indicating the “statistical” or the “dating” websites was counterbalanced (“key”).

“Order” and “key” were not significant in any analysis (p’s > .4) and are not discussed further.

Results and Discussion

A contingency table with the count of responses (30 sub- jects�5 Chinese characters) is shown in table 1.

As predicted, a high spatial placement increased the likelihood that an image was categorized as the brand logo of a Chinese statistics website, whereas a low spatial place- ment increased the likelihood that it was categorized as the brand logo of a Chinese dating website. To analyze the data, we used a generalized estimating equation model—a generalized form of logistic regression for choices ob- served under a within-subjects design (Ge, Häubl, and Elrod 2012; Liang and Zeger 1986)—with “position” (higher/lower) and the specific “image” as independent variables and choice (statistical/dating website) as the de- pendent variable. Vertical position was indeed a significant predictor of subjects’ choices (B¼�870, Wald 5.49, 1 df, p < .05), whereas the image used was not significant (Wald 3.99, 4 df, p > .4).

These results support hypothesis 1 that vertical position affects perceivers’ interpretation of ambiguous stimuli in ways that reflect the metaphorical connection of up with rationality and down with emotion identified in study 1.

Ruling Out Valence as an Explanation. Note that these results are difficult to account for on the basis of a vertical- ity-valence association identified in other work (Meier and Robinson 2004). For undergraduate marketing students, dating websites have a more positive valence than statistics websites (a truism we confirmed in a pretest;

FIGURE 3

CHINESE CHARACTERS USED IN STUDY 2

TABLE 1

RESPONDENTS’ CATEGORIZATION (STATISTICAL VS. DATING) OF THE FIVE CHINESE CHARACTERS SHOWN IN

STUDY 2

Position

TotalLower Higher

Response Chinese statistical website

30 44 74 39% 60%

Chinese dating website

47 29 76 61% 40%

Total 77 73 150

640 JOURNAL OF CONSUMER RESEARCH

Mdating¼6.21, Mstatistics¼3.23; F(1, 51)¼44.23, p < .01; 9 point scales anchored at Bad/dislike/unpleasant and Good/like/pleasant; a¼ .94). Hence the verticality-valence association would predict that a given ideograph is more likely to be identified as a dating website (positive valence) when shown in a high position, and as a statistics website (negative valence) when shown in a low position. This pat- tern is the opposite of what we observed.

STUDY 3: WHERE DOES THE MUSIC SECTION VERSUS THE SCIENCE

SECTION GO ON A WEBSITE?

Overview

Study 2 showed that the physical placement of a stimu- lus along the vertical dimension can influence its interpre- tation along the metaphorically related rational-emotional dimension. This influence follows the direction from con- crete sensory perception (physical space) to abstract con- cepts, which is often assumed to be the primary direction of influence (Landau et al. 2010). In study 3, we tested whether the metaphorical association of verticality and ra- tionality/emotionality can also exert an influence in the op- posite direction (i.e., if the rational vs. emotional connotations of a stimulus affect its spatial placement). To test this, we asked people to freely place two sections of a website onto an empty page.

Pretests

Pretest 1. We asked 62 people from an online subject pool to rank various sections of an online newspaper on ra- tionality and emotionality. We chose the business, culture, news, science, technology, art, entertainment, food, music, and style sections for this purpose, assuming that the first five would be more representative of rationality and the lat- ter five with emotionality. Music was ranked as the most emotional (M¼7.56) and science as the most rational (M¼1.97) on 9 point scales anchored at 1 (Extremely ra- tional) and 9 (Extremely emotional).

Pretest 2. We assessed whether the music section and the science section differed on dimensions other than ratio- nality/emotionality. Specifically, we wanted to ensure that the two sections did not vary on valence, importance, and prototypicality. Fifty-one participants from an online pool participated in this second pretest. Participants were told that they would be rating a website newspaper section on several dimensions. Each participant was randomly as- signed to view one of two sections, presented in the cen- ter of a computer screen. To measure valence, participants were asked to rate their attitudes toward the section (1¼Bad/dislike/unpleasant; 9¼Good/like/pleas- ant; a¼ .97; Mitchell and Olson 1981). Then we asked subjects how important they thought the section was

(1¼Not important at all; 9¼Extremely important). Finally, to test prototypicality and other possible conscious associations with verticality, we asked respondents if they would expect the section to be in a high or low position (1¼Low; 5¼Neither high nor low; 9¼High); if they would associate the section with something that goes up or down (1¼Goes down; 5¼Goes neither up nor down; 9¼Goes up); and if they would associate the section with something that is up or down in position (1¼Up; 5¼Neither up nor down; 9¼Down, a¼ .72). There was no significant difference between the two sections (all p’s > .4) in all the dimensions considered.

Main Study

Study 3 had two between-subjects conditions (section: science vs. music). Forty-seven people from an online pool participated in this study. First, participants were told that we wanted to create a new website and needed to decide where to place different sections. We instructed partici- pants that they were going to see two pages. On the first page, they were going to see the name of the specific sec- tion they had to place. On the second page, they were go- ing to see a blank square. Participants were invited to think of the blank square as an empty website and use the com- puter mouse to click a position where they would place the section. After these instructions, each participant was ran- domly assigned to view one of two web sections: science or music. The name of the section was presented in the middle of a white page. Finally, we showed the participants a blank square of 500�500 pixels.

We programmed the survey such that when participants clicked with their mouse, the website section (music or sci- ence) that they had previously viewed appeared on the screen. They were allowed to click as many times as they wished until they were satisfied with their placement of the section. The placement of the section, as measured in pixel coordinates, represented our dependent variable of interest.

To ensure that participants were able to see the blank square, we showed them a 500�500-pixel blue square at the very beginning of the questionnaire and asked them if they were able to see the square without scrolling. Participants who had to scroll were automatically directed to the end of the questionnaire and then paid. Finally, we programmed the survey to automatically detect the participant screen resolu- tion to make sure it was not a confounding factor.

Consistent with hypothesis 1, we predicted that subjects would place the science section higher along the vertical axis than the music section.

Results and Discussion

An ANOVA with section (science or music) as the inde- pendent variable and pixel coordinates along the vertical axis as the dependent variable revealed that the science

CIAN ET AL. 641

section was placed significantly higher along the vertical axis than the music section (Mscience¼401.68 pixels, Mmusic¼332.14 pixels; F(1, 45)¼5.32, p < .05). Section (science or music) was not a significant predictor of pixel coordinates on the horizontal-axis (Mscience¼204.40 pix- els, Mmusic¼222.05 pixels; F (1, 45)¼ .21, p > .6).

Study 3 reveals that the relationship between verticality and rationality/emotionality is bidirectional: much as spa- tial placement can affect the interpretation of stimuli (study 2), the connotations of stimuli can affect their spatial place- ment (study 3). Furthermore, this study reveals that people not only hold the metaphorical association in their mind, but they tend to recreate it when they are free to manipulate rational and emotional stimuli along the vertical dimension.

Note that participants placed the science section higher than the music section even though pretest 2 for this study found no conscious expectations for this placement. This is consistent with the observation that people are often not aware of the conceptual metaphors that structure their thoughts (Lakoff and Johnson 1999). In fact, being aware of the metaphorical association at the time of judgment may undermine its influence (hypothesis 3). We test this prediction directly in study 4.

STUDY 4: AWARENESS OF THE METAPHORICAL ASSOCIATION

Overview and Method

Study 4 had a 2 (vertical awareness: no, yes)�2 (web- site section: rational, emotional) between-subjects design and used 138 people from an online pool. Study 4 repli- cated study 3 but added “awareness” conditions. The de- sign for the conditions in which we did not elicit the association awareness was the same as in study 3. In the conditions where subjects were made aware of the meta- phorical association, there was a small variation in the in- structions. Adapting the awareness procedure from Schwarz and Clore (1983), at the end of the instructions, we added the following sentence: “We are interested in studying if a website section that is more related to rational or emotional contents affects the vertical position in which you will place it.”

Results and Discussion

We conducted a 2�2 ANOVA with awareness and website section as the independent variables and pixel co- ordinates along the vertical axis as the dependent variable. Neither the main effect of awareness nor the main effect of the website section was significant (p > .1). However, the interaction between the two factors was significant (F (1, 134)¼4.13, p < .05). Planned contrasts revealed that within the unaware condition, participants placed the

science section above the music section (Mscience¼390.36 pixels, Mmusic¼335.74; F (1, 134)¼6.89, p < .05), repli- cating the results of study 3. However, within the aware condition, no significant difference emerged between the two sections (Mscience¼356.40 pixels, Mmusic¼362.02; p > .7). Finally, we found no difference between all four conditions regarding the section position along the horizon- tal axis (all p’s > .7)

We also captured the time participants spent until their last click on the blank square. Here the aware and unaware conditions differed, with participants in the aware condi- tion spending more time before making their final click compared to participants in the unaware condition (Munaware¼7.21 seconds, Maware¼10.53 seconds; F(1, 136)¼5.64, p < .05). This difference in the time spent pro- vides some indirect support for automaticity because a lon- ger time (vs. shorter time) spent in a choice setting is usually an indicator of a more controlled (vs. automatic) process (Dhar and Nowlis 1999; Suri and Monroe 2003). Making subjects aware of the metaphorical association seems to motivate them to think more about the vertical position in which the section is located. However, in the unaware conditions, participants seem to have used auto- matic and subconscious processing, which is both less ef- fortful and quicker than systematic processing (Chaiken 1980; Eagly and Chaiken 1993).

Study 4 corroborates the results of study 3, showing how the rational/emotional connotations of stimuli can affect their spatial placement. However, supporting hypothesis 3, we found that conceptual metaphors are less influential when attention is drawn to them. Note that the “awareness” manipulation of this study only made participants aware that the researchers were studying whether emotional versus ra- tional content affected vertical placement, without providing information about what the expected effect might be. This renders the test in study 4 fairly conservative.

STUDY 5: EFFECT OF VERTICALITY ON EVALUATION OF A POLITICAL

SLOGAN

Overview

We have proposed that a match between physical verti- cality (higher/lower position) of a stimulus and its charac- teristics (rational/emotional associations) should lead to more positive evaluations of the stimulus (hypothesis 2). Study 5 tests this prediction. Participants were exposed to a rational or emotional political slogan presented high or low on a computer screen. Of central interest is whether a slo- gan is evaluated more favorably when its vertical place- ment is consistent with its rational versus emotional connotation.

642 JOURNAL OF CONSUMER RESEARCH

Pretest

We created two political slogans for a nonpartisan elec- tion that differed in only one word. One had more of a ra- tional appeal and the other more of an emotional appeal.

Rational appeal: “Intelligence in serving the community”

Emotional appeal: “Passion in serving the community”

We pretested these descriptions with an online panel (n¼47) to ensure that they conveyed the desired appeal. Each participant was shown one description (presented in the center of a computer screen) and then asked to rate the rational/emotional appeal and political claim valence. We measured rational/emotional appeal with two items, namely, “Does this slogan have a rational or an emotional appeal?” (1¼Rational; 5¼Neither rational nor emotional; 9¼Emotional), and “Does this slogan evoke more feelings or thoughts?” (1¼Thoughts; 5¼Neither thoughts nor feel- ings; 9¼Feelings; a¼ .95). These items were adapted from Shiv and Fedorikhin (1999). To measure valence, par- ticipants were asked to rate their attitudes toward the claim (1¼Bad/dislike/unpleasant; 9¼Good/like/pleasant; a¼ .96; Mitchell and Olson 1981).

As expected, the rational slogan was rated as having a more rational appeal (M¼3.37), whereas the emotional slogan was rated as having a more emotional appeal (M¼7.21; F (1, 45)¼57.66, p < .01), both significantly different from the neutral point of 5 (p’s < .01). In terms of valence, both the slogans were rated close to the neutral point of the scale, with the emotional slogan having a (marginally significant) higher positive valence than the rational logo (Mrational¼5.52, Memotional¼6.57; F(1, 45)¼3.33, p¼ .07).

Main Study

Study 5 had a 2 (appeal: emotional, rational)�2 (screen position: higher, lower) between-subjects design. A total of 122 undergraduates from a large university in the midwest- ern United States completed the study in exchange for course credit. Each participant was told that he or she would be evaluating a political slogan for a nonpartisan po- litical election (that is, we specified that the candidate had no affiliation with a political party). They were then ran- domly assigned to one of the four conditions (i.e., rational slogan in the higher part of the screen, rational slogan in the lower part of the screen, emotional slogan in the higher part of the screen, and emotional slogan in the lower part of the screen). As in study 2, the higher and lower positions of the screen were equidistant from the middle. Participants were then asked for their attitudes toward the slogan (1¼Bad/dislike/unpleasant; 9¼Good/like/pleas- ant; a¼ .96; Mitchell and Olson 1981), the slogan

relevance (“This slogan was meaningful to me,” “This slo- gan is relevant to me,” and “This slogan is important to me” anchored on 1¼Not at all to 9¼Very much; a¼ .96; Williams and Drolet 2005), and their intention to vote for the candidate (on a scale from 1 to 100).

Results and Discussion

Table 2 shows the raw means of each dependent variable.

Attitude toward the slogan, slogan relevance, and inten- tion to vote were all highly correlated (a¼ .74 on the z-scores). Accordingly, we created an overall “slogan pref- erence index” (which ranged from�1.80 to 1.92). We then conducted a 2�2 ANOVA with position and appeal as the independent variables, and slogan preference index as the dependent variable.

The main effect of rational/emotional appeal was in the same direction of the pretest (Mrational¼�.13, Memotional¼ .13), although it was not statistically signifi- cant (p > .1). The main effect of position was not significant (p > .8). However, as expected, the interac- tion between appeal and position was significant (F (1, 118)¼9.39, p < .01). Planned contrasts reveal that the rational slogan was evaluated more favorably when it was presented higher versus lower on the screen (Mhigher¼ .10, Mlower¼�.33; F (1, 118)¼4.13, p < .05). Conversely, the emotional slogan was evaluated more fa- vorably when presented at a low versus high position on the screen (Mhigher¼� .13, Mlower¼ .36; F (1, 118)¼5.30, p < .05; results of the individual measures are included in the online appendix).

Study 5 supports our hypothesis 2 that a match between physical verticality (higher/lower position) of a stimulus and its metaphorically associated characteristics (rational/ emotional connotation) results in more positive stimulus evaluations. Thus the association between verticality and rationality/emotion has downstream consequences, influencing people’s preferences.

Note also that the obtained pattern is incompatible with an alternative pathway through the metaphorical associa- tion of verticality and valence (Meier and Robinson 2004). The verticality-valence metaphor predicts a main effect of

TABLE 2

STUDY 5 RAW MEANS

Slogan Position

Attitude toward

the slogan Slogan

relevance Intention to vote

Slogan preference

index

Rational Higher 5.10 4.76 50.14 .10 Lower 4.39 3.77 43.55 �.33

Emotional Higher 5.11 3.85 46.28 �.13 Lower 6.13 4.80 53.09 .36

CIAN ET AL. 643

position (up is good), rather than the observed interaction between position and type of appeal.

STUDY 6: HEALTHY, TASTY FOOD AND THE MODERATING EFFECT OF

FAMILIARITY

Overview

We have proposed that familiarity with the brand will moderate the effect of the match of physical verticality with the rational/emotional nature of stimuli on preferences for the stimuli. A familiar brand triggers brand associations that exist within a consumer’s knowledge set. Thus with greater familiarity, prior knowledge and evaluation may at- tenuate the effect of the metaphorical match on product evaluation (hypothesis 4).

In study 6, we provide product ads with brand names that are either familiar or not familiar, and we explore the impact of metaphorical match/mismatch on evaluations of the product. We ran two pretests to assess the rational/emo- tional connotation of the headlines and to choose the brands. We first present the pretests and explain the stim- uli, and then continue with the main experiment.

Pretests and Stimuli

We created two headlines for an ad depicting a granola bar. The headlines differed only in one word and had a dif- ferent (rational/ emotional) appeal.

Rational appeal: “The healthy choice” Emotional appeal: “The tasty choice”

Pretest 1. Sixty participants from an online pool evalu- ated the headlines. Using the same rationality/emotionality scale as used in study 5 (a¼ .96), participants evaluated the tasty headline as having a more emotional appeal (M¼2.53), whereas they evaluated the healthy headline as having a more rational appeal (M¼6.95; both significantly different from the midpoint 5, p’s < .01). In terms of va- lence (a¼ .91), we found no significant difference between the two headlines (Mhealthy¼6.24, Mtasty¼6.38; p > .7).

Pretest 2. We asked 20 people from an online pool to list, in a free response format, three brands of granola bars familiar to them. Kashi, Nature Valley, and Quaker were most frequently listed. We chose Kashi as the familiar brand. Building on these pretests, we created eight versions of an ad for a granola bar, with the product centered in the middle (Figure 4). In four of these ads, we used Kashi as the familiar brand; in the other four, we created a fictitious brand called “Dori”—this was used as the unfamiliar brand. Both for Kashi and Dori, two of the ads had the ra- tional headline (either above or below the product), and the other two ads had the emotional headline (either above or

below the product). As in the previous studies, the higher and lower positions were equidistant from the middle.

Main Study

Design. Study 6 had a 2 (brand: familiar, unfamil- iar)�2 (appeal: emotional, rational)�2 (position: higher, lower) between-subjects design. Each subject was ran- domly assigned to one of the eight conditions. Our depen- dent variables were the same as in study 4 (i.e., attitudes, relevance, and behavioral intention). Since we used food as stimuli, we also controlled for respondents’ gender and level of hunger (Brunstrom et al. 2008; Lozano, Crites, and Aikman 1999; Wardle et al. 2004).

Procedure and Measures. We ran study 6 using an on- line pool. Because brand familiarity was a critical factor in our study, we asked in the survey description that only peo- ple who were familiar with the Kashi brand (and unfamiliar with the Dori brand) take the survey. After reading the de- scription, 360 participants started the questionnaire in ex- change for money. To further screen for familiarity, we first asked participants to evaluate their familiarity with the “Dori”/“Kashi” brand name on a 7 point scale (7¼Very familiar, 6¼Quite familiar, 5¼Slightly familiar, 4¼Neither unfamiliar nor familiar, 3¼Slightly unfamil- iar, 2¼Quite unfamiliar, 1¼Very unfamiliar; Berlyne and Parham 1968). Participants who were “quite familiar” or “very familiar” with Kashi and also “quite unfamiliar” or “very unfamiliar” with Dori were allowed to continue with the survey, whereas the others (14 participants [4% of the total]) were automatically directed to the end of the ques- tionnaire and paid without being asked any further ques- tions. Thus 346 respondents passed our screening questions and completed the main survey.

After this screening, participants completed an unrelated five minute filler task and then reported their hunger (on a scale from 1¼Very full to 9¼Very hungry). Finally, they were directed to the main study. Here, participants were told that they would be evaluating an ad for a granola bar and were randomly assigned to one of the eight conditions. Similar to study 4, after viewing the ad, we asked partici- pants for their attitudes toward the granola bar, the head- line relevance, and their behavioral intentions. The attitude and relevance scales were adapted from study 5 (attitude scale a¼ .97; relevance scale a¼ .94). With respect to the behavioral intentions, similarly to study 5, we asked them to evaluate their intention to eat and to buy the granola bar (both on a scale from 1 to 100; a¼ .89).

Results and Discussion. Table 3 shows the raw means of each dependent variable. Attitude toward the product, slogan relevance, intention to eat, and intention to buy were all highly correlated (a¼ .90 on the z-scores). Consequently, we created an overall “preference index” (ranging from �1.99 to 1.62). We then conducted a

644 JOURNAL OF CONSUMER RESEARCH

FIGURE 4

ADS USED IN STUDY 6

CIAN ET AL. 645

2�2�2 analysis of covariance with brand familiarity, ra- tional/emotional appeal, and position as the independent variables; preference index was the dependent variable, and hunger and gender were covariates.

The main effect of brand familiarity was significant (F (1, 336)¼16.62, p < .01), with the familiar brand pre- ferred to the unfamiliar one. The main effect of rational/ emotional appeal and position were not significant (p’s > .8). Hunger had a significant effect on the preference index (F (1, 336)¼4.91, p < .05), supporting previous lit- erature showing that hunger enhances attitudes toward the food and purchase intentions (Lozano et al. 1999). Gender was marginally significant (p¼ .08), with females having higher preference for the granola ad than males, both in the familiar and unfamiliar condition.

None of the two-way interactions were significant (p > .1), except for the interaction between rational/emo- tional appeal and position (F (1, 336)¼10.53, p < .01). However, this is qualified by a significant three-way inter- action between familiarity, rational/emotional appeal, and position (F (1, 336)¼4.24, p < .05). Planned contrasts re- veal that—within the unfamiliar brand—participants evalu- ated the ad with the rational headline more favorably when the rational headline was presented above rather than be- low the granola bar (F(1, 336)¼6.87, p < .01). Conversely, participants evaluated the ad with the emo- tional headline more favorably when the emotional slo- gan was presented below rather than over the granola bar (F (1, 336)¼7.29, p < .01). Within the familiar brand, however, neither planned contrast was significant (p > .4). Results of the individual measures are included in the on- line appendix.

Study 6 further corroborates hypothesis 2 and provides support for hypothesis 4. Results shows that a match be- tween physical verticality (higher/lower position) of a stimulus and its metaphorically associated characteristics (rational/emotional connotation) results in more positive evaluations (hypothesis 2), unless the brand is familiar (hy- pothesis 4). A further contribution of study 6 is that it uses

stimuli that resemble real ads, which gives our research more external validity and generalizability to the kinds of ads used in the market.

GENERAL DISCUSSION

This series of six main studies (and a pilot) shows a ro- bust association between the abstract concepts of rational- ity and emotions and the physical dimension of verticality. First, we demonstrate an implicit association between ver- ticality and rationality/emotion using an IAT procedure (study 1). Second, we find this metaphorical association to be bidirectional, going from concrete to abstract (study 2) as well as from abstract to concrete (study 3). Specifically, ambiguous stimuli are perceived as having a more rational meaning when presented high rather than low on a screen (study 2). Conversely, stimuli with rational connotations are placed higher on the screen than stimuli with emotional connotations (study 3). We show how this association has downstream effects, affecting people’s preferences: a given stimulus is evaluated more favorably when its spatial placement matches its rational/emotional connotations (studies 5 and 6). Moreover, we explore two boundary con- ditions. In study 4, we show how the metaphorical effect fades when people become aware of the association. In study 6, we also show that familiarity with the brand atten- uates the effect of the metaphorical match (mismatch) on perceivers’ preferences toward the stimuli.

Theoretically, our work contributes to the conceptual metaphor and embodied cognition literature (Barsalou 2008, Krishna and Schwarz 2014; Lakoff and Johnson 1980, 1999; Landau et al. 2010; Lee and Schwarz 2014a) by showing a metaphorical association between verticality and rationality/emotionality—that rationality is associated with up or higher and emotion with down or lower. This re- search also contributes to the literature exploring the im- pact of visual cues on consumer behavior (Greenleaf and Raghubir 2008; Hoegg and Alba 2011; Hoegg, Alba, and

TABLE 3

STUDY 6 RAW MEANS

Unfamiliar brand Headline Position

Attitude toward the product

Headline relevance

Intention to eat

Intention to buy

Overall preference index

Dori Rational Higher 6.46 4.16 61.40 52.71 .00 Lower 5.34 3.38 44.23 37.25 �.51

Emotional Higher 6.03 3.27 50.00 40.61 �.36 Lower 7.13 4.04 63.39 57.09 .11

Familiar brand Headline Position

Attitude toward the product

Headline relevance

Intention to eat

Intention to buy

Overall preference index

Kashi Rational Higher 7.45 4.66 65.21 63.17 .30 Lower 7.02 4.43 64.45 60.95 .19

Emotional Higher 6.81 3.95 66.51 55.56 .08 Lower 6.98 4.45 68.72 59.47 .21

646 JOURNAL OF CONSUMER RESEARCH

Dahl 2010; Patrick and Peracchio 2010; Vadiveloo, Morwitz, and Chandon 2013) by demonstrating how visual positioning of information along the vertical dimension can affect consumer attitudes, preferences, and interpreta- tion of ambiguous information.

Theoretical Implications

Verticality has been associated with several conceptual metaphors, such as valence (good is up; Meier and Robinson 2004), power (powerful is up; Schubert 2005), and morality (moral is up; Meier et al. 2007). Our studies add rationality (rational is up) to this list. Methodologically, this requires that competing metaphori- cal meanings of identical sensory inputs are controlled for, as we have done in the present studies. More important, it raises the conceptual issue of what determines which of the multiple metaphorical meanings of a given sensory attrib- ute a perceiver adopts. We assume that the answer parallels findings in other domains: perceivers adopt the interpreta- tion that is rendered most accessible by the task and the context in which it is presented. This issue has been exten- sively explored in the domain of fluency research, where the same metacognitive experience (i.e., the subjective ex- perience of ease or difficulty with which people process in- formation) can inform a wide variety of judgments, taking on different meanings in different contexts (Alter and Oppenheimer 2009; Schwarz 2004, 2010, 2015). For exam- ple, fluently processed stimuli are judged as more positive (Winkielman and Cacioppo 2001), less risky and more fa- miliar (Song and Schwarz, 2009), more beautiful (Reber et al. 2004), more likely to be true (Reber and Schwarz 1999), and so on (Schwarz 2010). This malleability reflects that people’s inferences from experienced fluency are guided by lay theories of mental processes that are them- selves recruited by the context and the task they face (Schwarz 2004, 2010). Moreover, once an accessible inter- pretation has been adopted, competing interpretations are less accessible and people arrive at judgments that are at odds with inferences they would have readily drawn from the same input in a slightly different context (Schwarz 2010). Similarly, the rational versus emotional connota- tions of the stimuli (e.g., science vs. music websites in study 3 or rational vs. emotional political slogans in study 5) or tasks (e.g., determining the brand logo of a statistics vs. dating website in study 2) elicited an interpretation of verticality information along the rational-emotional dimen- sion, which was brought to mind by the tasks themselves. Once this interpretation was adopted, perceivers no longer seemed to use other metaphorical associations that reflect a link of verticality and valence (Meier and Robinson 2004) or verticality and power (Schubert 2005). Exploring the context sensitivity of metaphor selection and application is a promising (and pressing) issue for future research (Krishna and Schwarz 2014).

A second issue of broad theoretical importance is the di- rectionality of metaphor effects. Studies 2 and 3 showed that the association between physical verticality and ratio- nality/emotionality is bidirectional. In contrast, Meier and Robinson (2004) found that the association between physi- cal verticality and affective valence is unidirectional. For example, perceivers’ affective states influence visual per- ceptions but not vice versa. A consensus on the directional- ity of conceptual metaphor effects has not been reached, although plentiful research exists on the subject (cf. Landau et al. 2010; Lee and Schwarz 2012; Zhang and Wang 2009; Zhong and Leonardelli 2008). Zhang and Li (2012) proposed that conceptual metaphors exert a unidi- rectional effect when the target domain is abstract and has no direct bodily connotations; conversely, they expect bidi- rectional effects when the target domain has bodily conno- tations. The fit of this account depends on what one considers “bodily.” On the one hand, the observed bidirec- tional association of verticality with rational/emotional fits the account, given that the association is rich in direct and clear body states (e.g., heart rate increase and headaches, as previously discussed). On the other hand, Meier and Robinson’s (2004) unidirectional association of verticality and affective valence is only compatible with the account if we assume that the valence stimuli they used have fewer bodily associations than the material used in the present studies. Suffice it to say that the determinants of metaphor- ical directionality remain a challenging issue that requires more extensive experimentation.

Finally, it is also worth noting that awareness of a poten- tial influence of the metaphorical link attenuated its impact in study 4. This is consistent with extensive research show- ing that awareness of potential contextual influences— from feelings (Schwarz and Clore 1983) to accessible con- cepts (Strack et al. 1993)—undermines their impact (for a review, see Schwarz 2012). This also suggests that meta- phors exert more influence when their introduction is sub- tle rather than blatant, which deserves more attention in future research.

Managerial Implications

Several directly implementable managerial implications are also worth nothing. First, the present studies used brands, website sections, food ads, and political slogans as stimuli. However, the lessons learned should apply to many other elements of marketing communications. Indeed, all visual formats—whether a printed page or a television, computer, or smartphone screen—integrate ver- tical placement. Our studies suggest that when deciding on the vertical positioning of products or information on a printed page or a computer screen, marketing managers should take into consideration the rational-emotional asso- ciation of their message. The information will be much more powerful when the vertical position matches rather

CIAN ET AL. 647

than mismatches the rational versus emotional content of the message. One must acknowledge, however, that verti- cality may at times have less of an effect on the interpreta- tion of the stimulus and thus on the consequent attitudes, such as when an individual is familiar with a brand and has preconceived opinions about it, as in study 6. In study 5, a political slogan with a rational (emotional) appeal was more influential when the appeal was presented in the higher (lower) part of the ad, resulting in differential voting intentions. As a real-world illustration, we end by noting how Shepard Fairey’s (Figure 5) iconic HOPE poster for the 2008 Obama campaign observed the metaphor match- ing principle, placing HOPE with its strong emotional as- sociations in the lower part of the image. This metaphorical match may not have influenced people with strong prior opinions about Obama. However, for the many undecided and less opinionated voters, the emotional slo- gan on the lower part of the poster may have led to a more positive attitude toward Obama than if the same slogan had been in the upper part.

DATA COLLECTION INFORMATION

The first author supervised the collection of data for all the studies by research assistants. The data were collected

from summer 2012 to spring 2015 either at the Behavioral Lab of the Ross School of Business (University of Michigan) or using the Amazon Mechanical Turk panel (details are described in the methods section of the stud- ies). The first author analyzed the data collected in all stud- ies. All the authors continually discussed data analyses and results. Data results were also discussed in multiple occa- sions in Professor Krishna’s Sensory Marketing Lab and in Professor Schwarz’s Social Cognition Lab.

REFERENCES

Alba, Joseph W. and J. Wesley Hutchinson (1987), “Dimensions of Consumer Expertise,” Journal of Consumer Research, 13 (March), 411–54.

Alter, Adam L. and Daniel M. Oppenheimer (2009), “Uniting the Tribes of Fluency to Form a Metacognitive Nation,” Personality and Social Psychology Review, 13, 219–35.

Alvis, John (1995), Divine Purpose and Heroic Response in Homer and Virgil: The Political Plan of Zeus, Lanham, MD: Rowman & Littlefield.

Appelhans, Bradley M. and Linda J. Luecken (2006), “Heart Rate Variability as an Index of Regulated Emotional Responding,” Review of General Psychology, 10 (3), 229–40.

Aquinas, Thomas (trans. 1981), Summa Theologica, trans. Fathers of the English Dominican Province, Westminster, MD: Christian Classics.

Ballester, Luis Garcia (2002), Galen and Galenism Theory and Medical Practice from Antiquity to the European Renaissance, Surrey, UK: Ashgate.

Bargh, John A. (1984), “Automatic and conscious processing of social information,” in Handbook Of Social Cognition, ed. Thomas K. Srull and Robert S. Wyer, Hillsdale, NJ: L. Erlbaum Associates, 1–43.

———. (2006), “What Have We Been Priming All These Years? On the Development, Mechanisms, and Ecology of Nonconscious Social Behavior,” European Journal of Social Psychology, 36 (2), 147–68.

Barsalou, Lawrence W. (2008), “Grounded Cognition,” Annual Review of Psychology, 59, 617–45.

Berendt, Erich A. and Keiko Tanita (2011), “The ‘Heart’ of Things: A Conceptual Metaphoric Analysis of Heart and Related Body Parts in Thai, Japanese and English,” Intercultural Communication Studies, 20 (1), 65–78.

Berlyne, Daniel E. and L. C. C. Parham (1968), “Determinants of Subjective Novelty,” Perception & Psychophysics, 3 (6), 415–23.

Bornstein, Robert F. and Paul R. D’Agostino (1992), “Stimulus Recognition and the Mere Exposure Effect,” Journal of Personality and Social Psychology, 63 (October), 545–52.

Brunel, Frederic F., Brian C. Tietje, and Anthony G. Greenwald (2004), “Is the Implicit Association Test a Valid and Valuable Measure of Implicit Consumer Social Cognition?” Journal of Consumer Psychology, 14 (4), 385–404.

Brunstrom, Jeffrey M., Peter J. Rogers, Emmanuel M. Pothos, Raff Calitri, and Katy Tapper (2008), “Estimating Everyday Portion Size Using a ‘Method of Constant Stimuli’: In a Student Sample, Portion Size Is Predicted by Gender, Dietary Behaviour, and Hunger, but not BMI,” Appetite, 51 (2), 296–301.

FIGURE 5

OBAMA HOPE POSTER

Source: This image was originally posted to Flickr by “Count to 10” at https://

www.flickr.com/photos/31057088@N02/8083830049. It is licensed under the

terms of the Creative Commons v.2.0.

648 JOURNAL OF CONSUMER RESEARCH

Campbell, Margaret C. and Kevin Lane Keller (2003), “Brand Familiarity and Advertising Repetition Effects,” Journal of Consumer Research, 30 (2), 292–304.

Casasanto, Daniel (2009), “New Directions in Cognitive Linguistics,” Vol. 24, ed. Vyvyan Evans and Stephanie Pourcel, Amsterdam, the Netherlands: John Benjamins.

Chae, Boyoun Grace and JoAndrea Hoegg (2013), “The Future Looks ‘Right’: Effects of the Horizontal Location of Advertising Images on Product Attitude,” Journal of Consumer Research, 40 (2), 223–38.

Chaiken, Shelly (1980), “Heuristic versus Systematic Information Processing and the Use of Source versus Message Cues in Persuasion,” Journal of Personality and Social Psychology, 39 (5), 752–66.

Clark, Herbert H. (1973), “Space, Time, Semantics, and the Child,” in Cognitive Development and the Acquisition of Language, ed. Timothy E. Moore, New York: Academic Press, 27–63.

Clore, Gerald L. (1992), “Cognitive Phenomenology: Feelings and the Construction of Judgment,” in The Construction of Social Judgment, ed. Leonard L. Martin and Abraham Tesser, Hillsdale, NJ: Erlbaum, 133–64.

Cosmides, Leda and John Tooby (2000), “Evolutionary Psychology and the Emotions,” in Handbook of Emotions, 2nd ed., ed. Michael Lewis and Jeannette M. Haviland-Jones, New York: Guilford, 91–115.

Crawford, L. Elizabeth (2009), “Conceptual Metaphors of Affect,” Emotion Review, 1 (2), 129–39.

Damasio, Antonio (1994), Descartes’ Error: Emotion, Reason, and the Human Brain, New York: Grosset/Putnam.

Deng, Xiaoyan and Barbara E. Kahn (2009), “Is Your Product on the Right Side? The ‘Location Effect’ on Perceived Product Heaviness and Package Evaluation,” Journal of Marketing Research, 46 (6), 725–38.

de Sousa, Ronald (1990), The Rationality of Emotion, Cambridge: MIT Press.

———. (2014), “Emotion,” Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta, http://plato.stanford.edu/archives/ spr2014/entries/emotion/.

Dhar, Ravi and Stephen M. Nowlis (1999), “The Effect of Time Pressure on Consumer Choice Deferral,” Journal of Consumer Research, 25 (4), 369–84.

Dodds, Eric R. (1951), The Greeks and the Irrational, Berkeley: University of California Press.

Eagly, Alice H. and Shelly Chaiken (1993), The Psychology of Attitudes, Fort Worth, TX: Harcourt Brace Jovanovich.

Elliott, Anthony (2002), Psychoanalytic Theory: An Introduction, Durham, NC: Duke University Press.

Evans, Dylan and Pierre Cruse, eds. (2004), Emotion, Evolution, and Rationality, Oxford, UK: Oxford University Press.

Evans, Vyvyan and Stephanie Pourcel, eds. (2009), New Directions in Cognitive Linguistics, Amsterdam, the Netherlands: John Benjamins.

Fetterman, Adam K. and Michael D. Robinson (2013), “Do You Use Your Head or Follow Your Heart? Self-Location Predicts Personality, Emotion, Decision Making, and Performance,” Journal of Personality and Social Psychology, 105 (2), 316–34.

Finger, Stanley (2001), Origins of Neuroscience: A History of Explorations into Brain Function, Oxford, UK: Oxford University Press.

Francis, W. Nelson and Henry Kucera (1967), Computational Analysis of Present-Day American English, Sudbury, MA: Dartmouth Publishing.

Gawronski, Bertram (2009), “Ten Frequently Asked Questions About Implicit Measures and Their Frequently Supposed, But Not Entirely Correct Answers,” Canadian Psychology/ Psychologie canadienne, 50 (3), 141–50.

Ge, Xin, Gerald Häubl, and Terry Elrod (2012), “What to Say When: Influencing Consumer Choice by Delaying the Presentation of Favorable Information,” Journal of Consumer Research, 38 (6), 1004–21.

Giessner, Steffen R. and Thomas W. Schubert (2007), “High in the Hierarchy: How Vertical Location and Judgments of Leaders’ Power Are Interrelated,” Organizational Behavior and Human Decision Processes, 104 (1), 30–44.

Greenleaf, Eric and Priya Raghubir (2008), “Geometry in the Marketplace,” in Visual Marketing: From Attention to Action, ed. Michel Wedel and Rik Pieters, New York: Taylor & Francis, 113–42.

Greenspan, Patricia S. (1988), Emotions and Reasons: An Inquiry into Emotional Justification, New York: Routledge.

Greenwald, Anthony G., Debbie E. McGhee, and Jordan L. K. Schwartz (1998), “Measuring Individual Differences in Implicit Cognition: The Implicit Association Test,” Journal of Personality and Social Psychology, 74 (6), 1464–68.

Helvig, Ashley W. and Ptlene Minick (2013), “Adolescents and Headaches: Maintaining Control,” Pediatric Nursing, 39 (1), 19–25.

Higgins, E. Tory (1996), “Knowledge Activation: Accessibility, Applicability, and Salience,” in Social Psychology: Handbook of Basic Principles, ed. E. Tory Higgins and Arie Kruglanski, New York: Guilford, 133–68.

Hilton, James L. and John M. Darley (1991), “The Effects of Interaction Goals on Person Perception,” in Advances in Experimental Social Psychology, Vol. 24, ed. Mark P. Zanna, New York: Academic Press, 236–67.

Hoegg, JoAndrea and Joseph W. Alba (2011), “Seeing Is Believing (Too Much): The Influence of Product Form on Perceptions of Functional Performance,” Journal of Product Innovation Management, 28 (3), 346–59.

Hoegg, JoAndrea, Joseph W. Alba, and Darren W. Dahl (2010), “The Good, the Bad, and the Ugly: Influence of Aesthetics on Product Feature Judgments,” Journal of Consumer Psychology, 20 (4), 419–30.

Josephs, Lawrence (1997), “The View from the Tip of the Iceberg,” Journal of the American Psychoanalytic Association, 45 (2), 425–63.

Kawakami, Kerry, Curtis E. Phills, Jennifer R. Steele, and John F. Dovidio (2007), “(Close) Distance Makes the Heart Grow Fonder: Improving Implicit Racial Attitudes and Interracial Interactions Through Approach Behaviors,” Journal of Personality and Social Psychology, 92 (6), 957–71.

Kent, Robert J. and Chris T. Allen (1994), “Competitive Interference Effects in Consumer Memory for Advertising: The Role of Brand Familiarity,” Journal of Marketing, 58 (July), 97–105.

Kirman, Alan, Pierre Livet, and Miriam Teschl (2010), “Rationality and Emotions,” Philosophical Transactions of the Royal Society B: Biological Sciences, 365 (1538), 215–19.

Krishna, Aradhna and Norbert Schwarz (2014), “Sensory Marketing, Embodiment, and Grounded Cognition: A Review and Introduction,” Journal of Consumer Psychology, 24 (2), 159–68.

Lakoff, George (1993), “The Contemporary Theory of Metaphor,” Metaphor and Thought, 2, 202–51.

CIAN ET AL. 649

Lakoff, George and Mark Johnson (1980), Metaphors We Live By, Chicago: University of Chicago Press.

———. (1999), Philosophy in the Flesh: The Embodied Mind and Its Challenge to Western Thought, New York: Basic Books.

Landau, Mark J., Brian P. Meier, and Lucas A. Keefer (2010), “A Metaphor-Enriched Social Cognition,” Psychological Bulletin, 136 (6), 1045–67.

Lane, Kristin A., Mahzarin R. Banaji, Brian A. Nosek, and Anthony G. Greenwald (2007), “Understanding and Using the Implicit Association Test: IV. What We Know (So Far) About the Method,” in Implicit Measures of Attitudes: Progress and Controversies, ed. Bern Bernd Wittenbrink and Norbert Schwarz, New York: Guilford Press, 59–102.

Lee, Angela and Aparna Labroo (2004), “The Effect of Conceptual and Perceptual Fluency on Brand Evaluation,” Journal of Marketing Research, 41 (May), 151–65.

Lee, Spike W. S. and Norbert Schwarz (2011), “Clean Slate Effects: The Psychological Consequences of Physical Cleansing,” Current Directions in Psychological Science, 20, 307–11.

———. (2012), “Bidirectionality, Mediation, and Moderation of Metaphorical Effects: The Embodiment of Social Suspicion and Fishy Smells,” Journal of Personality and Social Psychology, 103 (5), 737–49.

———. (2014a), “Metaphors in Judgment and Decision Making,” in The Power of Metaphor: Examining Its Influence on Social Life, ed. Mark J. Landau, Michael D. Robinson, and Brian P. Meier, Washington, DC: American Psychological Association, 85–108.

———. (2014b), “Framing Love: When It Hurts to Think We Were Made for Each Other,” Journal of Experimental Social Psychology, 54, 61–67.

Levine, Lawrence W. (1988), Highbrow/Lowbrow, Cambridge: Harvard University Press.

Liang, Kung-Yee, and Scott L. Zeger (1986), “Longitudinal Data Analysis Using Generalized Linear Models,” Biometrika, 73 (1), 13–22.

Lozano, Dora I., Stephen L. Crites, and Shelley N. Aikman (1999), “Changes in Food Attitudes as a Function of Hunger,” Appetite, 32 (2), 207–18.

Mandler, Jean M. (1992), “How to Build a Baby: II. Conceptual Primitives,” Psychological Review, 99 (4), 587–604.

Meier, Brian P., David J. Hauser, Michael D. Robinson, Chris Kelland Friesen, and Katie Schjeldahl (2007), “What’s ‘Up’ with God? Vertical Space as a Representation of the Divine,” Journal of Personality and Social Psychology, 93 (5), 699–710.

Meier, Brian P. and Michael D. Robinson (2004), “Why the Sunny Side Is Up: Associations Between Affect and Vertical Position,” Psychological Science, 15 (4), 243–47.

Meier, Brian P., Martin Sellbom, and Dustin B. Wygant (2007), “Failing to Take the Moral High Ground: Psychopathy and the Vertical Representation of Morality,” Personality and Individual Differences, 43 (4), 757–67.

Messner, Claude and Joachim Vosgerau (2010), “Cognitive Inertia and the Implicit Association Test,” Journal of Marketing Research, 47 (2), 374–86.

Mitchell, Andrew A. and Jerry C. Olson (1981), “Are Product Attribute Beliefs the Only Mediators of Advertising Effects on Brand Attitudes?” Journal of Marketing Research, 18 (3), 318–22.

Mohammad, Saif and Peter Turney (2010), “Emotions Evoked by Common Words and Phrases: Using Mechanical Turk to

Create an Emotion Lexicon,” in Proceedings of the NAACL- HLT 2010 Workshop on Computational Approaches to Analysis and Generation of Emotion in Text, June, Los Angeles, CA.

Nosek, Brian A., Anthony G. Greenwald, and Mahzarin R. Banaji (2007), “The Implicit Association Test at Age 7: A Methodological and Conceptual Review,” in Automatic Processes in Social Thinking and Behavior, ed. John A. Bargh, London: Psychology Press, 265–92.

O’Meara, Dominic J., ed. (1978), Neoplatonism and Christian Thought, Vol. 3, Albany: SUNY Press.

Patrick, Vanessa M. and Laura A. Peracchio (2010), “‘Curating’ the JCP Special Issue on Aesthetics in Consumer Psychology: An Introduction to the Aesthetics Issue,” Journal of Consumer Psychology, 20 (4), 393–97.

Pham, Michel Tuan (2007), “Emotion and Rationality: A Critical Review and Interpretation of Empirical Evidence,” Review of General Psychology, 11 (2), 155–78.

Reber, Rolf and Norbert Schwarz (1999), “Effects of Perceptual Fluency on Judgments of Truth,” Consciousness and Cognition, 8 (3), 338–42.

Reber, Rolf, Norbert Schwarz, and Piotr Winkielman (2004), “Processing Fluency and Aesthetic Pleasure: Is Beauty in the Perceiver’s Processing Experience?” Personality and Social Psychology Review, 8 (4), 364–82.

Robinson, Michael D. (2007), “Lives Lived in Milliseconds: Using Cognitive Methods in Personality Research,” in Handbook of Research Methods in Personality Psychology, ed. Richard W. Robins, R. Chris Fraley, and Robert F. Krueger, New York: Guilford, 345–59.

Schmitter, Amy M. (2010), “17th and 18th Century Theories of Emotions,” Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta, http://plato.stanford.edu/entries/emotions- 17th18th/.

Schubert, Thomas W. (2005), “Your Highness: Vertical Positions as Perceptual Symbols of Power,” Journal of Personality and Social Psychology, 89 (1), 1–21.

Schwarz, Norbert (1990), “Feelings as Information: Informational and Motivational Functions of Affective States,” ed. E. T. Higgins and R. Sorrentino, Handbook of Motivation and Cognition: Foundations of Social Behavior, Vol. 2, New York: Guilford Press, 527–61.

———. (2004), “Meta-Cognitive Experiences in Consumer Judgment and Decision Making,” Journal of Consumer Psychology, 14 (4), 332–48.

———. (2010), “Meaning in Context: Metacognitive Experiences,” in The Mind in Context, ed. B. Mesquita, L. F. Barrett, and E. R. Smith, New York: Guilford, 105–25.

———. (2012), “Feelings-as-Information Theory,” in Handbook of Theories of Social Psychology, ed. Paul A. Van Lange, Arie W. Kruglanski, and E. Tory Higgins, Thousand Oaks, CA: Sage, 289–308.

———. (2015), “Metacognition,” in APA Handbook of Personality and Social Psychology: Attitudes and Social Cognition, ed. Mario Mikulincer, Phillip R. Shaver, Eugene Borgida, and John A. Bargh, Washington, DC: American Psychological Association, 203–29.

Schwarz, Norbert and Gerald Clore (1983), “Mood, Misattribution, and Judgments of Well-Being: Informative and Directive Functions of Affective States,” Journal of Personality and Social Psychology, 45, 513–23.

Shiv, Baba and Alexander Fedorikhin (1999), “Heart and Mind in Conflict: The Interplay of Affect and Cognition in Consumer

650 JOURNAL OF CONSUMER RESEARCH

Decision Making,” Journal of Consumer Research, 26 (3), 278–92.

Song, Hyunjin and Norbert Schwarz (2009), “If It’s Difficult to Pronounce, It Must Be Risky: Fluency, Familiarity, and Risk Perception,” Psychological Science, 20 (2), 135–38.

Steidle, Anna, Eva-Verena Hanke, and Lioba Werth (2013), “In the Dark We Cooperate: The Situated Nature of Procedural Embodiment,” Social Cognition, 31 (2), 275–300.

Stewart, David W. (1992), “Speculations on the Future of Advertising Research,” Journal of Advertising, 21 (September), 1–17.

Strack, Fritz, Norbert Schwarz, Herbert Bless, Almut Kübler, and Michaela M. Wänke (1993), “Awareness of the Influence as a Determinant of Assimilation versus Contrast,” European Journal of Social Psychology, 23, 53–62.

Suri, Rajneesh and Kent B. Monroe (2003), “The Effects of Time Constraints on Consumers’ Judgments of Prices and Products,” Journal of Consumer Research, 30 (1), 92–104.

Swan, Toril (2009), “Metaphors of Body and Mind in the History of English,” English Studies, 90, 460–75.

Vadiveloo, Maya, Vicki Morwitz, and Pierre Chandon (2013), “The Interplay of Health Claims and Taste Importance on Food Consumption and Self-Reported Satiety,” Appetite, 71, 349–56.

Wardle, Jane, Anne M. Haase, Andrew Steptoe, Maream Nillapun, Kiriboon Jonwutiwes, and France Bellisie (2004), “Gender Differences in Food Choice: The Contribution of Health Beliefs and Dieting,” Annals of Behavioral Medicine, 27 (2), 107–16.

Whitehead, Alfred N. (1978), Process and Reality, New York: The Free Press.

Williams, Lawrence E. and John A. Bargh (2008a), “Experiencing Physical Warmth Promotes Interpersonal Warmth,” Science, 322 (5901), 606–7.

———. (2008b), “Keeping One’s Distance: The Influence of Spatial Distance Cues on Affect and Evaluation,” Psychological Science, 19 (3), 302–8.

Williams, Lawrence E., Julie Y. Huang, and John A. Bargh (2009), “The Scaffolded Mind: Higher Mental Processes Are Grounded in Early Experience of the Physical World,” European Journal of Social Psychology, 39 (7), 1257–67.

Williams, Patti and Aimee Drolet (2005), “Age-Related Differences in Responses to Emotional Advertisements,” Journal of Consumer Research, 32 (3), 343–54.

Winkielman, Piotr and John T. Cacioppo (2001), “Mind at Ease Puts a Smile on the Face: Psychophysiological Evidence That Processing Facilitation Elicits Positive Affect,” Journal of Personality and Social Psychology, 81 (6), 989–1000.

Zajonc, Robert (2000), “Feeling and Thinking: Closing the Debate over the Independence of Affect,” in Feeling and Thinking: The Role of Affect in Social Cognition, ed. Joseph P. Forgas, New York: Cambridge University Press, 31–58.

Zhang, Meng and Xiuping Li (2012), “From Physical Weight to Psychological Significance: The Contribution of Semantic Activations,” Journal of Consumer Research, 38 (6), 1063–75.

Zhang, Meng and Jing Wang (2009), “Psychological Distance Asymmetry: The Spatial Dimension versus Other Dimensions,” Journal of Consumer Psychology, 19 (3), 497–507.

Zhong, Chen-Bo and Geoffrey J. Leonardelli (2008), “Cold and Lonely: Does Social Exclusion Literally Feel Cold?” Psychological Science, 19 (September), 838–42.

CIAN ET AL. 651

Copyright of Journal of Consumer Research is the property of Oxford University Press / USA and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.