Impression Management Using Typeface Design

Pamela W. Henderson Joan L. Giese Joseph A. Cote

Conditional Acceptance at the Journal of Marketing Ms 02-168 April 2004

Copyright 2002-2004 Pamela W. Henderson, Joan L. Giese, and Joseph A. Cote all rights reserved

Pamela W. Henderson is Associate Professor of Marketing, Washington State University, 2710 University Drive, Richland, WA 99352, 509-372-7207, 509-372-7512 (fax), phenders@tricity.wsu.edu. Joan L. Giese is Assistant Professor, Department of Marketing, Washington State University, Pullman, WA 99164-4730, 509-335-6354, 509-335-3865 (fax), giesej@wsu.edu. Joseph A. Cote is Professor, Department of Marketing, Washington State University, 14204 NE Salmon Creek Avenue, Vancouver, WA 98686 360-546-9753, cote@vancouver.wsu.edu. The authors gratefully acknowledge Donovan Follette, Theresa Grate, Jeff Boettcher, James Hutton, Andrew Eads, numerous professional graphic designers, the editor, and the reviewers for their contribution to this research.

Impression Management Using Typeface Design Abstract This paper develops empirically-based guidelines to help managers select typefaces that impact strategically valued impressions. The potential tradeoffs among the impressions created by typeface pleasing, engaging, reassuring and prominent are addressed. Selecting typeface can be simplified by use of six underlying design dimensions; elaborate, harmony, natural, flourish, weight, and compressed.

Impression Management Using Typeface Design The visual aspects of a corporations marketing materials are receiving increasing attention in marketing research (Childers and Jass 2002; Henderson and Cote 1998; Shapiro 1999; Tavassoli 2001, 2002; Tavassoli and Han 2001, 2002; Veryzer and Hutchinson 1998). By far, the most pervasive design element in marketing materials is typestyle. Both academicians and practitioners recognize that typeface design is an important visual tool for accomplishing corporate communication objectives (Childers and Jass 2002; Hutton 1987; McCarthy and Mothersbaugh 2002; Pan and Schmitt 1996; Tantillo, DiLorenzo-Aiss, and Mathisen 1995). Initial research indicates that typeface design impacts perceptions of advertised brands, influences the readability and memorability of ads (Childers and Jass 2002; McCarthy and Mothersbaugh 2002), creates strategically important impressions (e.g., general positive image to more specific impressions of innovativeness, change, power, or warmth, see Craig 1980; Dolen 1984; Hinrichs and Hirasuna 1990; Hutton 1987, 1997; O'Leary 1987; Solomon 1991; Somerick 2000; Spaeth 1995; Tantillo, DiLorenzo-Aiss, and Mathisen 1995), affects the appropriateness of the typeface for different products (Pan and Schmitt 1996; Walker, Smith, and Livingston 1986), and may affect a companys financial performance (Bloch 1995; Hertenstein and Platt 2001; Hutton 1997; Wallace 2001). Despite increasing research on the subject, little guidance is available to assist corporations in selecting typeface to create strategically important impressions. Research on the impressions created by typeface has assessed only a few of the many characteristics that differentiate their design. Characteristics studied include serif versus sans serif (Tannenbaum, Jacobson, and Norris 1964; Tantillo, DiLorenzo-Aiss, and Mathisen 1995), general use versus novel display, weight (Rowe 1982), and italics (Tannenbaum, Jacobson, and Norris 1964). In addition, most

studies have examined only a small set of typefaces (10 or fewer); thus limiting the variance seen in the extensive pool of fonts from which corporations make their selections. Furthermore, conflicting results continue to surface (e.g., Rowe 1982; Tantillo, DiLorenzo-Aiss, and Mathisen 1995). The only conclusion researchers generally agree about is that typeface design impacts responses, yet the nature of its effects are not well known. This lack of guidelines led McCarthy and Mothersbaugh (2002) to call for research to create a set of principles linking the features of type with the impressions they create. In this paper, we develop a set of empirically-based guidelines to improve managers ability to choose typefaces that impact strategically valued impressions. In particular, we address four questions: 1. 2. 3. 4. What are the strategically relevant impressions created by typeface design? What characteristics are most useful for describing typeface design? What is the impact of design on each type of impression? What guidelines should corporations follow to achieve their communication goals through the use of typeface design?

We conducted an exploratory study of current typeface designs and the impressions they create. Being exploratory in nature, the study is not based on a specific theoretical perspective. However, there are a few theories related to the possible dimensions of impressions and design characteristics that will help direct the analysis. Thus, we briefly review theories and empirical research related to the dimensionality of design and the impressions they create for other stimuli. Review of Typeface Design Effects There are two classes of studies that allow us to anticipate the influence of typeface design characteristics on consumer responses. First, several empirical studies have directly explored the relationship between a typefaces characteristics (e.g., serifs) and a response (e.g., innovativeness). For example, Tantillo, DiLorenzo-Aiss, and Mathisen (1995) investigated the

influence of typestyle on selected affective responses (e.g., happy/sad and young/old) and found numerous response differences between the sans serif and the serif typefaces while, in a similar study, Rowe (1982) found very few differences. Although each of these studies was a worthy initial effort, each study fails to investigate a sample of fonts that have a representative range of design characteristics. These studies also fail to consider the full array of responses to fonts. A second class of studies that can offer insight into the influence of typeface design characteristics on consumer responses is the research on aesthetics. Perception-based theories (such as Gestalt psychology) suggest that simple and harmonious designs will be liked more than complex and disharmonious designs (Bornstein and DAgostino 1992; Klinger and Greenwald 1994; Van den Bergh and Vrana 1998; Veryzer and Hutchinson 1998; Whitfield and Slater 1979; Whittlesea 1993). Motivation-based theories suggest more elaborate designs will increase arousal and lead to liking (-shape relationship, see Berlyne 1971; Hirschman 1980). Other theories, such as prototypicality and conditioning, attempt to explain why some stimuli are more pleasing than others, but develop no clear link between the design characteristics and responses (Martindale 1988; Martindale and Moore 1988; Veryzer 1999; Veryzer and Hutchinson 1998; Whitfield and Slater 1979, 1983). Purely empirical work on logo design indicates a linear positive relationship between a harmonious design and a pleasingness response and an -shape relationship between an elaborate design and a pleasingness response (Henderson and Cote 1998). Henderson and Cote (1998) also found a positive relationship between natural designs and liking, but they present no theory to explain this link nor do they address dimensions of response other than pleasingness and liking. In summary, theory and prior research provide insufficient guidance for managing the range of impressions influenced by typeface design.

Responses to Typeface Characteristics In determining impressions created by typeface design, we first consider what constitutes a response. From a purely empirical standpoint, the design literature has documented a wide variety of responses to art and design (e.g., honest, distinctive, happy, warm, graceful, beautiful, masculine, powerful, interesting, intense, emotional, etc., see Craig 1980; Ernst 1977; Rowe 1982; Solomon 1991; Tantillo, DiLorenzo-Aiss, and Mathisen 1995). More broadly, we consider what might drive response. First, design adds meaning to the stimulus (beyond simply the depiction of letters). In their seminal research on rating the meaning of concepts, Osgood, Suci, and Tannenbaum (1957) conclude that responses could be accounted for by three underlying dimensions: evaluation (e.g., good, pleasant, beautiful); activation (e.g., hot, active, fast); and potency (e.g., strong, brave, rugged). Design also conveys emotion. Mehrabian and Russell (1974) have shown that basic emotions or affective responses are captured by three dimensions: pleasingness (e.g., pleased, contented, relaxed); arousal (e.g., excited, aroused, jittery); and dominance (e.g., important, influential, controlling). Finally, in the context of a corporation, design communicates something about the company. As such, we turn to the spokesperson literature. Much as a spokesperson dresses-up the advertisers spoken words, so typeface design dresses-up the written word. Spokesperson research has involved conceptualizing and measuring responses to celebrity endorsers as three underlying dimensions: attractiveness (e.g., elegant, beautiful, attractive); trustworthiness (e.g., honest, sincere, dependable); and expertise (e.g., skilled, qualified, experienced, see Ohanian 1990). There are noticeable similarities to the responses studied by Osgood, Suci, and Tannenbaum (1957), Mehrabian and Russell (1974), and Ohanian (1990). Similarities are evident in the evaluation/pleasingness/attractiveness dimensions, the activation/arousal dimensions, and the

potency/dominance dimensions. In addition, the trustworthiness dimension is discussed in the typology literature and is critical to spokesperson response and to businesses in general. Expertise, on the other hand, seems irrelevant to typeface design. Design Characteristics of Typeface Typeface design can be distinguished by universal and typeface-specific characteristics. Universal design characteristics are subjective descriptions of the typeface and include characteristics such as symmetry, activity, and complexity. As such, universal design characteristics are holistic descriptions that rely on perception and could be used to describe a wide variety of stimuli. Typeface-specific design characteristics are graphemic descriptions of the fonts and include characteristics such as short/tall, serif/sans serif, and condensed/extended. Typeface-specific characteristics are not as subjective and provide an opportunity to explain additional variance in responses specific to typeface design. In effect, examining universal characteristics allows for greater generalizability of findings, whereas examining typefacespecific characteristics provides an opportunity to hypothesize about additional design factors and/or engineer a typeface to meet specific goals. Although there is no direct research on the universal characteristics of typeface design, research on logo design has proposed three universal dimensions; elaborateness (complex, active, and depth), naturalness (representative and organic), and harmony (balance & symmetry, see Henderson and Cote 1998; Henderson et al. 2003). These dimensions may or may not hold for typeface. There is no research that indicates the dimensionality of typeface-specific characteristics. It is also difficult to anticipate whether typeface-specific dimensions will, or will not, be independent of universal design dimensions. On the one hand, typeface-specific characteristics, such as serifs and ascenders, appear to be so specific to type since they do not

clearly relate to one of the universal dimensions. On the other hand, some typeface-specific characteristics, such as handwritten versus typed, appear to relate to universal dimensions found for logos, such as the naturalness dimension. Unfortunately, neither empirical nor theoretical information regarding typeface characteristics is sufficient to provide much guidance on dimensionality. Goal of Research Our review of the design literature suggests that there are no meaningful guidelines for typeface design. The lack of guidelines may lead to designs that do not achieve corporate objectives. For example, reports on corporations changing their logotypes typically discuss the image management hopes to communicate through the new typeface (Spaeth 1994; 1999; 1995). Yet, the implicit assumption of a single response to a logotype is probably incorrect. It is more likely that there are multiple responses to a logotype and that the corporation must consider tradeoffs among responses when developing its communication goals. Thus, guidelines are needed to assist corporations in managing the range of impressions created through their choices. In order to develop meaningful guidelines for selecting fonts, we conducted an empirical investigation to determine the design dimensions that best capture differences among typefaces, the response dimensions typefaces generate, and how typeface design dimensions relate to response dimensions. Based on the empirical findings, we then provide guidelines for corporations when selecting typeface. Research Method Data collection required four stages. First, appropriate typeface design characteristics were identified and a sample of representative typefaces selected. Next, typefaces were rated on the selected design characteristics by professional graphic designers and advertisers. A list of

strategically relevant impressions was then identified. Lastly, consumers responded to the typefaces on those impression measures. Phase 1: Selection of Design Characteristics and Typefaces As noted earlier, universal design dimensions, such as natural, harmony, and elaborate, should be relevant for all types of stimuli. These universal dimensions may be captured by either universal design variables (e.g., the symmetry variable loads onto the harmony dimension for brand logos) or typeface-specific variables (e.g., handwritten is more organic than machine-made type). In addition, there are also likely to be design dimensions that are unique to the specific stimulus. These typeface-specific dimensions can only be captured by typeface-specific variables. In Phase 1, a list of universal and typeface-specific design characteristics was developed. Design characteristics included in publications on typology were used to construct an initial list. Next, five professional graphic designers from different firms that work with a wide range of corporations were asked to list the primary characteristics that differentiate typeface design. The final list consisted of 16 universal and eight typeface-specific design characteristics. Table 1 lists the typeface design characteristics and provides illustrative fonts. -----------------------------Insert Table 1 about here -----------------------------The five professional designers each provided an extensive list (40-150) of commercially available typefaces representative of variation on the design characteristics they had identified. To further broaden the range of typefaces, additional typeface software was purchased. Based on this input, the researchers selected 210 typefaces. Typefaces were selected that represented the full range for each of the 24 design characteristics (e.g., extremely complex typefaces to

extremely simple typefaces) and had both an upper and lower case (some specialty typefaces only have upper case). See Table 1 for examples of the design characteristics. Phase 2: Ratings of Design Characteristics Eighty-two professional graphic designers, working in agencies and corporations, rated the 210 typefaces on seven-point semantic differential scales for each of the 24 characteristics. To minimize fatigue, each designer rated between 10 and 30 typefaces on 12 of the 24 characteristics. Typefaces were presented on white paper in 16-point font size in full alphabetic (upper and lower case) and numeric forms. A paper-and-pencil method was used to ensure that typefaces appeared true to form since computer and software differences across design agencies pose difficulties in maintaining consistent typeface appearances for research purposes. In all, 17,683 individual ratings were obtained from professional designers. Phase 3: Selection of Impression Responses Because our goal is to provide guidance to corporations, we researched impressions that corporations and designers seek to create through typeface. Managerially-oriented design literature and input from five professional graphic designers revealed the responses thought to result from typeface design and that are relevant to the general communication goals of corporations. These impressions include innovative, calm, liking, interesting, formal, strong, warm, honest, familiar, emotional, masculine/feminine, and attractive. To confirm the relevance of these impressions, 35 additional professional designers rated their perceived ability to select typefaces that elicited these responses. Results indicated that these designers believed that they could select typefaces to create these impressions and that these impressions were meaningful to them and their corporate clients.

10

We purposely did not refer to scales found in the meaning (Osgood, Suci, and Tannenbaum 1957), emotion (Mehrabian and Russell 1974), or spokesperson (Ohanian 1990) literatures, as we wanted the results of this study to address specific corporate goals. This approach provides a stronger test of these dimensions of response in the context of type than if we had used scales from these studies. Still, there is enough similarity in the evaluations that meaningful comparisons should be possible. Phase 4: Ratings of Impressions Created by Typefaces The next phase involved obtaining responses to the typefaces. Because of the enormity of gathering a large sample of evaluations of 210 typefaces on each of the 12 response variables, a laboratory-based computer task was used. The computer display of each typeface was visually examined and measured to guarantee that the appearance, size, and resolution was the same on the monitor as on the printed pages to which graphic designers responded. One typeface was dropped because its appearance on the computer was slightly different than its appearance in print. Software was written to randomly select 20 typefaces to present to the respondents. Each typeface was presented individually as a complete alphabet and number set accompanied by six of the 12 seven-point semantic differential scales (e.g., like/dislike; strong/delicate; etc.). The participants controlled the speed of typeface viewing and responding. Over 60,000 response ratings from 336 upper-division students at a large university were obtained with an average of 24.3 responses per typeface. The use of students in design research has been repeatedly justified by findings that show a surprising consistency in aesthetic response across age groups (Berlyne 1971; Eysenck 1988).

11

Analysis and Results The data were analyzed with the approach used in experimental aesthetics as well as in research on language processing [e.g., \Berlyne, 1974 #409;Bradshaw, 1984 #472]. Analyses were conducted at the stimulus- rather than the individual-level. This requires averaging across individual ratings of a stimulus (typeface) on a particular characteristic or impression response to obtain a score for that stimulus on that variable. All remaining analyses were conducted using these stimulus scores. Thus, the unit of analysis is the typeface, and the sample size for each analysis is the number of typefaces (209). This approach is particularly appropriate for marketing management because it recognizes that stimuli are designed for, managed for, and responded to by groups of people rather than individuals. The variables used in the analyses were 23 averaged design characteristics and the 12 averaged responses. We dropped one of the original 24 design characteristics (frequency of use) because we concluded that it did not reflect design, but instead reflected designers behavior. What are the Strategically Relevant Response Dimensions? Consistent with Henderson and Cote (1998), we started by conducting an exploratory factor analysis (EFA) of the impressions data. The analysis produced three factors. However, one factor included positive loadings for interesting, emotional, and innovative; and negative loadings for calm, formal, honest, and familiar. This factor was cumbersome to interpret since it combined the activation/arousal dimension (Mehrabian and Russell 1974; Osgood, Suci, and Tannenbaum 1957) with a variant of the trustworthy response dimension suggested by the spokesperson literature. An attempt to replicate the exploratory factor results using confirmatory factor analysis (CFA) indicated the problems with the EFA factor structure (CFI = .658, and low loadings for emotional and interesting). The CFA was respecified to include four factors, which

12

roughly corresponded to the evaluation/pleasingness/attractiveness, activation/arousal, potency/dominance, and trustworthiness dimensions found in previous research on emotions, evaluations of objects, and spokesperson perceptions (see Table 2). The model fit the data reasonably well (CFI = .806) and all the loadings were quite high ( > .7, with the exception of strong/delicate). The correlations among the impression factors were modest but significant (see Table 3). In addition, half the correlations were negative, indicating an implicit tradeoff between impression responses. Based on these results, four dimensions were used to describe the impression variables. Pleasing/Displeasing was comprised of liked, warm and attractive. Engaging/Boring was comprised of interesting and emotional. Reassuring/Unsettling consisted of calm, formal, honest, familiar and a negative loading for innovative (whose opposite endpoint was mainstream). Finally, Prominent/Subtle included strong (whose opposite endpoint was delicate) and masculine. Summated scores were used to capture each of the dimensions. -----------------------------Insert Table 2 about here -----------------------------What Dimensions Best Capture Typeface Design? As described earlier, the design characteristics consisted of universal and typeface-specific characteristics. Universal characteristics included distinctive, ornate, special use, conveys meaning, depth, uniform, balanced, smooth, symmetric, curved, organic, slanted, active, readable (which is known to capture simplicity in type) and handwritten/typed (which indicates natural versus man-made in type). Typeface specific characteristics included serif, ascenders, descenders, heavy, repeat, fat, condense, and x-height. These two groups of characteristics were

13

analyzed separately to ensure that generalizable dimensions were identified and linked to responses so as to better advance design research across stimuli. Exploratory factor analyses were performed using principal component analysis with varimax rotation. The factor analysis of the universal characteristics revealed three design dimensions, which explained 69.7 % of the variance (see Table 3). The first factor, Elaborateness, included ornate, depth, distinctive, meaningful, and negative loadings for readable and common (versus special purpose) use. The second factor, Harmony, included balance, smoothness, symmetry, and uniformity. The third factor, Naturalness, included active, curved, organic, slant, and a negative loading for looks-typed (as opposed to handwritten). These results are very similar to Henderson and Cotes (1998) findings, albeit using different design characteristics specified by practitioners. A CFA confirmed the appropriateness of this factor structure (CFI = .858). However, subsequent analyses used the EFA results since orthogonal factor scores could be created for examining the relationship between design and response. Factor analysis of the typeface-specific variables also uncovered three dimensions which explained 60.4% of the variance (see Table 3). Flourish is comprised of serif, ascenders, and descenders. It might appear that flourish and elaborate would tap the same dimension. However, they were only correlated at .183. Weight is comprised of heavy, fat, and repeated elements. Finally, Compressed is comprised of condensed and x-height. Again, a CFA confirmed the appropriateness of the factor structure (CFI = .862), and the EFA factor scores were used for the regression analyses. -----------------------------Insert Table 3 about here ------------------------------

14

How do Design Characteristics Influence Responses to Fonts? Four separate regression analyses were conducted using pleasing, engaging, reassuring, and prominent as the dependent variables. The three universal design dimensions (elaborateness, naturalness, and harmony) and the three font-specific design dimensions (flourish, weight, and compressed) were used as predictors. First, all the design dimensions were included in the model. We then added non-linear relationships using stepwise regression. We also tested for interactions amongst the design dimensions, but none were significant. Overall, the design dimensions were strongly related to the impressions created by typeface. Explained variance was particularly high (adjusted R2 from .514 to .734). In addition, both universal and typefacespecific design dimensions influenced response; however, the universal dimensions consistently explained more variance. A summary of the regression results is presented in Table 4. -----------------------------Insert Table 4 about here -----------------------------Pleasing/Displeasing. The design dimensions explained 51.4% of the variance in the pleasing/displeasing response factor. Elaborate, harmony, natural, flourish and compressed all had significant impacts. Natural had the largest impact (R2 = .320) creating more pleasing fonts, with the effect leveling off at high values. Elaborateness explained 7.9% of the variance and had a negative impact on pleasingness. Harmony, flourish, and compressed all had modest effects (R2 = .045, .039, and .031 respectively). Harmony and flourish both increased pleasingness (harmony had slight non-linearity in the nature of the positive relationship). A curvilinear relationship indicated that moderate values of compressed created the most pleasing fonts.

15

Engaging/Boring. Design explained a large portion of the variance (68.0%) in the engaging/boring dimension of response. Natural and elaborate had the largest effects (R2 = .268 and .211 respectively) with higher levels of natural and elaborate creating more engaging typefaces. Harmony was also important and created less engaging fonts (R2 = .170), with this effect diminishing at higher levels of harmony. Compressed and flourish had nominal effects (R2 = .018 and .013 respectively); and both increased the engagingness of the font. Reassuring/Unsettling. The design elements explained 73.4% of the variance in the reassuring/unsettling response dimension. Harmony and elaborate had the most influence explaining 38.5% and 33.0% of the variance respectively. Harmony made fonts more reassuring while elaborateness made them more unsettling. Flourish explained a nominal amount of variance (R2 = .018) and made fonts more reassuring. Prominent/Subtle. Natural, weight, flourish and harmony explained 52.8% of the variance in the prominent/subtle response factor. Natural explained the most variance (R2 = .291) with more natural fonts being perceived as less prominent and more subtle. Weight increased perceptions of prominence and explained 17.1% of the variance. Flourish and harmony created less prominent designs, explaining 4.3% and 2.4% of the variance respectively. Developing Guidelines In addition to conducting regression analyses to assist in developing guidelines for corporations use of typefaces, we conducted a cluster analysis to assess tradeoffs. The focus of the cluster analysis was on identifying response profiles that could be achieved through a range of commercially available fonts. The response clusters were built using the 12 raw response variables, rather than the factors, so as to not lose any richness in the data. The number of clusters was determined by looking at the average distance between clusters and comparing this

16

to the within-cluster distances. In addition, we avoided creating clusters with too few fonts. Six clusters appeared to best describe the data. While creating additional clusters reduced the distance values within clusters, cluster sizes became very small. The results and discussion of the cluster analysis, as well as an extended discussion of guidelines for impression management through typeface usage, are presented in the next section. Guidelines for Impression Management Before examining specific guidelines for selecting typeface, a number of general conclusions are supported by our analysis. The results provide broad empirical support for the contention by CEOs, corporate identity analysts, and creative agencies that typefaces convey a variety of strategically important impressions (Dolen 1984; Hutton 1997; Somerick 2000; Spaeth 1995). Just as a spokesperson projects an image of the company, typeface appears to have the potential to influence the impressions created by corporate communications. In addition, the strength of the relationship between typeface design and the resulting impressions (adjusted R2 from .514 to .734) suggests that corporations can have significant control over the resulting impressions (all content issues being equal). Since type is inherent to most corporate communications, companies can cost effectively leverage the benefits of an appropriately designed typeface. Our findings further reveal that corporations should take into consideration all four responses that their typefaces create. Thus, typeface should be carefully selected to ensure consistency with other elements of the corporate identity strategy. For example, Hiltons original logomark was redesigned using a script look to make it more friendly (Spaeth 1999). from to

Our results show that natural, scripted typefaces produce more reassuring and pleasing fonts. However, Hiltons new font is only average on elaborateness and harmony. The

17

combined result is a font that is less prominent, (more subtle), and only moderately engaging. Simply changing to a script look in an attempt to make a friendlier impression had a more complex effect in that a combination of responses (pleasing, less prominent, and only moderately engaging) resulted. Focusing only on a single response may lead to unintended consequences in the other types of response. Firms must recognize the implications of design for all responses, since multiple responses may be elicited. Not only must firms attend to the breadth of impressions resulting from their font selections, they will, in many cases, need to make some tradeoffs with respect to the responses sought. Ideally, corporations would be able to create any combination of the above impressions. However, since the design factors have different effects on impressions, a practitioners ability to create high values on all four response dimensions is limited. Specifically, elaborate designs increase how engaging the design is, but decrease how pleasing and reassuring it is. Harmony increases pleasing and reassuring responses, but decreases engaging and prominence responses. Lastly, natural designs are pleasing and engaging, but less prominent. As such, some tradeoff between responses appears to be necessary. To illustrate, the change in the Citibank logo might achieve their goal of being . . .softer, less aggressive, and cozier (Spaeth 1999)" but may also be too uninteresting (i.e., too low on the engaging dimension) and less pleasing. from to

A similar font used in our study, corporate mono, is reassuring, but below average in pleasing, low in engaging, and average in prominence. Common Response Combinations We draw on the cluster analysis results to illustrate the tradeoffs. Among the wide range of commercially available fonts in this study, we found six general profiles (see Table 5).

18

-----------------------------Insert Table 5 about here -----------------------------The first cluster comprises pleasing, subtle (not prominent), and engaging fonts that were average on the reassuring dimension. These are liked, warm, attractive, interesting, emotional, feminine, and delicate (Scheherezade). The means on design dimensions for this cluster confirm predictions from the regression results. Namely, fonts that evoke these responses will be high in harmony and flourish and low in weight. A good example is Anglewizard Films (see Table 5). The second group is comprised of unsettling but engaging fonts. These fonts are interesting, emotional, exciting, informal, dishonest, unfamiliar, and innovative (Paintbrush). While most companies would not want to be referred to as unsettling (not reassuring), these fonts communicate an edginess that is of value in many communication efforts. For example, Terrwear.com (see Table 5) produced clothing for mountain biking. Again the design dimension means for this cluster confirm the regression analysis predictions that fonts that evoke these responses are natural, somewhat elaborate and lack harmony. The next cluster of fonts is unlikely to be used heavily by corporations. These fonts are displeasing and unengaging (i.e., boring) but are average on the reassuring and prominence dimensions. They are disliked, cold, unattractive, uninteresting and unemotional (Chainlink). The cluster means confirmed regression analysis predictions that these fonts would be unnatural, low on compressed, and have little flourish. Even the tradeoffs predicted by the regression (elaborate designs are displeasing but engaging, and harmony creates boring but pleasing fonts) were consistent with the cluster results. Although their use in marketing will be limited, there may be some communication situations where a font such as this would be used; for example, to

19

display the characteristics or claims of a competing brand or in non-profit advertising to describe undesirable behaviors (e.g., antismoking advertisements). However, some companies may want to produce this type of displeasing image. For example, Cleopatra Records (see Table 5) produces albums such as Zeromancers Clone Your Lover, and This is NeoGoth. The fourth profile of fonts is prominent while being average on pleasing, engaging, and reassuring. These fonts are masculine and strong and are characterized by designs with weighty lines (NewYorkDeco). They may also have some elaborateness. It appears that any font can be made to fit this category simply by making it thicker. For example, Canon (see Table 5) uses a fairly simple font with thick lines. Again the regression results would predict fonts are made more prominent by increasing weight (although the regression results suggest these fonts should be less natural and have less flourish than was found with the cluster analysis). The fifth cluster of fonts is low in reassuring (i.e., unsettling), displeasing, and yet engaging while average on the prominent dimension. These fonts will be interesting, emotional, exciting, informal, dishonest, unfamiliar, innovative, cold, disliked and unattractive (AluminumShred). Consistent with the regression results, these responses are created by designs low in harmony, below average in naturalness, and above average in elaborateness. Like cluster three, these are fonts that are unlikely to be used heavily by corporations unless they want to convey negative information (e.g., fear appeal ad) or are targeting a niche market. For example, Abominable Records (see Table 5) carries bands like Helicopterejectionseat which plays, "an eclectic species of mathematical rock and/or roll. The final profile is highly reassuring, but low in engaging (i.e., boring). It comprises fonts that are average on pleasing and prominent. This cluster contains many common, highly readable fonts (Georgia). Consistent with regression results, they are low on elaborateness and

20

high on harmony. Such fonts are commonly used in the content of print ads as well as by stalwarts of the community. Just looking at the logotype of Mark Rushing & Associates (see Table 5) conveys this reassuring impression. The cluster analysis results provide evidence of the tradeoffs necessary when selecting fonts commercially available fonts. Yet, further examination of the regression results reveals that new fonts can be created to achieve additional arrays of response profiles. The regression results provide guidance to corporations for enlisting graphic designers to modify existing fonts or create new, corporation-specific fonts, rather than using commercially available fonts that may limit the combination of impressions created. Following are guidelines for creating response arrays beyond those described above. Designing Corporation-Specific Fonts Several strategically attractive response profile options emerge from examining the regression analyses. Pleasing, Engaging and Prominent Fonts (average on reassuring) - This combination is very similar to Cluster 1 except it is prominent rather than subtle. As noted earlier, simply making the lines thicker can make any font more prominent. For example, Fluf is a much more prominent font than the similar Kidstuff. Disney (see Table 5) uses this strategy to create a more prominent looking logotype than is common for cluster 1 designs. Pleasing, Reassuring and Prominent Fonts (average on engaging) - Creating pleasing and reassuring fonts should be fairly easy since no tradeoff is required. Pleasing fonts are natural and simple while reassuring fonts are harmonious and simple. In addition, while harmony does not have a large effect on pleasingness, it tends to be positive. The same is true for the effect of naturalness on reassuringness. Thus, pleasing, reassuring fonts can be created through with

21

natural, simple, harmonious designs. These fonts can be made more prominent simply by making the lines thicker (which does not affect evaluations of pleasing or reassuring). Surprisingly, none of the fonts in our sample had the design characteristics needed to create pleasing and reassuring fonts. All the fonts that were high in both naturalness and harmony were also very elaborate. The closest example to the desired font was Hamburger which was high in naturalness but only average in harmony and was slightly below average in elaborateness. Hallmark (see Table 5) is a good example of this type of design. Pleasing, Reassuring and Subtle Fonts (average on the engaging) - As noted above, creating pleasing and reassuring fonts is fairly easy. To also make them subtle, lines should be thin and natural. There was no example of this type of font in our sample, and it was extremely difficult to find a corporate example. However, Imagination Web Design (see Table 5) provides a good example. Combining Fonts - A final option is to combine fonts with different response profiles to create a hybrid response. For example, first initials and delimiters might be used to create a sense of pleasingness and subtlety. This could be followed by generic fonts to create a reassuring impression. Elkins and Associates1 provides a good example of just this strategy. Similar approaches might be used to grab attention, through engaging first initials, without making the typeface overly unpleasant. Creating Differentiation through Typeface Design The nine design profiles of typefaces, along with the added possibility of combining fonts from different profiles, provide corporations with great flexibility in achieving communication goals and creating differentiated marketing materials. Even greater flexibility and creativity is

22

afforded corporations in that they may achieve the described design dimensions by manipulating any of the design characteristics underlying those dimensions. For example, if a corporation chooses to use a natural design in order to communicate a pleasing, subtle image; a typeface may be chosen or created that emphasizes any one of several aspects of naturalness, such as it lookshandwritten, exhibits curvature, or is slanted. The typeface need not have all characteristics of naturalness to be high on this dimension. Thus, rather than constrain a corporation in its use of typeface, the guidelines serve to direct the already extensive work that goes into selecting, modifying, and using typefaces towards accomplishing the goals corporations state they desire. In addition, by simplifying the myriad of characteristics graphic designers consider into six design dimensions, by simplifying the numerous impressions corporations desire into four response dimensions, and by combining these into nine broad typeface profiles, these guidelines should serve to improve the communication of corporate and brand image goals between executives and their creative partners. Finally, the guidelines should improve a corporations ability to distinguish themselves in meaningful ways from the communications of their competitors. Without guidance on the design dimensions and profiles that distinguish typefaces, it would be difficult to determine whether a selected font differentiates the firm from competitors or conveys an identical response profile. These guidelines should provide greater insight to both designers and corporations in the audits they regularly conduct of their competitions communication materials. Creating Integrated Communication Campaigns The guidelines are useful in creating any type of marketing communication that uses print. In addition, these guidelines should assist corporations in creating and integrating their entire

23

portfolio of communications that use the written word. Corporations are increasingly sensitive to integrating communications across a wide variety of media. Our findings suggest that typeface is a medium with its own message. This makes it critical that the fonts message/impression be chosen carefully and held consistent across the wide variety of communications in which a corporation engages. In some cases, the firms preferred font may not be appropriate for a particular medium, e.g., a billboard, or a web page. By following the guidelines provided, corporations can choose alternative fonts that still have the same response profile. This will increase their flexibility while simultaneously increasing the consistency in the messages they are communicating. While this would not be appropriate for a brand logo which must appear the same in all communications, it is appropriate for other forms of written communication a corporation generates. Future Research While this study has sought to provide corporations with guidelines for managing impressions created by their designed materials, more research is needed in the area of design, in general, and typeface, in particular. Fundamentally, research is needed to determine if people have an innate, hardwired inclination to respond to design based on daily interactions with the natural world (Colarelli and Dettmann 2003). This research should focus on the consistency with which people respond to design elements across a range of stimuli and purposes. Another productive area for future research is in examining design characteristics and their influence on perception and information processing; e.g., Do elaborate fonts attract greater attention?; Do elaborate fonts increase cognitive load and require greater processing time?. In addition, future research should establish baseline or control responses in order to assess whether typeface design enhances or

24

undermines impressions, as well as to determine the degree to which design can hurt or help achieve desired responses. It is important to determine the impact of the impressions created by typeface on other responses of interest to corporations such as brand attitudes, customer retention, click-through behaviors on web sites, purchase behavior, and corporate identity. Initial research indicates that typeface affects important responses to advertising (e.g., Childers and Jass 2002; McCarthy and Mothersbaugh 2002). More research is needed to determine the extent of impression transfer from typeface to the brand and company itself, as well as its impact on the variety of responses and behaviors studied in marketing. Another avenue of research will be to explore the extent to which responses to typeface and other designed stimuli vary across individuals. The approach taken in the current paper was a stimulus-level analysis which is particularly appropriate in marketing where efforts are directed at entire segments of consumers. Still, research indicates that individual differences can impact attentiveness to aesthetics (Bloch, Brunel, and Arnold 2003). In addition, there may be differences across nations as some countries write using symbols or characters opposed to an alphabet, which may impact perceptions of letters in alphabetic name brands (Pan and Schmitt 1996). Interestingly, research indicates that perceptions of logos are fairly similar between the U.S. and Asia (Henderson et al. 2003) , but it is unclear if this will hold with typeface. Research is also needed to determine the public policy implications of typeface selection. For example, typeface selection is of special concern in providing information to the elderly. While the focus is typically on readability for the elderly, variables such as engagingness may weigh heavily in the extent to which warning labels are read.

25

More generally, research is needed on the relationship between design and response for other design objects. Empirically-based guidelines are needed to help corporations manage, in a more informed and strategic fashion, their entire design portfolio including products, advertisements, packaging, websites, signage and physical design such as retail outlets. The present research and previous research on logos (Henderson and Cote 1998) suggests there may be universal design dimensions that are generalizable across stimuli. In addition, the responses to these designs may also be relatively universal and generalizable. As such, the beginning elements of examining design impact for other classes of design stimuli and assessing the further impact of design on a wider variety of responses, may be identified. Such research can only help to improve the profitability of design for corporations while providing the basis for a more universal theory of design.

26

TABLE 1 Examples of Design Characteristics Design Characteristic Ornate vs. Plain Special Use vs Common Use Depth vs. Flat Distinctive vs. Not Distinctive Conveys Meaning-Does Not Convey Meaning Readable vs. Not Readable Balanced vs. Unbalanced Smooth vs. Rough Symmetric vs. Asymmetric Uniform vs. Not Uniform Organic vs. Geometric2 Looks Typed vs. Looks Handwritten Active vs. Passive Slanted vs. Straight Curved vs. Angular Heavy vs. Light Short and Fat vs. Tall and Thin Repeated vs. no Repeated Elements Serif vs. Sans Serif Ascenders Pronounced vs. Not Pronounced3 Descenders Pronounced vs. Not Pronounced Condensed vs. Extended4 x height Tall vs Short5 High Low Universal - Specifc Univeral Specific Univeral Univeral Univeral Specific Univeral Univeral Univeral Univeral Univeral Specific Univeral Univeral Univeral Specific Specific Specific Specific Design Factor Elaborate Elaborate Elaborate Elaborate Elaborate Elaborate Harmony Harmony Harmony Harmony Natural Natural Natural Natural Natural Weight Weight Weight Flourish Flourish Flourish Compressed Compressed

ABCYabcy

ABCYabcy ABCYabcy ABCYabcy ABCYabcy ABCYabcy ABCYabcy ABCYabcy ABCYabcy ABCYabcy ABCYabcy ABCYabcy
ABCYabcy

ABCYabcy
TUVltuv ABCYabcy ABCYabcy
ABCYabcy

ABCYabcy
ABCYabcy ABCYabcy

ABCYabcy ABCYabcy ABCYabcy

ABCYabcy ABCYabcy

ABCYabcy
ABCYabcy
ABCYabcy ABCYabcy

ABCYabcy

ABCYabcy
ABCYabcy ABCYabcy

ABCYabcy

ABCYabcy
ABCYabcy

ABCYabcy
ABCYabcy

ABCYyabcy Specific ABCYabcy ABCYabcy ABCYabcyx Specific Specific Specific

ABCYabcy
ABCYabcy ABCYabcyx

The authors can provide a more complete set of examples including examples for the response variables.

2 3

4 5

Organic looks more irregular, unplanned or natural while geometric looks more like objects that are man-made, planned, or measured. Ascenders (descenders) are the parts of the letter that go above (below) the main body e.g., the top of a lower case h (the tail of the lower case y). Pronounced ascenders (descenders) appear to go significantly above (below) the body of the letter or stand out in their influence on the appearance of the letter. Condensed refers only to the width of the letter. Condensed letters are narrow while extended letters have a wider base. x-height refers to the height of the lower-case x. Tall letters are letters whose tops are far above the height of the x. Short letters do not rise above the height of the x very far.

27

TABLE 2 Confirmatory Factor Analysis of the Response Dimensions Pleasing .834 .838 .855 Engaging Reassuring Prominent

Like-Dislike Warm-Cold Attractive-Unattractive Interesting-Uninteresting Emotional-Unemotional Calm-Not Calm Formal-Informal Honest-Dishonest Familiar-Unfamiliar Innovative-Mainstream Strong-Delicate Masculine-Feminine

.892 .886 .919 .880 .921 .916 -.826 .591 1.000

Correlations among Response Dimensions Pleasing Engaging Reassuring Prominent 1 Pleasing .341 1 Engaging .607 -.589 1 Reassuring -.543 -.578 .044 1 Prominent
chi-square = 536.540 based on 48 degrees of freedom probability value for the chi-square statistic is less than .001 the normal theory rls chi-square for this ml solution is 447.318. Bentler-Bonett normed fit index = .806 Bentler-Bonett nonnormed fit index = .751 comparative fit index (CFI) = .819

28

TABLE 3 Exploratory Factor Analysis of the Design Dimensions

DistinctiveNot Distinctive Ornate-Plain Special UseCommon Use Readable-Not Readable Conveys Meaning-Does Not Convey Meaning Depth:Flat-Multidimensional Uniform-Not Uniform Balanced-Unbalanced Smooth-Rough Symmetric-Asymmetric Curved-Angular Organic-Geometric Slanted-Straight Looks Handwritten-Looks Typed Active-Passive Serif-Sans Serif Ascenders :Pronounced-Not Pronounced Descenders :Pronounced-Not Pronounced Heavy-Light Repeated ElementsNo Repeated Elements Short and Fat-Tall and Thin Condensed-Extended x-Height:Tall-Short Elaborate Harmony Natural Flourish Weight Compressed -.247 .132 .787 .780 -.720 -.711 .687 .659 -.213 -.218 -.371 -.406 .180 .243 .116 -.307 .380 -.334 .449 .475 -.081 -.101 .816 .759 .693 .600 .313 -.373 -.215 .431 -.503 .244 -.180 -.076 .296 .208 -.257 -.277 .135 -.353 .804 .742 .720 -.718 .615 .754 .720 .639 -.333 .307 -.382 .001 .281 .158 -.226 -.197 .771 .705 .677 -.024 -.017 .008 .137 .156 .033 -.180 .397 -.817 .583

Principal Component Analysis with Varimax Rotation Elaborate, Harmony, and Natural explained 69.7% of the variance in the perception design characteristics Flourish, Weight, and Compressed explained 60.4% of the variance in the quantifiable design characteristics

29

TABLE 4 Summary of Regression Results Design Elements Natural Pleasing Elaborate Harmony Flourish Compressed Natural Elaborate Engaging Harmony Compressed Flourish Harmony Reassuring Elaborate Flourish Natural Weight Flourish Harmony positive, linear positive, linear positive, linear negative at a decreasing rate positive, linear positive, linear positive, linear negative, linear positive, linear increasingly negative positive, linear negative, linear Direction of Effect positive, plateaus at high values negative, linear Beta Coefficient .428 - .138N2 -.275 .029 + .208H
3

Size of Effect (adjR2) .320 .079 .045 .039 .031 .268 .211 .170 .018 .013 .385 .330 .018 .291 .171 .043 .024

Total Adj R2

.514

.209 .118 .128S2 .412 .403 -.361 + .0645H2 .153 .126 .586 -.600 .168 -.436 .150N2 .329 -.220 .258 .229H3

.680

.734

Prominent

.528

30

TABLE 5 Cluster Analysis Results

Cluster Size Responses
Pleasing Engaging Reassuring Prominent

Level
High High Average Low

Design
Elaborate Natural Harmony Flourish Compressed Weight Elaborate Natural Harmony Flourish Compressed Weight Elaborate Natural Harmony Flourish Compressed Weight Elaborate Natural Harmony Flourish Compressed Weight Elaborate Natural Harmony Flourish Compressed Weight Elaborate Natural Harmony Flourish Compressed Weight Elaborate Natural Harmony Flourish Compressed Weight Elaborate Natural Harmony Flourish Compressed Weight Elaborate Natural Harmony Flourish Compressed Weight

Level
Average High Average High Average Low Above Average High Low Average High Average Above Average Low Average Below Average Low Above Average Above Average Average Average Average Average High Above Average Below Average Low Below Average Average Average Low Average High Average Average Average Average High Average High Average High Low High High Average Average High Low High High Average Average Low

Fonts
Scheherezade Informal Roman AncientScript Enviro Pepita MT Baphomet Edda Chiller Stonehenge Paintbrush

Examples

37

37

Pleasing Engaging Reassuring Prominent

Average High Low Average

21

Pleasing Engaging Reassuring Prominent

Low Low Average Average

Playbill Logan Onyx Industria Inline StencilSet

41

Pleasing Engaging Reassuring Prominent

Average Average Average High

NewYorkDeco Bandstand SunSplash Middle Ages Fisherman AluminumShred BigDaddy Integrity Ransom Amazon Georgia Verdana Janson Text Century Gothic Times New Roman Century Schoolbook Maiden Word Author 6 Viner Hand ITC

19

Pleasing Engaging Reassuring Prominent

Low High Low Average

54

Pleasing Engaging Reassuring Prominent

Average Low High Average

Pleasing Engaging Reassuring Prominent Pleasing Engaging Reassuring Prominent Pleasing Engaging Reassuring Prominent

High High Average High High Average High High High Average High Low

Hamburger

None in Dataset

Author and Viner Hand ITC are presented in bold to make them more consistent with the design characteristics needed to create a prominence response.

31

REFERENCES Berlyne, Daniel E. (1971), Aesthetics and Psychobiology. New York NY: Meredith Corporation. ---- (1974), Studies in the New Experimental Aesthetics. New York: Wiley. Bloch, Peter H. (1995), "Seeking the Ideal Form: Product Design and Consumer Response," Journal of Marketing, 59 (July), 16-29. Bloch, Peter H., Frederic F. Brunel, and Todd J. Arnold (2003), "Individual Differences in the Centrality of Visual Product Aesthetics: Concept and Measurement," Journal of Consumer Research, 29 (March), 551-65. Bornstein, Robert F. and Paul R. DAgostino (1992), "Stimulus Recognition and the Mere Exposure Effect," Journal of Personality and Social Psychology, 63 (October), 545-52. Bradshaw, John L. (1984), "A Guide to Norms, Ratings and Lists," Memory and Cognition, 12 (2), 202-06. Childers, Terry L and Jeffrey Jass (2002), "All Dressed Up with Something to Say: Effects of Typeface Semantic Associations on Brand Perceptions and Consumer Memory," Journal of Consumer Psychology, 12 (2), 93-106. Colarelli, Stephen M. and Joseph R. Dettmann (2003), "Intuitive Evolutionary Perspectives in Marketing Practices," Psychology and Marketing, 20 (9), 837-65. Craig, James (1980), Designing with Type: A Basic Course in Typography. New York: WatsonGuptill Publications. Dolen, Michael (1984), "Creating Effective Logos with Words-Graphics Teamwork," Business Marketing, 69 (December), 92-6. Ernst, Sandra B. (1977), The ABC's of Typography. New York: Art Direction Book Company. Eysenck, Hans J. (1988), "Personality and Scientific Aesthetics," in The Foundations of Aesthetics, Art, and Art Education, Frank H. Farley and Ronald W. Neperud, eds. New York: Praeger Publishers. Henderson, Pamela W. and Joseph A. Cote (1998), "Guidelines for Selecting or Modifying Logos," Journal of Marketing, 62 (April), 14-30. Henderson, Pamela W., Joseph A. Cote, Siew Meng Leong, and Bernd Schmitt (2003), "Building Strong Brands in Asia: Selecting the Visual Components of Image to Maximize Brand Strength," International Journal of Research in Marketing, 20 (4), 297313.

32

Hertenstein, Julie H. and Marjorie B. Platt (2001), "Valuing Design: Enhancing Corporate Performance Through Design Effectiveness," Design Management Journal, 12 (Summer), 10-9. Hinrichs, Kit and Delphine Hirasuna (1990), Typewise. Cincinnati, OH: F&W Publications, Inc. Hirschman, Elizabeth C. (1980), "Innovativeness, Novelty Seeking, and Consumer Creativity," Journal of Consumer Research, 7 (3), 283-95. Hutton, Jim (1987), "How to Think Corporate Identity," Public Relations Journal, 43 (May), 258. ---- (1997), "The Influence of Brand and Corporate Identity on Consumer Behavior: A Conceptual Framework," Journal of Brand Management, 5 (November), 428-39. Klinger, Mark R. and Anthony G. Greenwald (1994), "Preferences need no Inferences: The Cognitive Basis of Unconscious Mere Exposure Effects," in The Hearts Eye: Emotional Influences in Perception and Attention, Paula M. Niedenthal and Shinobu Kitayama, eds. San Diego: Academic Press. Martindale, Colin (1988), "Aesthetics, Psychobiology, and Cognition," in The Foundations of Aesthetics, Art, and Art Education, Frank H. Farley and Ronald W. Neperud, eds. New York: Praeger Publishers. Martindale, Colin and Kathleen Moore (1988), "Priming, Prototypicality and Preference," Journal of Experimental Psychology: Human Perception and Performance, 14 (4), 66170. McCarthy, Michael S. and David L. Mothersbaugh (2002), "Effects of Typographic Factors in Advertisning-Based Persuasion: A General Model and Initial Empirical Tests," Psychology and Marketing, 19 (July-August), 663-91. Mehrabian, Albert and James Russell (1974), An Approach to Environmental Psychology. Cambridge, MA: The MIT Press. Ohanian, Roobina (1990), "Construction and Validation of a Scale to Measure Celebrity Endorsers Perceived Expertise, Trustworthiness, and Attractiveness," Journal of Advertising, 19 (39-52.). O'Leary, Noreen (1987), "Legibility Lost: Why Are So Many Ads So Hard to Read?," Adweek, October 5, D7-D10. Osgood, Charles E., George J. Suci, and Percy H. Tannenbaum (1957), The Measurement of Meaning. Urbana, IL: University of Illinois Press.

33

Pan, Yigang and Bernd H. Schmitt (1996), "Language and Brand Attitudes: Impact of Script and Sound Matching in Chinese and English," Journal of Consumer Psychology, 5 (3), 26377. Rowe, Camille L. (1982), "The Connotative Dimensions of Selected Display Typefaces," Information Design Journal, 3 (1), 30-37. Shapiro, Stewart (1999), "When an Ads Influence is Beyond Our Conscious Control: Perceptual and Conceptual Fluency Effects Caused by Incidental Ad Exposure," Journal of Consumer Research, 26 (June), 16-36. Solomon, Martin (1991), "Typography: More Than Words," in Designer's Guide to Typography, Nancy Aldrich-Ruenzel and John Fennell, eds. New York: Watson-Guptill Publications. Somerick, Nancy M. (2000), "Practical Strategies for Avoiding Problems in Graphic Communication," Public Relations Quarterly, 45 (3), 32-4. Spaeth, Tony (1994), "Do Logos Really Matter," Across the Board, 31 (March), 51-3. ---- (1999), "Powerbrands," Across the Board, 36 (2), 23-28. ---- (1995), "What Does It All Mean?," Across the Board, 32 (February), 53-5. Tannenbaum, Percy H., Harvey K. Jacobson, and Eleanor L. Norris (1964), "An Experimental Investigation of Typeface Connotation," Journalism Quarterly, 41, 65-73. Tantillo, John, Janet DiLorenzo-Aiss, and Richard E. Mathisen (1995), "Quantifying Perceived Differences in Type Styles: An Exploratory Study," Psychology & Marketing, 12 (5), 447-57. Tavassoli, Nader T. (2001), "Color Memory and Evaluations for Alphabetic and Logographic Brand Names," Journal of Experimental Psychology: Applied, 7 (June), 104-11. ---- (2002), "Spatial Memory for Chinese and English," Journal of Cross-Cultural Psychology, 33 (July), 415-31. Tavassoli, Nader T. and Jin K. Han (2002), "Auditory and Visual Brand Identifiers in Chinese and English," Journal of International Marketing, 10 (2), 13-28. ---- (2001), "Scripted Thought: Processing Korean Hancha and Hangul in a Multimedia Context," Journal of Consumer Research, 28 (December), 482-93. Van den Bergh, Omer and Scott R. Vrana (1998), "Repetition and Boredom in a Perceptual Fluency/Attributional Model of Affective Judgment," Cognition and Emotion, 12 (4), 533-53.

34

Veryzer, Robert W. Jr. (1999), "A Nonconscious Processing Explanation of Consumer Response to Product Design," Psychology and Marketing, 16 (6), 497-522. Veryzer, Robert W. Jr. and J. Wesley Hutchinson (1998), "The Influence of Unity and Prototypicality on Aesthetic Responses to New Product Designs," Journal of Consumer Research, 24 (March), 374-95. Walker, Peter, Sylvia Smith, and Alan Livingston (1986), "Predicting the Appropriateness of a Typeface on the Basis of its Multimodal Features," Information Design Journal, 5 (1), 29-42. Wallace, Rob (2001), "Proving our Value: Measuring Package Designs Return on Investment," Design Management Journal, 12 (Summer), 20-7. Whitfield, T. W. A. (1983), "Predicting Preference for Familiar, Everyday Objects, An Experimental Confrontation between two Theories of Aesthetic Behaviour," Journal of Environmental Psychology, 3, 221-37. Whitfield, T. W. A. and P. E. Slater (1979), "The Effects of Categorization and Prototypicality on Aesthetic Choice in a Furniture Selection Task," British Journal of Psychology, 70, (February) 65-76. Whittlesea, Bruce W. A. (1993), "Illusions of Familiarity," Journal of Experimental Psychology: Learning, Memory, and Cognition, 19 (November), 1235-53.

35