Computers in Human Behavior 38 (2014) 240–247

Contents lists available at ScienceDirect

Computers in Human Behavior

journal homepage: www.elsevier .com/locate /comphumbeh

The effects of home page design on consumer responses: Moderatingrole of centrality of visual product aesthetics

http://dx.doi.org/10.1016/j.chb.2014.05.0300747-5632/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +1 (252)737 2874; fax: +1 (252)328 5655.E-mail addresses: yooj@ecu.edu (J. Yoo), Minjeong.Kim@oregonstate.edu

(M. Kim).1 Tel.: +1 (541)737 3468; fax: +1 (541) 737 0993.

Jungmin Yoo a,⇑, Minjeong Kim b,1

a Department of Merchandising, East Carolina University, Greenville, NC 27858, USAb College of Business, 228 Milam Hall, Oregon State University, Corvallis, OR 97331, USA

a r t i c l e i n f o

Article history:

Keywords:Home page designVisual fluencyCentrality of visual product aestheticsConsumer behaviorOnline retailing

a b s t r a c t

This study examined the effect of home page design on consumer responses based on the Hierarchy ofEffects model and impression formation theory. The design of the study was a one factor (home pagedesign: image- vs. text-oriented) between-subjects design with two moderators, brand familiarity andthe centrality of visual product aesthetics (CVPA). College women (N = 658) participated in this onlineexperiment. The findings revealed that: (1) an image-oriented design is more effective in enhancing ahome page’s visual fluency, and (2) people in both high and low CVPA groups preferred an image-ori-ented home page to a text-oriented home page.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Online advertising spending in the U.S. reached $32 billion in2012 with a 23.3% annual growth rate and is projected to reach$62 billion by 2016 (Go-Gulf, 2012), rapidly outpacing the growthrates of traditional advertising media like print and TV. In today’sera of multichannel retailing, the retailer’s home page plays a keyrole in this new advertising medium (Singh & Dalal, 1999). Asthe main entry point to the website, a home page not only drawsconsumers into the retailer’s website, but also builds a brandimage through the stories being communicated on the page(Lindgaard, Fernandes, Dudek, & Brown, 2006; Schenkman &Jönsson, 2000).

During the last decade, website design has received growingattention from scholars and practitioners. Extant online literatureand industry reports provide a wealth of information regardingthe creation of effective online product presentations that will leadto positive consumer responses (Blanco, Sarasa, & Sanclemente,2010: Kim & Lennon, 2008; Yoo & Kim, 2012). On the other hand,little is known about the factors involved in effective home pagedesign. Although home pages do not directly generate sales, theyhave become an effective advertising medium that attracts cus-tomers into the website (Singh & Dalal, 1999), building brand/storeimages and, most importantly, creating positive impressions that

last. Further research on the design of website home pages is there-fore clearly warranted.

Drawing on an advertising theory (the Hierarchy of Effects) anda psychology theory (impression formation), this study aims to (1)identify key design factors that determine the effectiveness of aretailer’s home page in creating a positive impression, (2) explainthe psychological process by which a home page impacts consum-ers’ behavioral intentions, and (3) examine how individual charac-teristics moderate the impact of this process. Postulating thatvisual fluency is a key factor in creating positive impressions ofan online store, this study’s objectives are to: (1) identify the homepage design factors that impact the visual fluency of a home page,(2) investigate the process by which visual fluency influencesconsumer preference and behavioral intentions, and (3) examinewhether individual characteristics such as brand familiarity andcentrality of visual product aesthetics (CVPA) moderate the effectof home page design on consumer responses.

Empirical findings from this study are expected to contribute tothe emerging literature related to online advertising. Consideringthe high costs associated with attracting new customers to anonline store (Evans, 2008), the findings will provide useful infor-mation for retailers seeking to build an effective home page thatwill bring consumers into their store by creating positive impres-sions and, ultimately, contributing to building a brand image.

2. Conceptual development

This section describes the study’s theoretical framework, whichintegrates the Hierarchy of Effects (HOE) model (Lavidge & Steiner,

Fig. 1. Proposed model of the study.

J. Yoo, M. Kim / Computers in Human Behavior 38 (2014) 240–247 241

1961) and impression formation theory (Asch, 1946) to explain theprocess by which a website home page influences consumerbehavior.

2.1. Theoretical framework

According to the HOE model, consumers take three steps torespond to advertising messages: cognition – affect – conation(Lavidge & Steiner, 1961). Cognition is defined as ‘‘a system ofbeliefs structured into some kind of semantic network’’(Holbrook & Batra, 1987, p. 405), while affect signifies theconsumer’s emotion or feelings and conation refers to their behav-ioral intentions, leading to actual behavior (Barry & Howard, 1990).Depending on the context, the sequence of these three steps in thetraditional HOE model may change. For example, for low involve-ment situations, Krugman (1965) suggested a cognition – conation– affect order, whereas Zajonc and Markus (1982) proposed affect-conation – cognition based on their views of affect primacy. Fol-lowing the traditional HOE model, the current study posited thathome page design as an advertisement influences consumers’ eval-uation of the visual fluency (cognition) of a home page, which thenhas an impact on their preference for a home page (affect) and,ultimately, on their behavioral intentions (conation).

Impression formation theory (Asch, 1946) also provides usefulinsights into understanding how home page design influences con-sumer behaviors in online retailing. Consumers form a positive ornegative impression about a brand even when they are exposedto only a small piece of information about a brand. This processis referred to as impression formation (Aquirre-Rodriguez,Bosnjak, & Sirgy, 2012). In traditional retailing, consumers formfirst impressions and observe or infer store-related informationfrom the store’s window displays (Sen, Block, & Chandran, 2002).In online retailing, a website home page is posited to have a similarimpact on the way online shoppers form their impressions of anonline store. As with window displays in traditional retailing, ahome page can communicate information about product offerings,events and promotions, resulting in the creation of certainimpressions about a website.

Although home page designs constantly change to reflect newproducts, seasonal events and/or promotions, three dimensionsprimarily determine a consumer’s perception of a website; (1)the information available on the site, (2) the usability of the site,and (3) the impression elicited from the site. Of these, the firstimpression of the website is what draws people in to spend moretime on the website (Schenkman & Jönsson, 2000). Lindgaardet al. (2006) found that first impressions of a home page are madeinstantaneously, within 50 ms, and these quickly formedimpressions remain remarkably consistent over time.

Furthermore, there is a so-called ‘halo effect’ of first impressionscreated by the visual appeal of websites. Psychology researchgenerally supports a long-term effect of first impressions in thatpeople often make judgments based on their positive first impres-sions (Lindgaard et al., 2006). Additionally, ‘confirmation bias’corroborates that people tend to seek out confirmatory facts sup-porting their initial assumption, while ignoring disconfirmatoryfacts (Nisbett & Ross, 1980). Positive impressions of a website werefound to improve a consumer’s usability perception despite actualpoor usability experiences (Lindgaard & Dudek, 2003). Campbelland Pisterman (1996) suggested that positive impressions from ahome page draw attention away from negative issues experiencedin subsequent web pages. Thus, it is critical for retailers to design ahome page that creates a positive first impression of a brand orstore.

Previous research has shown that visual images play an impor-tant role in creating a positive impression of the retailer andfacilitating positive emotional experiences (Park, Choi, & Kim,

2005; Tractinsky & Lowengart, 2007), which then affect consum-ers’ subsequent decision making processes (Janiszewski, 1993;Kim & Moon, 1998). Applying the HOE model and impression for-mation theory to the online retailing setting, this study postulatesthat a visually fluent home page will create a positive impressionof an online store, impacting consumers’ affective response (e.g.,preference) and, ultimately, their conative response (e.g., behav-ioral intentions) with regard to that online store (see Fig. 1). Thefollowing section provides a detailed review of the literaturesupporting the development of the study hypotheses.

2.2. Home Page Design and Consumer Responses

Advertising research largely supports the contention thatpictures are more easily recalled and recognized than text inadvertisements (Paivio, 2007). Rossiter and Percy (1978) foundthat image-oriented advertisements had a more positive effect onconsumers’ affective learning than text-oriented advertisements.Similarly, Hirschman (1986) found that consumers perceivedimage-oriented advertisements as being more aesthetic and emo-tional than text-oriented advertisements and that image-orientedadvertisements increased consumers’ perceptions of familiarity.

Compared to traditional media such as newspapers and maga-zines, websites provide more flexibility and capacity to createand communicate information using pictures and text (Wiebe &Howe, 1998). Currently, some home pages are predominantlyimage-oriented with dominant visual images and minimal text,whereas others are predominantly text-oriented, containing manytext-based navigational links to other pages on the website andminimal image content. The results of a content analysis of thetop 50 online retailers’ websites (Internet Retailer, 2011) revealedthat more than half the websites analyzed were text-oriented andless than third were image-oriented. However how pictures andtext on home pages contribute to the creation of positiveimpressions remains largely conjectural.

A retail home page is like a store front that people pass by whilebrowsing the Internet. Rather than engaging in high cognitiveworks, consumers tend to glance over a home page and decide tobrowse further or go to another website. Based on the way con-sumers interact with home pages and the extant research, thisstudy posits that the visual fluency of the page plays a critical rolein forming an impression and enticing consumers into browsing awebsite.

2.2.1. Visual fluencyVisual fluency is defined as ‘‘the ease with which visual stimuli

are processed’’ (Winkielman, Schwarz, Reber, & Fazendeiro, 2003, p.77). When exposed to visual stimuli, people perform cognitivework to process the information. The amount of cognitive work

242 J. Yoo, M. Kim / Computers in Human Behavior 38 (2014) 240–247

required varies depending on the characteristics of the visual stim-uli. For example, both the clarity and familiarity of visual stimuliinfluence the ease of information processing because clearer andmore familiar visual stimuli are easier to process and are thusprocessed faster and more accurately (Jacoby, Kelley, & Dywan,1989).

Winkielman et al. (2003) discussed two levels of informationprocessing for visual fluency: perceptual and conceptual fluency.Perceptual fluency refers to low-level perceptual processes dealingwith features of the target (Jacoby et al., 1989). In general, repeti-tion, contrast, form priming, or duration influences perceptual flu-ency, thus increasing the speed and accuracy of perceptualrecognition (Jacoby, 1983). In contrast, conceptual fluency isrelated to high-level interpretation processes such as processingthe meaning of a target and is therefore influenced by semanticpriming, semantic predictability, context congruity, and rhyme(Whittlesea, 1993). Perceptual and conceptual fluency are inter-connected in that both generate similar outcomes on judgmentsand each domain of fluency influences the other domain. For exam-ple, perceptual manipulation using repetition affects conceptualjudgments of truth (Jacoby et al., 1989), whereas conceptualmanipulation using semantic priming influences perceptual judg-ments, such as visual clarity (Masson & Caldwell, 1998). Giventhe nature of a website home page, where low-level interpretationis generally required, this study focused on perceptual fluency as acognitive response to a home page.

2.2.2. Visual fluency of a home pageThe importance of images is well supported in the prior litera-

ture (Levine, 1995; Wiebe & Howe, 1998). According to Hodes(1994), properly designed visual images are dominant tools forinformation encoding. Images help people synthesize a largeamount of disparate information on a web page and review andorganize that information while expending minimal resources(Wiebe & Howe, 1998). Hirschman (1986) demonstrated that peo-ple require less cognitive effort to process visual images than forcomparable verbal information and suggested that visual imagesare more appropriate in advertising, especially when people feeluncertain and unfamiliar with products.

In advertising, images are generally more visually fluent thantexts (Hirschman, 1986). Similarly, an image-oriented home pageis likely to be more visually fluent than a text-oriented page. Peo-ple tend to be overwhelmed when exposed to hyper-text condi-tions because of the vast amount of information to process (Vora,1998). When processing information from text, people make a ser-ies of rapid eye movements while keeping their eyes fixed on a partof the text to obtain useful information between eye movements.Hence, the information obtained from the text is an accumulationof information gathered while pausing on sections of the textbetween eye movements (Marks & Dulaney, 1998). According toVora, people can process only a certain amount of informationbecause they select and process only the information relevant totheir needs.

In this study, an image-oriented home page refers to a homepage predominantly composed of visual images with minimal text(e.g., brand names), whereas a text-oriented home page refers to ahome page that is predominantly text-based with minimal images(e.g., brand logos). Based on the review of the literature, an image-oriented home page is posited to be more visually fluent than atext-oriented home page. A visually fluent home page is expectedto facilitate consumers’ processing of information and create morepositive impressions than a less visually fluent page.

H1. An image-oriented home page is more visually fluent than atext-oriented home page.

2.3. Individual differences

How people respond to a different home page design may differas a function of individual characteristics. The two main individualcharacteristics investigated in this study are brand familiarity andthe centrality of visual product aesthetics (CVPA).

2.3.1. Brand familiarityAs consumers’ accumulate experiences within a brand (Alba &

Hutchinson, 1987), brand familiarity is enhanced by frequent expo-sure to the brand and thus serves as a significant source of informa-tion about a brand. Consumers tend to consider a brand as familiarwhen the brand is frequently advertised in the media. Additionally,a well-known brand is more familiar to consumers and easier torecall and recognize than an unknown brand (Kent & Allen, 1994).The effects of mere exposure on consumer behaviors are widelysupported. The mere exposure effect explains that when a personis exposed to a stimulus repeatedly it creates certain positive effectsrelated to that stimulus (Zajonc & Markus, 1982).

The existing literature supports this positive relationshipbetween familiarity and visual fluency. Hirschman (1986) demon-strated that image-oriented advertisements increase consumers’ratings on familiarity because images are easier and more comfort-able to process than texts. Winkielman et al. (2003) agreed, argu-ing that familiarity boosts the visual fluency of an object becausepeople process previously exposed stimuli more readily than novelstimuli. Hence, repeated exposure stimulates information process-ing, which subsequently leads to faster stimuli recognition andclearer judgments than novel stimuli. When exposed to a familiarbrand home page, people may therefore find it easier to identifythe brand logo or design styles and to process information, increas-ing the visual fluency of a website. This study postulates that howwebsite design (image-oriented vs. text-oriented) impacts visualfluency will differ as a function of brand familiarity.

H2. The effect of home page design on visual fluency is greater fora familiar brand than an unfamiliar brand.

2.3.2. Centrality of visual product aesthetics (CVPA)According to Hartmann, Sutcliffe, and De Angeli (2007), users’

evaluation of website aesthetics differed depending on their back-grounds. The researchers examined the effects of participants’ aca-demic majors as an analogue for individual background and foundthat design students tended to rate website aesthetics as moreimportant than students taking technical courses, suggesting thatvisual aesthetics vary based on how important an individual con-siders design to be. This individual characteristic can be explainedby the CVPA (Bloch, Brunel, & Arnold, 2003), which refers to thesignificance level assigned by an individual to visual aestheticsand is considered to be a general consumer characteristic. CVPAis divided into three dimensions: (1) an individual’s perceivedvalue of aesthetic product design, (2) an individual’s ability to eval-uate product design, and (3) an individual’s different responses toproduct aesthetics. Although CVPA does not determine an individ-ual’s preference for a particular aesthetic style, consumers withhigh CVPA are more concerned with visual aesthetics than thosewith low CVPA, which influences their preference for particularbrands and products and causes them to tend to make productand brand inferences based on visual product aesthetics.

In the context of this study, high CVPA consumers are likely tohave a more positive response to an image-oriented home pagethan low CVPA consumers. Additionally, high CVPA consumersmay acquire information and make judgments faster on animage-oriented home page than low CVPA consumers, suggesting

J. Yoo, M. Kim / Computers in Human Behavior 38 (2014) 240–247 243

that CVPA influences perceptions of visual fluency. Thus, it is pos-ited that the CVPA of individual consumers will influence howwebsite design impacts visual fluency.

H3. The effect of image-oriented home page on visual fluency isgreater for high CVPA consumers, whereas the effect of text-oriented home page is greater for low CVPA consumers.

2.4. Visual fluency and preference

As ‘‘the outcome of a consumer’s evaluation process’’ (Lefkoff-Hagius & Mason, 1993, p. 102), preference is a consumer’s affectiveresponse to a stimulus and can explain how consumers chooseproducts. Prior research suggests that visual fluency plays a rolein influencing consumers’ liking of the stimulus. Reber,Winkielman, and Schwarz (1998) found that stimuli that are visu-ally fluent increase the preference for those stimuli, even with asingle exposure. In Winkielman et al.’s (2003) research, peopleshowed a preference for pleasing and familiar stimuli because theywere more visual fluent. For example, Winkielman et al. manipu-lated drawings of everyday objects such as cars, and animals andabstract objects such as dots, and circles to measure visual fluency.They also manipulated the symmetry of the pictures and the pre-sentation duration to examine the effects of fluency. The resultsshowed that people simply liked visual stimuli that were easy toprocess, regardless of how the visual objects were manipulated.Thus, the following hypothesis was developed.

H4. The visual fluency of a home page has a positive impact onconsumers’ preference for the website.

2.5. Visual fluency, preference, and behavioral intentions

Behavioral intentions refer to ‘‘what the person intends to do’’in a specific domain of behaviors (O’Keefe, 2002, p. 101). The pur-pose of this study is to examine the effect of home page design onconsumers’ behavioral intentions. On a home page, consumersoften decide whether to browse the website further or not. Theimpression formed on a home page can also determine whetherpeople are likely to revisit a website or not. Therefore, consumers’behavioral intentions in this study were operationalized as inten-tion to browse more and intention to revisit.

Compared to the effects of visual fluency on memory, recall,affective responses, and evaluative judgments, little research hasbeen done to identify the effects of visual fluency on behavioralintentions. In one recent study by Im, Lennon, and Stoel (2010),consumers’ perceptual visual fluency as a result of the visual qual-ity on apparel websites impacted their behavioral intentions. Con-sumers exposed to a website presenting high visual quality images(i.e., big pictures, black font color, and a sharp contrast betweenpicture foreground and background) showed higher behavioralintentions than those directed to a website with low visual quality(i.e. small pictures, gray font color, and low contrast with addednoise). Poorly presented visual information confused shoppersand complicated their decision making, subsequently loweringtheir behavioral intentions. Based on our review of the literature,the following hypothesis was therefore proposed.

H5. The visual fluency of a home page has a positive impact onconsumers’ behavioral intentions.

A great deal of research has demonstrated that an individual’sbehavioral intention is determined by their affective responses,such as preference (Morris, Woo, Geason, & Kim, 2002; O’Keefe,2002). Suh and Yi (2006) examined the effects of brand attitudes

(i.e., degree of liking and favorability) on the customer satisfac-tion-loyalty relationship (i.e., intention to repurchase and intentionto recommend) in an advertising context. Their results showedthat consumers’ preference (liking) significantly affected theirbehavioral intentions to repurchase and recommend. Advertisingresearch further supports the that brand preferences formedthrough advertising are fundamental for increasing consumers’purchase intentions (Cobb-Walgren, Ruble, & Donthu, 1995). Thusthe following hypothesis was developed.

H6. Consumer preference for a website has a positive impact onbehavioral intentions.

3. Method

An online experiment using mock websites was used for thisstudy. A one factor (website design: image-oriented vs. text-ori-ented) between-subjects design was utilized with two moderators:brand familiarity and CVPA.

For instrument development, visual fluency, preference, andCVPA measures were adopted from Labroo, Dhar, and Schwarz(2008), Costley and Brucks (1992), and Bloch et al. (2003), respec-tively, with adequate reliabilities (Cronbach’s alphas > .70). Allitems were based on a seven-point scale. Behavioral intentionswere measured using a two-item scale developed for this study:intention to revisit the website and intention to browse more prod-ucts on the website. Demographic items such as age, ethnicity, andonline shopping experiences were also included.

In order to avoid potential confounding factors (i.e., gender andage differences, levels of online shopping experience), this studyused college women. College students are frequent online shoppersand thus a good representative sample of online shoppers (Kim,Ferrin, & Rao, 2009; Pavlou & Fygenson, 2006). In particular, collegewomen frequently shop online for apparel products (Denis & Fenech,2004; Nielsen, 2011) and were thus deemed suitable for the study.

A pretest was conducted for the brand familiarity manipulationusing six brands chosen for their availability in the U.S. marketand their relevance to young female consumers: three familiar(Ralph Lauren, J. Crew, and Anthropologie) and three unfamiliar(Club Monaco, French Connection, and agnés b). College women(N = 83) participated in the pretest and answered questionsdesigned to measure brand familiarity. The results showed thatRalph Lauren was the most familiar brand (M = 4.72, SD = 1.78)and agnés b was the least familiar brand (M = 1.32, SD = .97), witha statistically significant difference (p < .05). Based on the pretestresults, four home pages were created for the main experiment:(1) an image-oriented home page + a familiar brand (Ralph Lauren),(2) an image-oriented home page + an unfamiliar brand (agnés b),(3) a text-oriented home page + a familiar brand (Ralph Lauren),and (4) a text-oriented home page + an unfamiliar brand (agnés b).Both image-oriented and text-oriented home pages used existingcommercial home pages that had been modified using Adobe Photo-shop to represent the selected brands and conditions (image vs. text-oriented). The mock websites created for this experiment reflectedtypical retail home pages that are either text- or image-oriented.

When participants logged on to the research website, they wererandomly assigned to one of the four experimental websites. Afterbrowsing the mock website, participants completed the surveyquestionnaire.

4. Results

Four thousand college women enrolled at a large U.S. universitywere invited to participate in the online experiment via emails. Ofthese, 658 participated in the web experiment, a response rate of

244 J. Yoo, M. Kim / Computers in Human Behavior 38 (2014) 240–247

16.5%. The majority of the participants were Caucasians (76.7%),with a mean age of 24.3 (SD = 8.49). In terms of Internet experi-ence, most participants used the Internet every day (91.2%) andshopped online frequently (83%). With regard to online apparelshopping, over 75% of the participants shopped online for apparelmore than once a week, and over half of these participants shoppedonline for apparel more than three times a week.

4.1. Preliminary analyses

Exploratory factor analyses (EFA) were performed using maxi-mum likelihood estimation with varimax rotation to test thedimensionality of CVPA. As a result, three factors were identified,congruent with original research (Bloch et al., 2003). Cronbach’salphas were also calculated to assess the reliability of all the depen-dent variables and indicate the internal consistency of the scales(.85 to .98) (see Table 1). The items within each scale were aver-aged and then used as the dependent variable.

In order to test the moderating effect of CVPA, a median-split(Med = 4.40) was used to divide the sample into high and low CPVAgroups after checking the distribution of CVPA scores; people witha high CVPA score are more concerned about visual aesthetics thanpeople with a low CVPA score. There was a significant difference inCVPA mean score between the high CVPA group (N = 327)(M = 5.42, SD = .66) and the low CVPA group (N = 331) (M = 3.32,SD = .89) (t = 34.33, df = 556, p < .001).

Manipulation checks were performed to determine whether ornot the manipulations of home page designs and brand familiaritywere effective. The t-test revealed a significant difference in con-sumers’ perception of the degree of image orientation betweenimage-oriented and text-oriented home pages (t = 56.73, df = 654,p < .001). A higher mean score indicates that participants perceivedthe home page design as being more image-oriented; the meanscore for the image-oriented home page was 5.91 (SD = 1.10) andfor the text-oriented home page 1.45 (SD = .92).

Manipulation of brand familiarity was also examined. Theresults of the t-test showed a significant difference in brand

Table 1Measurement items, convergent validity, and internal consistency reliability.

Variable Measurement item

Visualfluency

F1: This website home page is: not at all eye-catching – very eye cat

F2: Information on this website home page is: difficult to process –F3: The way this website displays its products is attractive

Preference P1: This is a good websiteP2: I like this website

Behavioralintentions

B1: How likely is it that you would revisit the website that you sawprobable not possible – possibleB2: How likely is that you would browse more products on the webshow likely is it that you would click to go to the product pages on thisprobable not possible – possible

Brandfamiliarity

I am familiar with this brand

I know a great deal about the brandI have a lot of previous experience with the brand

CVPA Owning products that have superior design makes me feel good aboI enjoy seeing displays of products that have superior designsA product’s design is a source of pleasure for meBeautiful product designs make our world a better place to liveBeing able to see subtle differences in product designs is one skill thI see things in a product’s design that other people tend to pass overI have the ability to imagine how a product will fit in with designs oI have a pretty good idea of what makes one product look better thaSometimes the way a product looks seems to reach out and grab meIf a product’s design really ‘‘speaks’’ to me, I feel that I must buy itWhen I see a product that has a really great design, I feel a strong ur

Note: All items were based on a 7-point scale, Critical ratios were all significant at p = .0

familiarity between the two brands, Ralph Lauren and agnés b(t = 27.01, df = 556, p < .001). Participants were familiar with RalphLauren as a brand (M = 4.14, SD = 1.57), but agnés b was unfamiliarto most participants (M = 1.40, SD = .97). Thus the manipulations ofboth home page design and brand familiarity were deemedsuccessful.

4.2. Measurement model specification

Based on a two-step modeling approach (Anderson & Gerbing,1988), a confirmatory factor analysis (CFA) was conducted viaAMOS 18.0. In order to estimate the reliability and validity of theitems for latent constructs, the measurement model was evaluatedand then adjusted based on theoretical and statistical consider-ations. Maximum likelihood estimation was employed to assessparameters for the CFA. The overall fit of the measurement modelwas evaluated and the results of the CFA showed a significant chi-square statistic (v2 = 38.59, df = 10, p = .000). Although the resultsof CFA showed the significant chi-square statistic due to the largesample size (N = 658), other fit indices indicated a satisfactory fit ofthe model to the data (RMSEA = .066 [90% C.I. = (.045; .089)],NIF = .98, CFI = .99, IFI = .99, GFI = .98, AGFI = .95 and TLI = .99).

In order to assess the construct validity, convergent validity anddiscriminant validity were examined. The results from the CFA ofthe measurement model revealed that all path coefficients inthe CFA model were greater than .8 and were statistically signifi-cant at a p value of .001 level (see Table 1), providing evidence ofconvergent validity. Discriminant validity was assessed usingchi-square difference tests between an unconstrained model andconstrained models (Anderson & Gerbing, 1988). Since all the chi-square difference tests were significant in this study (see Table 2),discriminant validity between the constructs was established.

4.3. Hypotheses Testing

Hypotheses 1 to 3 were tested using Analysis of Variance(ANOVA) (see Fig. 1). First, a 2 (home page design: image-oriented

Factorloading

CR AVE Cronbach’salpha

ching .90 29.74 .81 .85

easy to process .82 25.37.97 33.56.97 34.11 .96 .98.99 35.07

today? unlikely – likely improbable – .94 32.16 .92 .97

ite that you saw today? In other words,website? unlikely – likely improbable –

.98 34.42 .98

.92

ut myself .86–.91

at I have developed over time

f other things I already ownn its competitors

ge to buy it

01.

Fig. 2. Home page design: CVPA interaction effect on visual fluency.

Fig. 3. Standardized estimates and significance tests for H4 to H6 (�p < .05, ��p < .01,���p < .001).

J. Yoo, M. Kim / Computers in Human Behavior 38 (2014) 240–247 245

vs. text-oriented) � 2 (brand familiarity: familiar vs. unfamiliar)between-subject ANOVA was performed. There was a significantmain effect for home page design on visual fluency [F (1,654) = 1541.32, p < .001, partial g2 = .70]. Participants exposed toan image-oriented home page perceived greater visual fluency(M = 5.48, SD = .97) than those exposed to a text-oriented homepage (M = 2.38, SD = 1.04), supporting H1. However, the interactioneffect between home page design and brand familiarity was notsignificant. Thus H2, the moderating effect of brand familiarity,was not supported.

Hypothesis 3 was tested using 2 (home page design: image-ori-ented vs. text-oriented) � 2 (CVPA: high vs. low) between-subjectANOVA. The results showed a main effect for home page designon visual fluency [F (1,654) = 1565.86, p < .001, partial g2 = .71],consistent with the support for H1. The ANOVA results alsorevealed a significant interaction effect between home page designand CVPA [F (1,654) = 9.43, p < .01, partial g2 = .014]. Simple effectstests further showed that the effect of home page design on visualfluency was significant for the participants in both the high [F(1,654) = 906.19, p < .001, partial g2 = .58] and low CVPA groups[F (1,654) = 668.32, p < .001, partial g2 = .51]. Participants in bothCVPA groups experienced greater visual fluency when a home pagedesign was image-oriented (Mhigh = 5.64, SDhigh = 1.00; Mlow = 5.30,SDlow = .92) than when a home page design was text-oriented(Mhigh = 2.29, SDhigh = 1.04; Mlow = 2.44, SDlow = 1.04). However,this difference in visual fluency between an image-oriented andtext-oriented website was greater for the high CVPA group thanfor the low CVPA group. Therefore H3 predicting moderating effectof CVPA was supported (see Fig. 2).

Hypotheses 4 to 6 predicted the process by which visual fluencyas a result of a home page design influences behavioral intentionsvia preference (see Fig. 1). Single group structural equation model-ing was performed using a maximum likelihood function, revealingan overall chi-square of 38.59 (df = 10, p < .001), NFI = .99, CFI = .99,IFI = .99, GFI = .98, AGFI = .95 and TLI = .99. The RMSEA was .066.Even though the chi-square statistic was significant due to thelarge sample size, other fit indices also suggest an acceptable fitto the data.

As shown in Fig. 3, all the hypothesized relationships were sig-nificant. The visual fluency of a home page increased participants’preference for the website (c11 = .96, t = 40.53, p < .001) and theirbehavioral intentions (c21 = .36, t = 3.73, p < .001), thus supportingboth H4 and H5. Participants who experienced greater visualfluency on a home page liked the website more and had higherbehavioral intentions than those who experienced less visual flu-ency. The results also revealed a significant positive relationshipbetween preference for a website and behavioral intentions(b21 = .55, t = 5.67, p < .001), supporting H6. Participants who likedthe website more also had greater behavioral intentions.

4.4. Decomposition of effects

It is possible that visual fluency impacts behavioral intentionsboth directly and indirectly. Decomposition of the effects revealed

Table 2Chi-square difference tests and confidence interval for discriminant validity.

Constraint v2 d

Unconstrained model 38.59 1Visual fluency M preference 121.28 1Visual fluency M behavioral intentions 333.33 1Preference M behavioral intentions 475.45 1

Note: Dv2 = v2 (constrained) – v2 (unconstrained); Ddf = df (constrained) – df (unconstr*** p < .001.

direct and indirect effects from visual fluency to behavioral inten-tions. Visual fluency influenced behavioral intentions both directly(.96, p < .05) and indirectly through preference (.53, p < .01).

5. Discussion

The findings of this study provide useful theoretical and practi-cal implications that will help both researchers and retailersunderstand the factors that contribute to the effectiveness of ahome page design. Following the HOE model, this study examined

f Dv2 Ddf C.I.

01 82.69*** 1 .97; .971 294.74*** 1 .87; .911 436.86*** 1 .88; .91

ained).

246 J. Yoo, M. Kim / Computers in Human Behavior 38 (2014) 240–247

how a website home page influences consumers’ cognitive (visualfluency), affective (preference), and conative (behavioral inten-tions) responses. Our findings revealed that an image-orientedhome page is more visually fluent than a text-oriented home pageregardless of individual differences in terms of brand familiarityand the individual significance of aesthetics. Visually fluent homepages led to greater preference and, ultimately, high browsing andrevisit intentions. Consistent with previous findings in advertisingresearch, images play an important role in influencing the visual flu-ency of a website and, potentially, facilitate brand image building bycreating a positive impression of a website (Alba & Hutchinson,1987). The findings of this study provide empirical support for theHOE model in an online context and shed new light on the processby which consumers are influenced by a home page design.

The findings of this study suggest that online retailers shouldutilize more visual images on their home pages when relevantbased on the important role images play in influencing consumerbehaviors in an online context. Visual fluency as a result of homepage design impacts consumers’ intentions to both browse a web-site further and revisit. Considering the importance of attractingcustomers to shop at online stores, where competitors are only aclick away, increasing the visual fluency of a home page is clearlycritical in enticing customers to visit and spend more time on awebsite.

The findings of the current study provide further insights intohow individual characteristics influence the way a home pageimpacts consumer responses. Contrary to the prior research(Winkielman et al., 2003), the effect of home page design on con-sumer responses did not differ as a function of brand familiarity.In the current study, brand familiarity did not increase the visualfluency of a home page. This unexpected finding provides usefulinformation for less well-known brands seeking to expand theirbusiness online. Website design per se was more important thanbrand familiarity for consumer perceptions of visual fluency. Avisually appealing home page is therefore likely to be a particularlycost-effective advertising medium for less known brands trying toattract consumers to visit their websites. The findings of this studydo not suggest that website design is unimportant for familiarbrands, but visual fluency was clearly more important than brandfamiliarity. Thus, familiar brands also need to devote attention tocreating visually fluent home pages in today’s highly competitiveretail environment.

In their exploration of the concept of CVPA, Bloch et al. (2003)suggested that CVPA is an individual difference that will moderatethe relationship between stimuli and consumer responses such asvisual fluency. Consistent with Bloch et al., this study providesempirical evidence that compared to consumers with low CVPAscores, consumers with high CVPA scores experience greater visualfluency when looking at image-oriented home pages. Although thedegree of visual fluency perceived from the same image-orientedhome page differed significantly between the high and low CVPAgroups, both groups preferred the image-oriented home page tothe text-oriented home page. Therefore, even when individual dif-ferences are considered, the visual images on a home page areimportant for generating favorable consumer responses.

The results of this study support previous research on informa-tion presentation formats (Marks & Dulaney, 1998; Vora, 1998). AsMarks and Dulaney pointed out, processing verbal informationtakes more time than processing visual information. Since peopleselect information based on their eyes’ fixed moments, theresearchers suggested the utility of colorful visual displays on awebsite to facilitate effective information processing by consum-ers. The findings of this study provide empirical support for thiscontention: an image-oriented home page design was consideredmuch easier to process than a text-oriented home page design,with a consequent impact on visual fluency.

The findings of this study corroborate the important role of awebsite home page in not only attracting consumers to a website,but also building brand images through forming positive impres-sions (Lindgaard et al., 2006; Schenkman & Jönsson, 2000), extend-ing impression formation theory to an online context. Onlineretailers need to approach the development of their website’shome page as part of their Integrated Marketing Communication(IMC) if they are to build a consistent brand image and go beyondsimply informing consumers about current promotions and newproducts. Considering how consumers use a home page whenshopping online, online retailers are advised to develop visuallyattractive websites that are easy to process with minimal text. Ahome page provides a unique opportunity to draw shoppers intoa store and at the same build a brand image that suits thecompany’s needs.

6. Limitations and future research

This study entails several limitations. The first was the use ofcollege women. While this group comprises a major portion ofonline shoppers, they may not adequately represent all such shop-pers. Thus, future research should expand the participant pool toinclude more diverse groups of Internet shoppers in order toenhance the generalizability of the findings.

The types of home page content considered in this study werelimited to images and text. In order to investigate the impact ofadditional elements of home page design on consumer responses,other aspects of home page design should be considered. For exam-ple, the use of flash designs, video clips, and/or promotion informa-tion, as well as factors such as downloading time, may also affectconsumers’ first impressions of a website and thus influence theirintention to continue on to browse product pages on the website.Therefore, future research into the effects of manipulating otherdesign aspects on a home page could provide very useful additionalinformation.

References

Alba, J. W., & Hutchinson, W. (1987). Dimensions of consumer expertise. Journal ofConsumer Research, 13, 411–454.

Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: Areview and recommended two-step approach. Psychological Bulletin, 103,411–423.

Aquirre-Rodriguez, A., Bosnjak, M., & Sirgy, J. (2012). Moderators of the self-congruity effect on consumer decision-making: A meta-analysis. Journal ofBusiness Research, 65, 1179–1188.

Asch, S. E. (1946). Forming impressions of personality. The Journal of Abnormal andSocial Psychology, 41(3), 258–290.

Barry, T. E., & Howard, D. J. (1990). A review and critique of the hierarchy of effectsin advertising. International Journal of Advertising, 9(2), 121–131.

Blanco, C., Sarasa, R., & Sanclemente, C. (2010). Effects of visual and textualinformation in online product presentations: looking for the best combinationin website design. European Journal of Information Systems, 19, 668–686.

Bloch, P. H., Brunel, F. F., & Arnold, T. J. (2003). Individual differences in thecentrality of visual product aesthetics: Concept and measurement. Journal ofConsumer Research, 29(4), 551–565.

Campbell, A., & Pisterman, S. (1996). A fitting approach to interactive service design:The importance of emotional needs. Design Management Journal, Fall, 10–14.

Cobb-Walgren, C. J., Ruble, C. A., & Donthu, N. (1995). Brand equality, brandpreference, and purchase intent. Journal of Advertising, 24(3), 25–40.

Costley, C. L., & Brucks, M. (1992). Selective recall and information use in consumerpreference. Journal of Consumer Research, 18, 464–474.

Denis, C., & Fenech, T. (2004). E-retailing. New York: Routledge.Evans, D. S. (2008). The economics of the online advertising industry. Review of

Network Economics, 7(3), 1–49.Go-Gulf (2012). Global online advertising spending statistics. <http://www.go-

gulf.com/blog/online-ad-spending/> Retrieved 12.10.13.Hartmann, J., Sutcliffe, A., & De Angeli, A. (2007). Investigating attractiveness in web

user interfaces. In Proceedings of SIGCHI conference on human factors incomputing systems, ACM (pp. 387–396).

Hirschman, E. C. (1986). The effect of verbal and pictorial advertising stimuli onaesthetic, utilitarian, and familiarity perceptions. Journal of Advertising, 15(2),27–34.

J. Yoo, M. Kim / Computers in Human Behavior 38 (2014) 240–247 247

Hodes, C. L. (1994). Processing visual information: implications of the dual codetheory. Journal of Instructional Psychology, 21, 36–43.

Holbrook, M. B., & Batra, R. (1987). Assessing the role of emotions as mediators ofconsumer responses to advertising. Journal of Consumer Research, 14, 404–420.

Im, H., Lennon, S. J., & Stoel, L. (2010). The perceptual fluency effect on pleasurableonline shopping experience. Journal of Research in Interactive Marketing, 4(4),280–295.

Jacoby, L. L. (1983). Perceptual enhancement: Persistent effects of an experience.Journal of Experimental Psychology, Learning, Memory, and Cognition, 9(1), 21–38.

Jacoby, L. L., Kelley, C. M., & Dywan, J. (1989). Memory attributions. In H. Roediger &C. Ferqus (Eds.), Varieties of memory and consciousness: Essays in honor of endeltulving (pp. 391–422). Hillsdale, NJ: Erlbaum.

Janiszewski, C. (1993). Preattentive mere exposure effects. Journal of ConsumerResearch, 20, 376–392.

Kent, R., & Allen, C. (1994). Competitive interference effects in consumer memoryfor advertising: The role of brand familiarity. Journal of Marketing, 58(3),97–105.

Kim, D. J., Ferrin, D. L., & Rao, H. R. (2009). Trust and satisfaction, two steppingstones for successful e-commerce relationships: A longitudinal exploration.Information Systems Research, 20, 237–257.

Kim, M., & Lennon, S. J. (2008). The effects of visual and verbal information onattitudes and purchase intentions in Internet shopping. Psychology andMarketing, 25(2), 146–178.

Kim, J., & Moon, J. Y. (1998). Designing towards emotional usability in customerinterfaces – Trustworthiness of cyber-banking system interfaces. Interactingwith Computers, 10(1), 1–29.

Krugman, H. E. (1965). The impact of television advertising: Learning withoutinvolvement. Public Opinion Quarterly, 29(3), 349–356.

Labroo, A. A., Dhar, R., & Schwarz, N. (2008). Of frog wines and frowning watches:Semantic priming, perceptual fluency, and brand evaluation. Journal ofConsumer Research, 34, 819–831.

Lavidge, R. J., & Steiner, G. A. (1961). A model for predictive measurements ofadvertising effectiveness. Journal of Marketing, 25, 59–62.

Lefkoff-Hagius, R., & Mason, C. H. (1993). Characteristic, beneficial, and imageattributes in consumer judgments of similarity and preference. Journal ofConsumer Research, 20, 100–110.

Levine, R. (1995). Guide to web style. <http://www.sun.com/styleguide/>.Lindgaard, G., & Dudek, C. (2003). What is this evasive beast we call user

satisfaction? Interacting with Computers, 15, 429–453.Lindgaard, G., Fernandes, G. J., Dudek, C., & Brown, J. (2006). Attention website

designers: You have 50 milliseconds to make a good first impression! Behaviorand Information Technology, 25, 115–126.

Marks, W., & Dulaney, C. L. (1998). Visual information processing on the world wideweb. In C. Forsythe, E. Grose, & J. Ratner (Eds.), Human factors and webdevelopment (pp. 25–44). Mahwah, NJ: Lawrence Erlbaum Associates.

Masson, M., & Caldwell, J. I. (1998). Conceptually driven encoding episodes createperceptual misattributions. Acta Psychologica, 98(2–3), 183–210.

Morris, J. D., Woo, C., Geason, J. A., & Kim, J. (2002). The power of affect: Predictingintention. Journal of Advertising Research, 42, 7–17.

Nielsen (2011). In U.S. men are shopping more than ever, while women arewatching more TV. <http://www.nielsen.com/us/en/newswire/2011/in-u-s-men-are-shopping-more-than-ever-while-women-are-watching-more-tv.html>Retrieved 01.05.14.

Nisbett, R. E., & Ross, L. (1980). Human inference: Strategies and shortcomings of socialjudgment. Englewood Cliffs, NJ: Prentice Hall.

O’Keefe, D. (2002). Persuasion: Theory and research. Thousand Oaks, CA: SagePublications.

Paivio, A. (2007). Mind and its evolution: A dual coding theoretical approach. Mahwah,NJ: Lawrence Erlbaum Associates.

Park, S., Choi, D., & Kim, J. (2005). Visualizing e-brand personality: Exploratorystudies on visual attributes and e-brand personalities in Korea. InternationalJournal of Human-Computer Interaction, 19(1), 7–34.

Pavlou, P. A., & Fygenson, M. (2006). Understanding and predicting electroniccommerce adoption: An extension of the theory of planned behavior. MISQuarterly, 20, 115–143.

Reber, R., Winkielman, P., & Schwarz, N. (1998). Effects of perceptual fluency onaffective judgments. Psychological Science, 29(1), 45–48.

Retailer, Internet. (2011). Top 500 guide. Chicago, IL: Vertical Web Media LLC.Rossiter, J. R., & Percy, L. (1978). Visual imagery ability as a mediator of advertising

response. Advances in Consumer Research, 5(1), 621–629.Schenkman, B. N., & Jönsson, F. U. (2000). Aesthetics and preferences of web pages.

Behaviour and Information Technology, 19(5), 367–377.Sen, S., Block, L. G., & Chandran, S. (2002). Window displays and consumer shopping

decisions. Journal of Retailing and Consumer Service, 9, 277–290.Singh, S. N., & Dalal, N. P. (1999). Web home pages as advertisements.

Communication of the ACM, 42(8), 91–98.Suh, J.-C., & Yi, Y. (2006). When brand attitudes affect the customer satisfaction-

loyalty relation: The moderating role of product involvement. Journal ofConsumer Psychology, 16(2), 145–155.

Tractinsky, N., & Lowengart, O. (2007). Web-store aesthetics in e-retailing: Aconceptual framework and some theoretical implication. Academy of MarketingScience Review, 11(1), 1–18.

Vora, P. R. (1998). Designing for the web: A survey. Interactions, 5(3), 13–30.Whittlesea, B. (1993). Illusions of familiarity. Journal of Experimental Psychology:

Learning, Memory, and Cognition, 19, 1235–1253.Wiebe, E. N., & Howe, J. E. (1998). Graphics design on the web. In C. Forsythe, E.

Grose, & J. Ratner (Eds.), Human factors and web development (pp. 225–239).Mahwah, NJ: Lawrence Erlbaum Associates.

Winkielman, P., Schwarz, N., Reber, R., & Fazendeiro, T. A. (2003). Cognitive andaffective consequences of visual fluency: When seeing is easy on the mind. In L.Scott & R. Batra (Eds.), Persuasive imagery: A consumer response perspective(pp. 75–89). NJ: Lawrence Erlbaum Associates.

Yoo, J., & Kim, M. (2012). Online product presentation: The effect of productcoordination and a model’s face. Journal of Research in Interactive Marketing,6(1), 59–72.

Zajonc, R. B., & Markus, H. (1982). Affective and cognitive factors in preferences.Journal of Consumer Research, 9, 123–131.