Analyzing the User Behavior toward Electronic Commerce … · · 2017-04-13S-O-R model) refer to...

fnbeh-10-00224 November 28, 2016 Time: 12:5 # 1

ORIGINAL RESEARCHpublished: 30 November 2016

doi: 10.3389/fnbeh.2016.00224

Edited by:Eduardo Fernandez,

Universidad Miguel Hernándezde Elche, Spain

Reviewed by:Eleni Konsolaki,

American College of Greece, GreeceMiguel Angel Lopez,

University of Granada, Spain

*Correspondence:Carlota [email protected]

Received: 22 July 2016Accepted: 07 November 2016Published: 30 November 2016

Citation:Lorenzo-Romero C, Alarcón-del-AmoM-d-C and Gómez-Borja M-Á (2016)

Analyzing the User Behavior towardElectronic Commerce Stimuli.

Front. Behav. Neurosci. 10:224.doi: 10.3389/fnbeh.2016.00224

Analyzing the User Behavior towardElectronic Commerce StimuliCarlota Lorenzo-Romero1*, María-del-Carmen Alarcón-del-Amo2 andMiguel-Ángel Gómez-Borja1

1 Business Department, University of Castilla-La Mancha, Albacete, Spain, 2 Business Department, Universidad de Alcalá,Alcalá de Henares, Spain

Based on the Stimulus-Organism-Response paradigm this research analyzes themain differences between the effects of two types of web technologies: Verbalweb technology (i.e., navigational structure as utilitarian stimulus) versus non-verbalweb technology (music and presentation of products as hedonic stimuli). Specificwebmosphere stimuli have not been examined yet as separate variables and theirimpact on internal and behavioral responses seems unknown. Therefore, the objectiveof this research consists in analyzing the impact of these web technologies –whichconstitute the web atmosphere or webmosphere of a website– on shopping humanbehavior (i.e., users’ internal states -affective, cognitive, and satisfaction- and behavioralresponses – approach responses, and real shopping outcomes-) within the retailonline store created by computer, taking into account some mediator variables (i.e.,involvement, atmospheric responsiveness, and perceived risk). A 2 (“free” versus“hierarchical” navigational structure) × 2 (“on” versus “off” music) × 2 (“moving”versus “static” images) between-subjects computer experimental design is used to testempirically this research. In addition, an integrated methodology was developed allowingthe simulation, tracking and recording of virtual user behavior within an online shoppingenvironment. As main conclusion, this study suggests that the positive responsesof online consumers might increase when they are allowed to freely navigate theonline stores and their experience is enriched by animate gifts and music background.The effect caused by mediator variables modifies relatively the final shopping humanbehavior.

Keywords: verbal (utilitarian) and non-verbal (hedonic) computer stimuli, webmosphere tools, Stimulus-Organism-Response model, shopping human behavior, computer experimental design

INTRODUCTION

Kotler (1973) defines atmospherics as “the conscious designing of space to create certain buyereffects, specifically, the designing of buying environments to produce specific emotional effects inthe buyer that enhance purchase probability.” Given the extended increasing of online stores, someauthors are focusing their research on studying an “extended” term: Web atmospheric, also calledwebmosphere by Childers et al. (2001), or virtual store atmosphere by Vrechopoulos et al. (2000),or online store environment by Manganari et al. (2009). This concept is defined by Dailey (2004)as “the conscious designing of web environments to create positive effects (e.g., positive affect,positive cognitions, etc.) in users in order to obtain more positive responses (e.g., site revisiting or

Frontiers in Behavioral Neuroscience | www.frontiersin.org 1 November 2016 | Volume 10 | Article 224

http://www.frontiersin.org/Behavioral_Neuroscience/

http://www.frontiersin.org/Behavioral_Neuroscience/editorialboard

http://www.frontiersin.org/Behavioral_Neuroscience/editorialboard

https://doi.org/10.3389/fnbeh.2016.00224

http://creativecommons.org/licenses/by/4.0/

https://doi.org/10.3389/fnbeh.2016.00224

http://crossmark.crossref.org/dialog/?doi=10.3389/fnbeh.2016.00224&domain=pdf&date_stamp=2016-11-30

http://journal.frontiersin.org/article/10.3389/fnbeh.2016.00224/abstract

http://loop.frontiersin.org/people/226594/overview

http://loop.frontiersin.org/people/232452/overview


http://www.frontiersin.org/

http://www.frontiersin.org/Behavioral_Neuroscience/archive


Lorenzo-Romero et al. An Empirical Analysis

browsing). When marketers design web interfaces in order toentice consumers, they are utilizing web atmospherics.”

Childers et al. (2001) explain that the use the use of video,images, humor, color, music, and other interactive aspects candefine and hedonic experience and help to create a moreenjoyable environment. On the contrary, a technology orientedperspective that focus on shopping as a cold information systems,with no enjoyable environments, is probable to be decreased,specially for products with hedonic attributes (Hirschman andHolbrook, 1982; Childers et al., 2001; Solaymani et al., 2012).

Most authors use the Stimulus-Organism-Response (S-O-R)paradigm proposed by Mehrabian and Russell (1974) to explainthe influence of web atmosphere on consumers. These authorssuggest that external stimulus (like web atmospheric cues) affectconsumers’ internal states and, in turn, they have an effecton behavioral responses. In the online shopping environment,some researchers use actual stimuli (Wang et al., 2010; Animeshet al., 2011; Kim and Lennon, 2013), and others use customerassessments of the stimuli to denote the stimulus segment of themodel (Nath, 2009; Koo and Ju, 2010; Manganari et al., 2011).Eroglu et al. (2001, 2003) proposed two types of webmosphericcues which named high and low task relevant cues. High task-relevant cues comprise verbal or pictorial contents directlyassociated with the shopping goal. The purpose of these verbalor pictorial descriptions (e.g., product information, price, anddelivery and return policies) is to assist online consumers to reachtheir shopping goals. Low task relevant cues, on the other hand,are peripheral contents (e.g., color, background patterns, andimages) not directly related to the shopping goals. Moreover, theorganism (i.e., user’s internal states of S-O-R model) is composedby affective, cognitive and satisfaction variables (Oliver, 1997;Bigné and Andreu, 2004). In addition, Eroglu et al. (2001, 2003)suggest that the internal states of the organism include affectand cognition and that satisfaction is one of the outcomes ofbehavior. And, finally, the response (i.e., behavioral responses ofS-O-R model) refer to website patronage intention (Jeong et al.,2009), purchase intention (Hsu et al., 2011; Liu et al., 2016), andintention to use and buy (Huang, 2013).

Many studies do not examine specific webmosphere stimuliseparately and their impact on internal and behavioral responses.Normally, these studies analyze global stimuli such as highand low task relevant cues (Eroglu et al., 2001, 2003) andtheir impact on consumer. Other authors use only one webstimulus like navigation structure (Dailey, 2004) including someaspects referred to web navigation. Other studies analyze specificstimuli attending to other online marketing variables such asprice, service quality, social variables, etc. (Chang et al., 2011;Nesset et al., 2011), not focusing the research toward webdesign exclusively. Moreover, some previous research analyzesbehavioral responses such as purchase intention against realshopping (although through simulation). Previous research doesnot examine the relationship between three mediator variables.Some authors analyze perceived risk or trust (Kim, 2012),or product involvement (Keng et al., 2012), or atmosphericresponsiveness (Eroglu et al., 2001). Nevertheless, any studyanalyzes the possible relationship and impact of three constructson internal and behavioral online buyer responses.

Thus, the main objective of this study is to establishthe main different effects of two webmosphere verbalcommunication stimulus (i.e., navigation structure) andnon-verbal communication stimuli (i.e., music and presentationof products), on internal (i.e., satisfaction, cognitive andaffective) and behavioral consumer responses (i.e., approachresponses and real shopping outcomes) within a virtual apparelshopping environment. Moreover, three mediator variables (i.e.,involvement, atmospheric responsiveness, and perceived risk)are considered in the study in order to analyze the impact ofthem on the relationships between webmospheric cues and userresponses.

Specifically, affective state refers to emotions that user feelstoward different stimuli (Eroglu et al., 2001, 2003). Cognitivestate is the thought processes and state of mind related to theacquisition, processing, retention, and retrieval of information(Eroglu et al., 2001). Satisfaction is understood as a globalevaluation or attitude that evolves over time resulting from arelationship between store and consumer (Flavián et al., 2004).Approach behaviors refer to all positive actions that might bedevelop toward a setting, for example, intentions to exploreor stay, while avoidance refers to the opposite (Mehrabianand Russell, 1974; Bitner, 1992; Gregg and Walczak, 2010).Real shopping outcomes are considered as behavioral contumerresponses. This construction refers to quantitative measures suchas time or money spent, products bought, etc.

Regarding mediator variables, involvement refers to thesignificant consciousness produced from product characteristics(Zaichkowsky, 1986). Atmospheric responsiveness is anenvironmental characteristic that influences users’ decisions onwhere and how to shop as well as the outcomes of the shoppingexperience (Eroglu et al., 2001, 2003). And, finally, perceived riskis the consumer’s level of uncertainty regarding the outcome of apurchase decision.

Respect to navigation structure, we propose compare twoweb manipulatios: “Free navigation structure” of the web sitemeans that consumer can move easily through the web site (i.e.,navigational bars and the same links to access the informationin all sites, location link, and searching tool) and, consequently,the user is not controlled by web marketer with restrictivenavigational bars. In contrast, in a web site with “hierarchicalnavigation structure,” the user only can use the “next” and“previous” navigational bars to move through the web site(i.e., user is exposed to restrictive navigational bars and, inconsequence, its navigation is more difficult). Regarding musicstimulus, we propose compare two web manipulations: Website“with music” versus website “non-music.” Finally, respect to“presentation of products” stimulus, we propose compare twoweb manipulations: “Moving images” (movement of photos in360◦, video, etc.) versus “static images.”

“One important feature of the new media that differs fromtraditional shopping channels is the absence of the experienceabout the online store visit and the unfeasibility of examininga product prior to purchase” (Lorenzo et al., 2007). Alba et al.(1997) indicate that a competitive advantage for marketers canbe lead through an effective design of web site. But, how canmarketers design web interfaces with high level of effectivity?







Web atmospherics may offer insight into this question (Dailey,2004).

Hoffman and Novak (1996) define network navigation as“the process of self-directed movement through the mediainvolving non-linear search and retrieval methods that permitgreater freedom of choice.” In traditional retail environments,consumers identify the spatial representations of the store’sdesign and recognize how products are grouped by theircommon characteristics or through orientation aids (i.e., displays,directory maps, store personnel, aisle markers, and so on) whenthey look for products (Titus and Everett, 1995).

Within the context of online behavior, we can introduce theusability concept. According to Nielsen (2000), the usability refersto the facility for the online navigation, and users are satisfiedwhen it is easy to use by them. Therefore, usability is a qualityattribute, and a key attribute to reach consumer satisfaction(Ranganathan and Ganapathy, 2002).

Dailey (2004) analyzes the impact of restrictive navigationcues as an explicit online webmosphere variable. She based onthe theory of flow experience and the psychological reactance toexplain that the restrictive navigation cues act as barriers thatmake users to be threated about the control over web navigation.This perceived barrier causes negative attitudes toward the website and an avoidance behavior, which has negative consequencesfor the web marketer.

As other authors affirm, an easy navigational structure, suchas a high task relevant (Eroglu et al., 2003), ease of use andusefulness (Childers et al., 2001), has a positive effect on theinternal and behavioral consumer responses. Therefore, wepropose the following hypothesis:

H1−1: An online shopping web with “free navigation structure”will affect in a more positive affective responses of users than“hierarchical navigation structure.”

The cognitive state in an online shopping refers issues relatedwith how online consumers understand information provided onthe screen to choose from different sites and products and theattitude toward the online store (Eroglu et al., 2003). Therefore,we considered the cognitive state as the consumers’ attitude andthe knowledge attained during the shopping experience. So, wepropose the following hypothesis:

H1−2: An online shopping web with “free navigation structure”will affect in a more favorable cognitive responses of users than“hierarchical navigation structure.”

Taking into account the double perspective of satisfactionmeaning commented in Section “Sample and Procedure:Computer Experimental Design,” our research is focused on theattitudinal satisfaction perspective because it is more relatedto the purchase intention (Shankar et al., 2003). Specifically,satisfaction is understood as a global evaluation or attitude asa result of relationships between store and consumer (Flaviánet al., 2004). For this reason, we will consider the satisfaction asan internal state which is affected by usability. So, the followinghypothesis is proposed:

H1−3: An online shopping web with “free navigation structure”will influence on higher levels of satisfaction of users than“hierarchical navigation structure.”

Donovan et al. (1994) and Sherman et al. (1997) foundthat, in traditional contexts, the shoppers’ environmentalperceptions affected their approach behaviors in the form oftime and money spent, returning, store exploration, and so on.Lorenzo et al. (2007) observe that “within the online shoppingenvironment, some works obtains similar approach/avoidancebehavior depending on the perceived “store” environment andthe mediating effects of individual traits and internal states(Eroglu et al., 2003).” In online environments, these effects havebeen less studied but literature indicates online atmospheric cuesaffect approach/avoidance responses. However, the relationshipbetween both dimensions (i.e., stimuli and behavioral responses)is mediated by internal states for two cases: (a) For onlineenvironments with only utilitarian elements (Dailey, 2004);(b) and for online environments with utilitarian and hedonicelements (Childers et al., 2001; Eroglu et al., 2001, 2003). Havinginto consideration these above literature, we suggest the followinghypothesis:

H1−4: An online shopping web with “free navigation structure”will affect in a more approach responses of users than“hierarchical navigation structure.”

Flavián et al. (2005) demonstrated a positive influence ofweb site usability on real shopping outcomes toward web site(mediated by confidence and satisfaction). In consequence, thefollowing hypothesis is proposed:

H1−5: An online shopping web with “free navigation structure”will influence on more positive real shopping outcomes than“hierarchical navigation structure.”

On the other hand, music in the website is other non-verbalweb technology stimulus. McMurray et al. (2008) argue that intraditional environments musical expectancy can influence onlower-level perceptual processes. Lorenzo et al. (2007) indicatethat “creating a more enjoyable environment may require the useof more powerful web languages, and the inclusion of images,video, color, humor, sound, music, games, animation, and allof the other interactive aspects that could define an enjoyableexperience. A technology oriented perspective that attempts totreat media shopping as cold information systems, rather thanimmersive, hedonic environments, is likely to be misguided,mainly for products with strong hedonic attributes, as can be thecase of apparel (Childers et al., 2001).”

Eroglu et al. (2001, 2003) explain that sounds or music arelow task-relevant cues, since they do not affect in a direct waythe realization of the task. Nevertheless, music or sounds canhelp to create an atmosphere that has the possibility of makingthe shopping experience more pleasurable. In consequence, wepropose the following hypothesis:

H2−1: An online shopping web with music will influenceon more positive affective responses of users than a webenvironment without music.







Low task-relevant cues affect positively users’ cognitive statesin online environments (Eroglu et al., 2001). In the literaturethere are several low-task relevant cues, but we will focus in musicand its effect on users’ cognitive states (learning/knowledge andattitudinal process). Therefore, related with cognitive responses,the following hypothesis is proposed:

H2−2: An online shopping web with music will influenceon more favorable cognitive responses of users than a webenvironment without music.

Customer’s satisfaction is defined as “a relative psychologicalstate which is a result of purchase/consumption experience”(Vanhamme, 2000). Most of research demonstrate that storeatmosphere influence on satisfaction and, in turn, behavioralresponses (Bigné and Andreu, 2004) and, specifically, withhedonic webmosphere stimuli (Childers et al., 2001). So, wesuggest the following hypothesis:

H2−3: An online shopping web with music will influence onmore satisfaction of users than a web site without music.

In some studies where consumers were exposed to differenttempo of background music within supermarkets, they obtainedsimilar responses (Milliman, 1986; Oakes, 2003). Eroglu et al.(2001, 2003) demonstrated that consumers behave in differentways (approach/avoidance behaviors) depending on the onlineperceived store environment and the mediating effects ofconsumer internal states. They analyzed whether the online storeinformation and the low-task relevant cues facilitate or impedethe attainment of shopping goals and, in turn, whether theonline shopper exhibited positive or negative behaviors towardthe particular web site. Finally, they obtained that a rise ofsome atmospheric cue (high or low cue) increased the approachresponse of consumer. Because our intention is focused on studyof specific atmospheric cue (i.e., music), we propose the followinghypothesis:

H2−4: An online shopping web with music will affect in moreapproach responses of users than a web site without music.

Finally, regarding behavioral responses we proposed thefollowing hypothesis, attending two groups of variables analyzed(loyalty and approach/avoidance behavioral). Specifically, asregards users’ loyalty toward online store after their visit inthe web site is measured by Eroglu et al. (2003) as satisfactionmeasurement. However, according to above works, loyalty isconsidered as the consequence of the satisfaction (Zeithaml et al.,1996; Bigné and Andreu, 2004; Flavián et al., 2004) and loyaltymeans recommended the online store, better results, etc. Inspite of this conceptual difference, the most works posit thatatmospheric cues (specifically, low-task relevant cues accordingto Eroglu et al. (2001, 2003) affect positively real shoppingoutcomes toward store, although this relationship is mediated byconsumers’ internal states. Taking everything into account, wesuggest the following hypothesis:

H2−5: An online shopping web with music will affect in morepositive real shopping outcomes than a web site without music.

Other non-verbal web technology stimulus, specificallyhedonic dimensions, can be the presentation of product.According to Lorenzo et al. (2007), there are three kindsof animations depending on the technical characteristics thatthey use: animated gifs, flash format and video. In addition,Lorenzo et al. (2007) indicate that “the use of images and theiranimations as design elements within stores offers advantages anddisadvantages depending on their use [. . .]. Animating imageshas a potential effect on human’s periphery vision, specifically;it is very difficult to concentrate on reading a text if there is arevolving logo on the web page. In this case, it is more convenientto reduce the use of animations. However, if animation is used tomake the purchase task easier, the use of animation will be moreconvenient.”

Schlosser (2003) examines how users process the images andtexts offered by virtual media and state that “the image processingrepresents with a more relevant magnitude the object’s interactiveeffects on user intention –purchasing or browsing– than theprocessing of a verbal or in written speech. In fact, internalstates as satisfaction variable is affected positively by animationelements on the website” (Childers et al., 2001; Eroglu et al., 2001,2003; Adelaar et al., 2003; Wang et al., 2010).

The presentation of virtual products with animation designinfluence positively on consumer’s internal states and shoppingoutcomes (Dahan and Srinivasan, 2000; Eroglu et al., 2003;Koernig, 2003; Yoh et al., 2003; Hong et al., 2004; McKinney,2004; Choura and Saber, 2005; Wang et al., 2010; Lian, 2011).In addition, Adelaar et al. (2003) demonstrated that verbal andvisual presence –through the use of image and/or video– improveusers’ emotional states. Therefore, we propose the followinghypothesis:

H3−1: An online shopping web with “moving images” willaffect in more positive affective responses of users than a webenvironment with “static images.”

In a similar vein, the introduction of “low task relevantcues” within the website, such as animation elements, influencefavorable on consumer internal states (Dahan and Srinivasan,2000; Childers et al., 2001; Eroglu et al., 2003; Hong et al., 2004).Moreover, Nielsen (2000) states that animations are efficaciouswhen their transitions allow users to control them by perceptivesystem. So, the hypothesis proposed is:

H3−2: An online shopping web with “moving images” willinfluence on a more favorable cognitive responses of users thana web environment with “static images.”

The two-dimensional structure of a computer screen makes itdifficult to understand a three-dimensional structure with onlyone image, regardless of its quality. So, animation is used to makethe understanding of product’s spatial visualization which theusers observe within the store easier. In this line, Hong et al.(2004) focus their research on the analysis of the use of Flash R©

format on e-consumer behavior. As a result, the animation canincrease the search for information if the animated element isrelevant to the search task. Furthermore, Eroglu et al. (2003)state that the inclusion of this web attribute within the store







facilitates the users’ information processing during the onlinepurchase which, in turn, improves their satisfaction (McKinney,2004). A lot of apparel websites lacks of movement of productsand, even, static images, using in consequence only somedescriptions about products. So, in order to analyze this type ofweb atmosphere cue, this research is focused in the comparisonof static images versus a combination of eight pictures which offeran optical illusion of movement as well as the use of streaming.

The literature offers some works related to the relationshipbetween web atmosphere and satisfaction. For instance, Childerset al. (2001) compared two types of web environments: oneof them included only utilitarian elements (i.e., order andreception system) and the other one included also somehedonic characteristics (i.e., images of products with highresolution, background music, and interactive plays). Grobelnyand Michalski (2015) also use an experimental study to analyzethe influence of hedonic stimuli (color) on user men and womensatisfaction. From Technology Acceptance Model, these authorsobtained that both web dimensions have an equal role in thedevelopment of consumer’s attitudes and behavioral. In thissense, studies such as Eroglu et al. (2001, 2003) and Adelaar et al.(2003) also used different web experiments in their researches.From S-O-R Model, these authors obtained similar results (i.e.,web elements such as images, and video, affect positively oninternal and behavioral responses). In contrast, Kim and Stoel(2004) used the Loiaconno’s WebQualTM instrument to measurethe effects of website quality dimensions (i.e., entertainment, webappearance, transaction capability, response time, informationalfit-to-task, and trust) on shopper satisfaction. These authorsobtained that only three dimensions (specifically, responsetime, transaction capability and informational fit-to-task) wereimportant predictors of consumer satisfaction. To sum up,according to literature, animation elements of the web site affectpositively on internal states, such as satisfaction (Childers et al.,2001; Eroglu et al., 2001, 2003; Adelaar et al., 2003; Hong et al.,2004). So, we proposed the following hypothesis:

H3−3: An online shopping web with “moving images” willinfluence on higher levels of satisfaction of users than a webenvironment with “static images.”

Schlosser (2003) examines how the users process the textsand images offered by online media, and demonstrate that “theimage processing represents with a more relevant magnitudethe object’s interactive effects on user intention –purchasingor browsing– than the processing of a verbal or in writtenspeech. So, the product’s interactivity causes intense mentalimages which will influence more strongly on user’s intentionthan in their cognitive processing” (Lorenzo et al., 2007). Inthe same line, Childers et al. (2001) state that the functionalaspects such as usability, navigational structure, etc., cause stronginfluences on the e-consumer responses. Moreover, the restof web elements not functional such as music, color, imageswith high resolution, etc., are equally relevant elements whichcontribute to the shaping of consumer attitudes and behavior(i.e., opinions about the store and unexpected purchase behavior).Based on previous literature, in webs which includes animations,

costumer’s show more positive internal states, which improvetheir shopping experience and, in consequence, their approachresponses. Having into consideration the above literature, wepropose the following hypothesis (Lorenzo et al., 2007):

H3−4: An online shopping web with “moving images” willaffect in more approach responses of users than a webenvironment with “static images.”

Eroglu et al. (2003) argue that the inclusion of entertainmentwithin the store raises the consumer’s positive behavior towardthe website. On the other hand, Koernig (2003) affirms that thehigher the product’s tangibility level represented through thescreen, the more positive behavioral response. Khakimdjanovaand Park (2005) affirm that the use of online visual displayfeatures that allow viewing the garment from different angles canhelp potential online consumers to make a purchase decision.

In conclusion, according to literature, the use of animationdesigns for presenting products influence positively onconsumer’s internal states and shopping real outcomes (Dahanand Srinivasan, 2000; Adelaar et al., 2003; Eroglu et al., 2003;Koernig, 2003; Yoh et al., 2003; Hong et al., 2004; McKinney,2004; Choura and Saber, 2005). In addition, Argyriou (2012)demonstrated that website characteristics do not affect revisitintentions directly but through the vividness of mental imagesthat consumers hold of the website as a whole. Vivid mentalwebsite imagery is stimulated by animation and facilitated byindividual tendencies to put faith in intuitive rather than rationalthinking. Then, it is proposed that (Lorenzo et al., 2007):

H3−5: An online shopping web with “moving images” willaffect in more positive real shopping outcomes than a webenvironment with “static images.”

Regarding mediator variables, Baron and Kenny (1986)analyze the differences between the properties’ mediator andmoderator variables. In general, mediator variables explainhow external physical events take on internal psychologicalsignificance. Whereas moderator variables specify when certaineffects will hold, mediators speak about how or why sucheffects occur (Lorenzo et al., 2007). Taking into account thisdistinction, in this research were studied three variables (i.e.,involvement, atmospheric responsiveness, and perceived risk) aspossible mediator between constructions analyzed.

Based on Eroglu et al.’s (2003) work, the relationshipbetween virtual atmosphere and internal states is mediated byuser personal characteristics (e.g., involvement and atmosphericresponsiveness).

Zaichkowsky (1986) defines product involvement as therelevant consciousness caused by product characteristics.Moreover, according to Goldsmith and Emmert (1991), thehigher the level of product involvement, the higher the searchfor product information. Therefore, product involvement is animportant variable in consumer behavior.

Then, variables such as involvement with the product (Kenget al., 2012), and specifically with virtual purchase (Hernándezet al., 2010), experience with the new media and advertisement(Fortin and Dholakia, 1999), should affect electronic shopping







(Yoh et al., 2003). In fact, atmospheric responsiveness isan environmental characteristic that influences consumers’decisions -where and how to shop- as well as the shoppingexperience (Eroglu et al., 2001, 2003). Lorenzo et al. (2007), basedon Huang’s (2006) research explain that “flow and involvementare known as motivational constructs (Csikszentmihalyi, 1975).The concept behind involvement lies in personal relevance,regardless of whether the locus of personal relevance residesin the consumer or the situation (Celsi and Olson, 1988).Enduring and situational involvement differ in aspects such as thetemporal pattern of occurrences, the motivations and the benefitssought (Huang, 2006). Enduring involvement is intrinsicallymotivated, whereas situational involvement is extrinsicallymotivated (Kapferer and Laurent, 1985). So, consumers who areenduringly involved are looking for hedonic benefits, whereasconsumers who are situational involved are engaged in goal-directed behaviors (Hoffman and Novak, 1996).”

Otherwise, perceived risk also influence on consumer behaviordepending on web stimulus (Yoh et al., 2003; Smith andSivakumar, 2004; Park and Stoel, 2005), so it is considered as amediator variable. Koernig (2003) affirm that if the tangibility ofthe service is increased in the online environment, the perceivedrisk related to the activity will decrease. In addition, Slovicet al. (1980) argue that the presence of images with low qualitywithin virtual stores cause a negative influence on consumerbehavior, and also it will increase the perceived risk. On theother hand, Chang and Wu (2012) study the effect of perceivedrisk by analyzing the moderating effects of decision-makingstyle, it means involvement or heuristics, in the virtual context,obtaining that perceived risk toward a website influences onpurchasing intention through cognition, and this affects onattitudes. Within this line, Keng et al. (2012) analyzed themoderating roles of needing for touching and implication withthe product.

Based on previous literature, four additional hypotheses foreach three webmosphere analyzed stimuli are proposed respectto mediator variables –two of them related to involvement– asfollow (a: Navigational structure; b: Music; c: Presentation ofproducts):

H4a,b,c: Involvement with apparel will have a positivemediator effect between the three different web stimuli (i.e.,music, navigational structure and presentation of products)and online user responses (i.e., internal and behavioralresponses)H5a,b,c: Involvement with virtual shopping will have a positivemediator effect between the three different web stimuli andonline user responsesH6a,b,c: Atmospheric responsiveness will have a positivemediator effect between the three different web stimuli andonline user responsesH7a,b,c: Perceived risk will have a positive mediator effectbetween the three different web stimuli and online userresponses

To test above hypotheses, an extended model is proposed(Figure 1) based on the S-O-R paradigm. The direct effects

between specific web technological attributes related to non-verbal (i.e., hedonic: music and visualization of products)and verbal (i.e., utilitarian: navigational structure) webdimensions on online user responses (i.e., internal states andbehavioral responses) were analyzed, as well as examining theinfluence of mediator variables (i.e., involvement, atmosphericresponsiveness, and perceived risk) between both constructs (i.e.,web stimuli and online user responses).

MATERIALS AND METHODS

Sample and Procedure: ComputerExperimental DesignA 2 (“free” versus “hierarchical” navigational structure) × 2(“on” versus “off” music) × 2 (“moving” versus “static” images)between-subjects computer experimental design was used tocarry out the research. Similar types of computer experimentaldesigns with more/less, different/similar web stimuli have beenanalyzed in previous studies (e.g., Eroglu et al., 2001, 2003;Dailey, 2004; Choura and Saber, 2005; Nesset et al., 2011; Changand Wu, 2012). A fictitious store was created for this research(named E-Fashion) to avoid the effects of prior experience witha e-retailer. That virtual shop offered fashion apparel (clothesand accessories) for men and women, in two designs (casualand formal). To standardized products to compare men andwomen, the number of clothes and accessories were the samefor the two genders. To create the content of the virtual shop,we based on a homepage (Schneiderman, 1998; Nielsen, 2000),which incorporates the similar links and web sites as other virtualapparel shops.

In sum, the experimental design consisted on the doublecombination between three types of web stimuli:

(1) Navigation structure stimulus. We used two different webnavigational structure used in this study:

• “Free navigation structure” of the website means thatcostumer can move easily through the website. Forthat, the following tools have been included within thisexperimental condition: “home link,” “location pattern,”“searching window,” “lateral menus in all webpages of thesite”). Consequently, the user is not controlled by marketerwith restrictive navigational bars.• “Hierarchical navigational structure” means that user only

can use the “previous” and “next” navigational bars tomove through the website. In consequence, the previoustools (“home link,” “location pattern,” “searching window,”“lateral menus in all webpages of the site”) have beeneliminated of this experimental condition in order to createmore restrictive navigation to user.

(2) Music stimulus. Two different music web stimuli were usedin this study:

• “On-music” within the webpage. In this case, the musicwas included within the experiment. Music was testedpreviously to analyze which kind of music was preferredby consumers, using music that already existed. The kind







FIGURE 1 | Stimulus-Organism-Response (S-O-R) proposed model to measure the effects of webmosphere non-verbal and verbal communicationstimuli on online user responses.

of music included was instrumental and very dynamic, butit was not readily identifiable with any particular singer orspecific group. It was used music of the album Traveler’sguide by Jahzzar, which is licensed under a Attribution-ShareAlike 3.0 International License.• “Off-music.” In this case, the music was eliminated of the

experiment.

(3) Presentation of product stimulus. We used two differentpresentations of products (Martin et al., 2005; Lorenzo et al.,2007):

• “Static images” web design presents the possibility of seeingfrontal images of all products and a window with staticphotos of different models showing several outfits.• “Moving images” web design allows consumers 360◦

visualization of products using gif images according toanimation definition defined by Dahan and Srinivasan(2000) and Sundar and Kalyanaraman (2004), as well awindow with videos in which the models show the fashionapparel of the store (Choura and Saber, 2005; Nesset et al.,2011).

As a result of combination the three webmosphericmanipulations (2 × 2 × 2), eight different websites weredefined:

(1) Hierarchical navigation+ on music+moving images;(2) Hierarchical navigation+ off music+moving images;(3) Hierarchical navigagion+ on music+ static images;(4) Hierarchical navigation+ off music+ static images;(5) Free navigation+ on music+moving images;(6) Free navigation+ off music+moving images;(7) Free navigation+ on music+ static images;(8) Free navigation+ off music+ static images.

In all experimental designs, the content of webpages is thesame (i.e., company description, products description, shoppingcar, security and privacy policy, promotions, fictitious moneyto buy, link to register, new arrivals and sale section, contactand search engine), excepting the changes of stimuli indicatedin each manipulation. From both links previously indicated,according to the each experiment, gifs animated or static images(in the case of “presentation of product” experimental design),and music or not (in the case of “music” experimental design)were included.

Before starting the experiment, a pre-test was conducted to adifferent group to ensure that subjects’ responses give us differentperception by inclusion of the above atmospheric manipulations.

The final sample consisted of 400 participants who weredivided into eight groups (two of them with the navigationalstructure manipulation, two of them with music manipulation,two of them with presentation manipulation). Participants wererandomly assigned to one of these experimental conditions.All members of each group were undergraduate students (18–25 years old, and half of the same were men and the other halfwere women), due to it was a group with enough experienceand positive predisposition to the Internet as to carry outthis experiment in adequate way (Ozok and Wei, 2010). Allgroups were exposed same external environment conditions(place and space) throughout the same day. Besides these aspects,all experiments were checked under same conditions explainedabove: time (50 min, since in the pre-test was demonstrated thatpeople spent on average this time until take a final decision),money (each subject had 200€; of budget, which is an average ofthe total budget that they usually have, according with a previousexploratory analysis) and elements of web site. In order to analyzean equilibrated model, each group was consisted of 50 subjects.

After participants finished the experimental task they werecomplete a final online questionnaire which includes affective,







TABLE 1 | Measurement of variables.

Variables Items Kind of Scale (five points) Source of the Scale

Internal states

Affective Attitude towards online purchase:· Disappointed/appointed· Unfavorable/favorable· Negative/positiveEmotion toward online purchase:· Bored/non-bored· Unhappy/Happy· Unaroused/aroused

Semantic differential question Mehrabian and Russell, 1974;Eroglu et al., 2003; Ivonin et al.,2013

Cognitive Believes about online apparel purchase:· Easy· Cheap· EnjoyLearning and knowledge obtained:· Learning about new media for shopping· Knowledge about online purchase

Likert scale Lorenzo et al., 2007;Schlosser, 2003

Satisfaction General satisfaction with the visit:· Browsing satisfaction· Security satisfaction· Satisfaction with the shopping experienceSatisfaction with design:· Design of the store is liked by users· Satisfaction with the presentation of products within the onlinestore· Navigational elements help users move across the website

Likert scale Anderson and Srinivasan,2003; Eroglu et al., 2003;Cristóbal, 2005; Flavián et al.,2004, 2005; O’Cass andFenech, 2003Lorenzo et al.,2007SUMI scale

Behavioral Responses

Approach responses Behavior intention and opinion about the website:· Revisiting the store· Recommendation of the store to other people· User would have spent more money in the website· User would have spent more time in the website· Navigational elements attract attention to usersNo expected shopping behavior:· Users buy more products than planned previously· Users spend more money than planned previously

Probability scale Zeithaml et al., 1996; Shermanet al., 1997; Eroglu et al., 2003;Bigné and Andreu, 2004;Dailey, 2004; Flavián et al.,2004, 2005; Gallarza and Gil,2006; Lorenzo et al., 2007;Ozok and Wei, 2010; Chen andTeng, 2013

Real shopping outcomes · Time spent during the visit· Products bought· Money spent

Click-through Eroglu et al., 2003

Mediators variables

Involvement Involvement with apparel:· I like going shopping to see or buy apparel· I like wearing fashion apparelInvolvement with online purchase:· I like buying on the Internet· It is enjoyable buying on the Internet

Likert scale Kapferer and Laurent, 1985;Yoh et al., 2003; Keng et al.,2012; Kim, 2012

Atmospheric responsiveness · When visit online apparel stores I usually perceive the design ofwebsite· The design of online stores influence on my possible visit to them

Likert scale Eroglu et al., 2003

Perceived risk · Internet transmits security in the purchases· It is safe to buy apparel without taking them previously· I would buy other products such as tickets, books, digital music,computer goods. . .

Likert scale Anderson and Srinivasan,2003; O’Cass and Fenech,2003; Flavián et al., 2004,2005; Gregg and Walczak,2010; Kim, 2012

cognitive and satisfaction measures (i.e., user internal states),behavioral responses (i.e., approach responses toward thewebsite), and mediators variables (i.e., involvement with appareland online purchase, atmospheric responsiveness, and perceivedrisk). All items, kinds of scales and sources used to generate theonline questionnaire are showed in Table 1. Participants of eightexperiments were asked by the same questionnaire, it means,

eight groups of questions corresponding to each scale, which itwas a total of 36 items.

Finally, based on Lorenzo et al. (2007), the web-based toolcreated for this research involved an automatic tracking process.This software was able to track and record all click-troughs andtimes related with the browsing behavior during the experimentto obtain behavioral responses in the proposed model (i.e.,







real shopping outcomes) to clarify the data captured by thesoftware.

It is worth noting that that this study was carried out inaccordance with the current ethical and legal recommendationsabout privacy of personal data. Throughout the whole researchwe take into account also the ICC/ESOMAR International Codeon Market and Social Research practices and norms.

A Model of Web Technology Effects onUser Responses. Measurement ofVariablesWe propose an extended model (Figure 1) based on the S-O-Rparadigm (Mehrabian and Russell, 1974), to test the influence ofthree specific atmospheric cues on online shopper responses. Themain objective is to analyze the differences of response betweenthe experimental groups, as well as examining the influence ofmediator variables between the constructs. The analyzed variablesin each construct are showed in Table 1.

Statistical TechniquesRegarding the statistical techniques used in this research, asour major objective is to analyze whether or not there weresignificant differences of behavior between the groups andthe effects of each online atmospheric manipulation cues onfive types of dependents constructs respect to the consumer:Affective, cognition and satisfaction (as internal states), andapproach/avoidance responses and real shopping outcomes(as consumer behavioral responses), a multivariate analysis ofvariance (MANOVA) is used to test this part. Additionally,a multivariate analysis of covariance (MANCOVA) is usedto analyze the mediator effect of involvement, atmosphericresponsiveness, and perceived risk factors between web verbaland non-verbal communication stimuli and user responses.

Previously, as the variable used as dependents and mediatorsare latent variables, it means, non-observable variables measuredthrough scales (see Table 1), we used exploratory factor analysisto describe variability among observed, correlated variables interms of a potentially lower number of unobserved variablescalled factors. Factor analysis searches for such joint variationsin response to unobserved latent variables.

RESULTS

In Tables 2 and 3 are showed the data obtained after factorialanalysis in each construct used, which are necessary to be used inthe multivariate analysis (MANOVA and MANCOVA). The firststep of the analysis is to test the model fit, that is, if it is a good ideato proceed with a factorial analysis for the data, which is showedin Table 2. The next step is to analyze the factors obtained fromthe factorial analysis, which are showed in Table 3.

Specifically, in Table 2 is showed the association levelindicators between analyzed variables which compose thedifferent constructs of our model. For example, “affective”construct shows that its variables are correlated, there is nomulticollinearity since the correlation matrix determinant isgreater than 0.00001, the K-M-O index is greater than 0.8 and

Bartlett’s test of sphericity is highly significant. These indicators,together with the measure of sample adequacy (>0.5), indicategood association level of items on its construct (Iacobucci, 1994;Hair et al., 2006). In the rest of construct (i.e., “cognitive,”“satisfaction,” “behavioral responses,” and “mediator variables”)the results are similar, meaning these results a good associationlevel of items on their constructs. Lower levels are obtainedin the “mediator variables” due to the inclusion of differenttypes of covariates with different meanings (i.e., involvement,atmospheric responsiveness, perceived risk). Nevertheless, theglobal statistics indicators are acceptable (i.e., K-M-O index0.549; p-value < 0.01; measure of sample adequacy withcoefficients between 0.49 and 0.58). It means a good associationlevel between mediator variables.

In Table 3 is showed the factorial analyses for each analyzedconstruct (i.e., “affective,” “cognitive,” “satisfaction” as internalvariables construct; “approach responses” as behavioral responsesconstruct; and “mediator variables or covariates” construct). Eachobtained factor shows the name of items with which is composed.For example, “affective variables construct” is composed by twofactors: “Attitude” and “Emotion.” In this case, each factor iscomposed by three items. The rest of constructs are composedby two factors, excepting “mediator variables” construct which iscomposed by four factors.

In the case of “real shopping outcomes” (within “behavioralresponses” construct) the factorial analysis is not possible dueto the variables are uncorrelated. These variables were obtainedthrough click-through -it has been indicated in Table 1- carriedout by users during their visit within online retail store.Nevertheless, the variables are used in the ANOVA and ANCOVApost analysis (Davino and Romano, 2014).

For each factor has been analyzed the Eigenvalues, thepercentage of explained variance, and specially, the CronbachAlpha index which measures the level of reliability of factor.In all cases, the eigenvalue is greater than 1.000, the level ofexplained variance is higher 0.5 and the Cronbach alpha near1.000 (Cronbach, 1970). It means a good fit of the factorialanalysis (Iacobucci, 1994; Hair et al., 1999).

In sum, as we can observe in Tables 2 and 3, the CronbachAlpha is used as reliability measurement of all analyzedconstructs (Cronbach, 1970). Moreover, the Wilks’ Lambdastatistic was used to test the global significance (Iacobucci, 1994;Hair et al., 1999), using the F-Snedecor statistic and, as acceptedconfidence statistic level, 95% (p-value = 0.05) and 90% (p-value = 0.1), as it is showed in Tables 4 and 5. In consequence,validity and reliability obtained from this factorial analysis showus the suitability of its application for the next multivariateanalysis. Moreover, the content validity is obtained through theapplication of literature in the development of our proposedmodel.

A comparative analysis (MANOVA vs. MANCOVA) isdeveloped in order to compare the effects of web stimuli onuser responses without (MANOVA) and with (MANCOVA) theinclusion of mediator variables, respectively. Each significanceitem (p-value < 0.05 and 0.1) confirm the webmospheric cuespreferred by users. The global contrasts are showed in Table 5(internal states) and Table 6 (behavioral responses).







TABLE 2 | Association level indicators between analyzed variables within the model.

Online user internal states

Affective Cognitive Satisfaction

Correlation matrix Correlated variables Correlated variables Correlated variables

Correlation matrix determinant 0.016 0.193 0.059

Kaiser-Meyer-Olkin index 0.894 0.687 0.872

Barlett’s sphericity proof 1633.101 p < 0.001 652.148 p < 0.001 1123.731 p < 0.001

Anti-image correlation matrix Partial correlation Reduced partial correlation Partial correlation

Measure of sample adequacy Coefficient between 0.85 and 0.91 Coefficient between 0.50 and 0.75 Coefficient between 0.85 and 0.89

Behavioral responses

Approach responses

Correlation matrix Correlated variables

Correlation matrix determinant 0.075

Kaiser-Meyer-Olkin index 0.748

Barlett’s sphericity proof 1027.201 p < 0.001

Anti-image correlation matrix Reduced partial correlation

Measure of sample adequacy Coefficients between 0.58 and 0.88

Mediator variables

Involvement; atmospheric responsiveness; perceived risk

Correlation matrix Partial correlated variables

Correlation matrix determinant 0.354

Kaiser-Meyer-Olkin index 0.549

Barlett’s sphericity proof 410.489 p < 0.001

Anti-image correlation matrix Partial correlation

Measure of sample adequacy Coefficients between 0.49 and 0.58

In order to clarify the specific impact of each stimulus oneach consumer response factor, the individual contrasts areshowed for each web manipulation in Table 6 (i.e., navigationalstructure), Table 7 (i.e., music) and Table 8 (i.e., presentation ofproducts), through ANOVA (univariate analysis of variance) andANCOVAS (univariate analysis of covariance) analysis (Davinoand Romano, 2014).

Effects of Verbal CommunicationStimulus (Utilitarian Dimension) on UserResponses: Navigational StructureThe global contrast MANOVA test (Table 4) shows significantdifferences between groups analyzed regarding the relationshipbetween this verbal web stimulus and all internal states factors (p-value < 0.05). In the ANOVA test (Table 6) we can see the positiveeffect of “free navigation” on user’s internal states. Specifically,the factors like emotion (as affective internal state), believes andlearning/knowledge (as cognitive states), and satisfaction with thedesign (as satisfaction state) are related positively with the “freenavigation” as verbal communication stimulus.

In Table 4, MANCOVA test includes four mediator factorsin which we can see the significance impact of involvement(with apparel and with online shopping) on all of three internalstates. Based on Table 6, the ANCOVA test shows that users whowere exposed to an online shopping environment with the “freenavigation” show more favorable internal responses than those

who were exposed to the “hierarchical navigation” during theirvisit within online retail store (the factors attitude and generalsatisfaction with the visit are not significantly different betweenboth navigational structure including involvement as mediatorfactors).

Regarding atmospheric responsiveness as mediator factor,Table 4 shows the inexistence of significant differences respectto affective and satisfaction internal states (p-value = 0.957 and0.747, respectively). Nevertheless, atmospheric responsivenessinfluence on cognitive states, and we can affirm that users preferfree navigation.

Finally, respect to perceived risk mediator factor, Table 4shows significant influence on affective and cognitive internalstates, whose preference is focus on “free navigation,” but perceiverisk do not influence on satisfaction.

Regarding the effects on behavioral responses, Table 5shows the inexistence of significant differences between analyzedgroups exposed to “hierarchical” versus “free” navigationalstructure on behavioral responses. The impact of internalvariables as mediator factors between this non-verbal stimuliand behavioral responses has not been analyzed. It couldbe the cause of this non-significant effect. The univariatetest of Table 6 does not show significant difference in anyanalyzed factors related to behavioral responses. Moreover,the inclusion of mediator factors (involvement, atmosphericresponsiveness, perceived risk) do not influence on the behavioralresponses.







TABLE 3 | Factorial analyses for each construct (variables and covariates).

Affective variables (a) Factors

Attitude Emotion

Dissapointed/appointed 0.822

Unfavorable/favorable 0.883

Negative/positive 0.860

Bored/non-bored 0.921

Unhappy/Happy 0.644

Unaroused/aroused 0.607

Eigenvalues of the factors 4.215 0.624

% Explained variance 70.250 10.403

Cronbach alpha 0.9061 0.8308

Cognitive variables (b) Factors

Believes about online apparel purchase Learning and knowledge obtained

Easy 0.873

Cheap 0.909

Enjoy 0.866

Learning about new media for shopping 0.847

Knowledge about online purchase 0.838




Satisfaction variables (c) Factors

General satisfaction with the visit Satisfaction with the design

Browsing satisfaction 0.835

Security satisfaction 0.709

Satisfaction with the shopping experience 0.863

Design of the store is liked by users 0.726

Satisfaction with the presentation of products within theonline store

0.597

Navigational elements help users move across thewebsite

0.871




Behavioral responses Factors

Approach responses to the website (d) Behavior intention and opinion about the website No expected shopping behavior

Revisiting the store 0.850

Recommendation of the store to other people 0.841

User would have spent more money in the website 0.731

User would have spent more time in the website 0.796

Navigational elements attract attention to users 0.464

Users buy more products than planned previously 0.892

Users spend more money than planned previously 0.911




(Continued)







TABLE 3 | Continued

Mediator variables or covariates (e) Factors

Involvement with theonline shopping

Involvement with theapparel

Atmosphericresponsiveness

Perceived risk withthe virtual shopping

I like buying on the Internet 0.846

It is enjoyable buying on the Internet 0.796

I like going shopping to see or buy apparel 0.848

I like wearing fashion apparel 0.854

When visit online apparel stores I usually perceive thedesign of website

0.760

The design of online stores influence on my possiblevisit to them

0.857

Internet transmits security in the purchases 0.603

It is safe to buy apparel without taking them previously 0.710

I would buy other products such as tickets, books,digital music, computer goods. . .

0.577

Eigenvalues of the factors 2.052 1.442 1.199 1.021

% Explained variance 22.798 16.021 13.317 11.346

Cronbach alpha 0.6400 0.6388 0.5290 0.1942

(a), (b), (c), (d), (e) Extraction Method: Principal Components Analysis; Rotation Method: Kaiser Varimax Normalization; (a), (b), (c) The rotation has converged in 3 iterations;(d) The rotation has converged in 4 iterations (e) The rotation has converged in 5 iterations.Bold values reflect which items load on each factor.

Effects of Non-verbal CommunicationStimulus (Hedonic Dimension) on UserResponses: Music in the WebsiteTable 4 shows significant differences between this non-verbalstimulus and internal states, specifically respect to cognitionstates. Moreover, analyzing the mean differences between bothtypes of manipulations (“on music” versus “off music”), in theTable 7 we can observe that users who were exposed to anonline shopping environment with the “on music” web designsshow more favorable learning and knowledge of the website andhigher satisfaction with the design than those who were exposedto “off music” conditions within the store. However, there arenot differences between groups respect to affective internal statebased on Table 4 (p-value= 0.932).

On the other hand, we can affirm that that music is anon-verbal stimulus which does not cause differences betweenuser’s behavioral responses, and neither considering mediatingvariables.

Effects of Non-verbal CommunicationStimulus (Hedonic Dimension) on UserResponses: Presentation of ProductThe results obtained in satisfaction construct analyzed (Table 4)show that consumers who were exposed to an online shoppingenvironment with the “moving images” web designs are moresatisfied those who were exposed to the “static images” conditionswithin the store. Moreover, this non-verbal stimulus influencepositively on believes about online apparel purchase (Table 8).

Regarding the effects of non-verbal communication stimuluson behavioral responses (i.e., approach responses and realshopping outcomes), MANOVA test shows significant differences

(Table 5). The ANOVA test (Table 8) indicates that users exposedto “moving images” stimulus show more favorable behaviorintention and opinion about the website that the users exposedto “static images.” Moreover, all real shopping outcomes (i.e.,products bought, time spent during the visit and money spent)are more positives in a “moving images” web design. Specifically,the involvement (with apparel and with online shopping)and perceived risk affect positively on approach responses. Incontrast, only involvement with apparel affect positively on realshopping outcomes. The rest of mediated factors considered havenot significant influence on behavioral responses.

DISCUSSION, LIMITATIONS ANDFUTURE RESEARCH

The S-O-R paradigm can be useful to illustrate the influenceof web atmosphere on consumers. This model indicates thatexternal stimuli (like web atmospheric cues) affect consumers’internal states and, in turn, they have an effect on behavioralresponses, within an online shopping context (Mehrabian andRussell, 1974; Eroglu et al., 2001). In this work, we obtainedthat three dimensions manipulated (i.e., navigational structureas verbal stimulus, and music and presentation of products asnon-verbal stimuli) affect significantly on internal and behavioraluser responses. The results are relevant for retail marketersbecause they must offer attractive online store as unique orcomplementary sale channel to entice people into their shops.

The main contribution of this study, compared to the previousone, is that we have considered many and different variablesand, what it can be more important, the inclusion of somemediator variables that can modify the consumer behavior andcan explain the modification of some established relationships







TABLE 4 | Global contrast: MANOVA vs. MANCOVA on online users’ internal states.

Factors Effects 3 Wilks F Snedecor Significance Hypotheses

MAN MANC MAN MANC MAN MANC MAN MANC

Internal states Affective NAV 0.984 0.982 3.124 3.461 0.045∗ 0.032∗ H1−1

MUS 1.000 1.000 0.071 0.068 0.932 0.935 H2−1

Present 0.997 0.997 0.581 0.611 0.560 0.543 H3−1

INV-A 0.916 17.739 <0.001∗ H4 a

INV-OS 0.961 7.856 <0.001∗ H5

AR 1.000 0.044 0.957 H6

PR 0.988 2.307 0.101 ∗∗ H7

Cognitive NAV 0.966 0.964 6.872 7.263 <0.01∗ <0.01∗ H1−2

MUS 0.991 0.988 1.816 2.374 0.164 0.094∗∗ H2−2

Present 0.992 0.995 1.557 0.903 0.212 0.406 H3−2

INV-A 0.917 17.613 <0.001∗ H4 b

INV-OS 0.946 11.043 <0.001∗ H5

AR 0.656 101.529 <0.001∗ H6

PR 0.979 4.096 0.017∗ H7

Satisfaction NAV 0.948 0.948 10.677 10.673 <0.001∗ <0.001∗ H1−3

MUS 0.993 0.991 1.446 1.748 0.237 0.175 H2−3

Present 0.986 0.986 2.707 2.669 0.068∗∗ 0.071∗∗ H3−3

INV-A 0.925 15.767 < 0.001∗ H4 c

INV-OS 0.968 6.342 < 0.001∗ H5

AR 0.998 0.292 0.747 H6

PR 0.991 1.708 0.183 H7

∗p-value < 0.05 confidence level: 95%. NAV, navigational structure; MUS, music; PRESENT, presentation of products; INV-A, involvement with apparel; INV-OS,involvement with online shopping; AR, atmospheric responsiveness; PR, perceived risk; MAN, MANOVA; MANC, MANCOVA. ∗∗p < 0.1.

TABLE 5 | Global contrast: MANOVA vs. MANCOVA on online users’ behavioral responses.

Factor and variables Effects 3 Wilks F Snedecor Significance Hypotheses

MAN MANC MAN MANC MAN MANC MAN MANC

Behavioral responses Approach responses (factor) NAV 0.999 0.999 0.195 0.166 0.823 0.847 H1−4

MUS 0.994 0.994 1.121 1.220 0.327 0.296 H2−4

Present 0.982 0.983 3.491 3.282 0.031∗ 0.039∗ H3−4

INV-A 0.951 10.074 <0.001∗ H4 d

INV-OS 0.985 2.979 0.052∗ H5

AR 0.989 2.136 0.120 H6

PR 0.977 4.654 0.010∗∗ H7

Real shopping outcomes (variables) NAV 0.997 0.996 0.407 0.534 0.748 0.659 H1−5

MUS 0.994 0.993 0.801 0.946 0.494 0.418 H2−5

Present 0.912 0.911 12.555 12.532 <0.001∗ <0.001∗ H3−5

INV-A 0.975 3.365 0.019∗ H4 e

INV-OS 0.989 1.400 0.242 H5

AR 0.999 0.153 0.928 H6

PR 0.990 1.295 0.276 H7

∗p-value < 0.05, ∗∗p-value < 0.1.

studied previously. Based on results, and in line with the resultsobtained by Dailey (2004), restrictive navigation act as barriersover web navigation. In this sense, a hierarchical navigation canbe perceived as more restrictive. In addition, consumer’s personalcharacteristics (i.e., involvement, atmospheric responsiveness,and perceived risk) mediate these results. In general, navigationaldesign, as utilitarian or verbal atmospheric cue, has a greater

effect especially on cognitive states (i.e., knowledge, learning andbeliefs on the Internet). It would be necessary and interestingto analyze the impact of internal variables as mediator factorbetween stimulus and behavioral responses (Lorenzo et al., 2007).

Moreover, positive user internal states (specifically,satisfaction) are more favorable with a “music” and “movingimages” web designs and the behavioral responses are more







TAB

LE6

|Ind

ivid

ualc

ont

rast

s:A

NO

VAS

vs.A

NC

OVA

So

nco

nsum

erre

spo

nses

wit

hth

eN

AV

IGA

TIO

NA

LS

TR

UC

TU

RE

asV

ER

BA

Lco

mm

unic

atio

nst

imul

i.

Verb

alw

ebm

osp

heri

ccu

e:na

vig

atio

nals

truc

ture

(“hi

erar

chic

alvs

.fre

e”)

Co

nsum

erre

spo

nses

Fact

ors

and

ítem

sF

Sig

nifi

canc

eM

ean

diff

eren

ces

(“hi

erar

chic

al–

free

”)

AN

OVA

AN

CO

VAA

NO

VAA

NC

OVA

AN

OVA

AN

CO

VA

Inte

rnal

stat

esA

ffect

ive

(fact

ors)

Att

itude

0.46

20.

869

0.49

70.

352

−0.

068

−0.

091

Emot

ion

5.81

25.

804

0.01

60.

016

−0.

240∗

−0.

235∗

Cog

nifiv

e(fa

ctor

s)B

elie

ves

abou

tonl

ine

appa

relp

urch

ase

3.94

03.

666

0.04

80.

056

−0.

197∗

−0.

154∗

Lear

ning

and

know

ledg

eob

tain

ed9.

572

10.3

740.

002

0.00

1−

0.30

7∗−

0.30

3∗

Sat

isfa

ctio

n(fa

ctor

s)G

ener

alsa

tisfa

ctio

nw

ithth

evi

sit

2.22

51.

964

0.13

70.

162

0.14

90.

136

Sat

isfa

ctio

nw

ithth

ede

sign

18.9

9919

.832

<0.

001

<0.

001

−0.

424∗

–0.4

30∗

Beh

avio

ralr

espo

nses

App

roac

hre

spon

ses

(fact

ors)

Beh

avio

rin

tent

ion

and

opin

ion

abou

tth

ew

ebsi

te0.

083

0.02

50.

773

0.87

40.

029

0.01

6

No

expe

cted

shop

ping

beha

vior

0.30

80.

311

0.57

90.

578

−0.

056

–0.0

55

Rea

lsho

ppin

gou

tcom

es(v

aria

bles

)Ti

me

spen

tdur

ing

the

visi

t0.

139

0.19

20.

709

0.66

20.

400

0.46

9

Pro

duct

sbo

ught

0.78

21.

120

0.37

70.

291

−0.

230

−0.

277

Mon

eysp

ent

0.38

30.

616

0.53

70.

433

−3.

980

−5.

076

∗p-

valu

e<

0.05

.AN

OVA

,var

ianc

eun

ivar

iabl

ean

alys

is;A

NC

OVA

,cov

aria

nce

univ

aria

ble

anal

ysis

.

TAB

LE7

|Ind

ivid

ualc

ont

rast

s:A

NO

VAS

vs.A

NC

OVA

So

nco

nsum

erre

spo

nses

wit

hth

eM

US

ICas

NO

N-V

ER

BA

Lco

mm

unic

atio

nst

imul

i.

No

n-ve

rbal

web

mo

sphe

ric

cue:

mus

icin

the

reta

ilo

nlin

est

ore

(“o

ffvs

.on

mus

ic”)

Co

nsum

erR

esp

ons

esFa

cto

rsan

dit

ems

FS

igni

fica

nce

Mea

nd

iffer

ence

s(“

Off

-O

nm

usic

”)

AN

OVA

AN

CO

VAA

NO

VAA

NC

OVA

AN

OVA

AN

CO

VA

Inte

rnal

stat

esA

ffect

ive

(fact

ors)

Att

itude

0.13

20.

127

0.71

60.

722

−0.

036

−0.

035

Emot

ion

0.01

00.

005

0.92

20.

942

0.01

0−

0.00

7

Cog

nifiv

e(fa

ctor

s)B

elie

ves

abou

tonl

ine

appa

relp

urch

ase

0.59

30.

218

0.44

20.

641

−0.

077

−0.

038

Lear

ning

and

know

ledg

eob

tain

ed2.

991

4.45

50.

085

0.03

5−

0.17

1∗∗

−0.

198∗

Sat

isfa

ctio

n(fa

ctor

s)G

ener

alsa

tisfa

ctio

nw

ithth

evi

sit

0.11

00.

159

0.74

00.

691

–0.0

33–0

.039

Sat

isfa

ctio

nw

ithth

ede

sign

2.80

43.

294

0.09

50.

070

−0.

163∗∗

−0.

175∗∗

Beh

avio

ralr

espo

nses

App

roac

hre

spon

ses

(fact

ors)

Beh

avio

rin

tent

ion

and

opin

ion

abou

tth

ew

ebsi

te1.

131

1.49

80.

288

0.22

2−

0.10

6−

0.12

0

No

expe

cted

shop

ping

beha

vior

1.12

00.

978

0.29

10.

323

0.10

60.

098

Rea

lsho

ppin

gou

tcom

es(v

aria

bles

)Ti

me

spen

tdur

ing

the

visi

t0.

570

0.79

00.

451

0.37

5−

0.81

0−

0.94

8

Pro

duct

sbo

ught

0.59

10.

681

0.44

20.

410

0.20

00.

215

Mon

eysp

ent

1.32

51.

506

0.25

00.

221

7.40

67.

918

∗p-

valu

e<

0.05

,∗∗p-

valu

e<

0.1.







TAB

LE8

|Ind

ivid

ualc

ont

rast

s:A

NO

VAS

vs.A

NC

OVA

So

nco

nsum

erre

spo

nses

wit

hth

eP

RE

SE

NTA

TIO

NO

FP

RO

DU

CT

Sas

NO

N-V

ER

BA

Lco

mm

unic

atio

nst

imul

i.

No

n-ve

rbal

web

mo

sphe

ric

cue:

pre

sent

atio

no

fp

rod

ucts

(“st

atic

vs.m

ovi

ngim

ages

”)

Fact

ors

and

ítem

sFa

cto

rsan

dít

ems

FS

igni

fica

nce

Mea

nd

iffer

ence

s(“S

tati

c-

Mo

ving

imag

es”)

AN

OVA

AN

CO

VAA

NO

VAA

NC

OVA

AN

CO

VAA

NC

OVA

Inte

rnal

stat

esA

ffect

ive

(fact

ors)

Att

itude

0.13

20.

090

0.71

60.

765

−0.

036

−0.

029

Emot

ion

1.03

61.

098

0.30

90.

295

−0.

101

−0.

102

Cog

nifiv

e(fa

ctor

s)B

elie

ves

abou

tonl

ine

appa

relp

urch

ase

3.01

61.

651

0.08

30.

200

0.17

3∗∗

0.10

3

Lear

ning

and

know

ledg

eob

tain

ed0.

083

0.12

00.

774

0.72

90.

029

0.03

2

Sat

isfa

ctio

n(fa

ctor

s)G

ener

alsa

tisfa

ctio

nw

ithth

evi

sit

0.67

20.

654

0.41

30.

419

−0.

082

−0.

078

Sat

isfa

ctio

nw

ithth

ede

sign

4.80

54.

576

0.02

90.

033

−0.

213∗

−0.

206∗

Beh

avio

ralr

espo

nses

App

roac

hre

spon

ses

(fact

ors)

Beh

avio

rin

tent

ion

and

opin

ion

abou

tth

ew

ebsi

te6.

672

6.22

40.

010

0.01

3−

0.25

8∗−

0.24

4∗

No

expe

cted

shop

ping

beha

vior

0.32

40.

323

0.56

90.

570

−0.

057

−0.

056

Rea

lsho

ppin

gou

tcom

es(v

aria

bles

)Ti

me

spen

tdur

ing

the

visi

t18

.963

19.2

72<

0.00

1<

0.00

1−

4.67

0∗−

4.67

7∗

Pro

duct

sbo

ught

20.2

3219

.535

<0.

001

<0.

001

−1.

170∗

−1.

151∗

Mon

eysp

ent

18.9

7918

.404

<0.

001

<0.

001

−28

.028∗

−27

.643∗

∗p-

valu

e<

0.05

,∗∗p-

valu

e<

0.1.

probable with, specially, “static images” conditions. In addition,users’ personal characteristics also modifies relatively the finalshopping human behavior.

It is necessary to take into account the inclusion of andappropriate music for target market (i.e., more dynamic foryoung people and more paused for old people) in order tooffer users a more attractive virtual shop window. Nevertheless,according our results, the music is a web stimulus not veryrelevant for the sample.

Moreover, the use of verbal communication stimulus, is alsoimportant to improve the user internal states, as can be thesatisfaction toward the use of a website, as Shankar et al. (2003)postulated. However, it is necessary to take into account thenon-verbal stimuli, as the music or the way of presenting theproducts, affect directly on behavioral responses, as satisfaction.In consequence, website designers should include full productscatalog, location information, searching engine and direct accessto home page, in all websites of site, improves the navigationdeveloped by user and, in turn, increase his/her internal statesand shopping responses.

Other recommendation for e-marketers is to analyze theclick-through. They represent real shopping outcomes such astime spent during the visit, products bought, and money spent.Web analytic allows studying the objective shopping outcomesthrough some free web applications like Google Analytics.The combination of subjective (i.e., internal states, approachresponses, mediator variables) and objective data (i.e., click-through) offer e-marketer full information on their customers inorder to analyze their habits and behaviors.

As main limitations, this research has been only focusedon the analysis of relationships between three specific onlineatmospheric cues and internal states/behavioral responses. It is alimitation because; the internal states mediated this relationship,aspect not tested in this research. On the other hand, accordingto literature, these kinds of variables mediate the relationshipbetween webtmospheric cues and consumers’ internal states(Choudhary, 2016). However, in this research their influenceon internal and behavioral responses was studied in order toanalyze the degree of significance in all the constructs. Moreover,the sample was composed of young people who were familiarwith web media (i.e., different web designs, a lot of kinds ofonline stores and experience with digital products, etc.). So,perhaps perceived risk mediator variable is not significance onthe satisfaction construct, due to sample is familiarity with onlineshopping.

Changes between treatments are not entirely ceteris paribus,however, there are many design conditions that change, especiallyin the navigation structure. This is an unavoidable problemin online experiments. So, the obtained conclusions should bestated carefully. This manuscript does not show the mixtureof three manipulations (e.g., free navigation with music anddynamic graphic). It shows that the particular free navigationchosen in this experiment is better than the particular hierarchicalnavigation, and that the particular music played is preferred thanno music, etc.

As future research, it would be interesting to analyze theuser experience toward other manipulations on webmospheric







(e.g., color, different navigational patterns, manipulations aboutmarketing mix elements such as prices, promotions, as suggestedLallemand et al., 2015), and the typology of products bought,main sections visited by users, etc. Moreover, our model couldbe improved using a structural equation modeling in order toanalyze the user’s internal variables [e.g., trust and familiarity,based on TAM (Kim, 2012) as mediator constructs betweenwebmosphere stimuli and behavioral responses (Chen andTeng, 2013)]. Finally, it would be interesting to compare thisresults with other different cultures (McMahan et al., 2014)where the use of the online shopping is different, as well asother characteristics that could affect to the online shopping(e.g., extroversion, familiarity, etc.), as Hofstede et al. (2010)proposed.

Moreover, this research opens a line of research in cognitiveneuroscience. According to Gazzaniga et al. (2002), cognitiveneuroscience is a branch of both psychology and neuroscience,overlapping with disciplines such as physiological psychology,cognitive psychology, and neuropsychology. Therefore, tocomplement the results obtained from a psychologicalperspective about consumer behavior and perceptions, it wouldbe necessary to develop some neuromarketing techniques tounderstand in a more detailed way how the consumers thinkand decide, which involves brain processes that our mindsare not aware of. When experimental designs are combinedwith neuromarketing techniques, they provide insights intoconsumer decisions and actions that are invisible to traditional

market research methodologies (Javor et al., 2013; Gani et al.,2015).

In addition, it would be necessary to study empirically the useof e-commerce by this kind of companies (Solaymani et al., 2012)in order to compare both perspectives, i.e., use of e-commerce bycompanies attending to webmosphere and opinion of consumersbased on webmosphere stimuli received through online store.

ETHICS STATEMENT

The study was exempt from ethics approval in accordance withthe policies of the University of Castilla-La Mancha.

AUTHOR CONTRIBUTIONS

All authors listed have made substantial, direct and intellectualcontribution to the work, and approved it for publication.

FUNDING

This study has been developed within the Research Projectsgranted by Junta de Comunidades de Castilla-La Mancha (Ref.PEII-2014-018-P) and Ministerio de Economía y Competitividad.Gobierno de España (Ref. PEJ-2014-A-74892).

REFERENCESAdelaar, T., Chang, S., Lancendorfer, K. M., Lee, B., and Morimoto, M. (2003).

Effects of media formats on emotions and impulse buying intent. J. Infr.Technol. 18, 247–266. doi: 10.1080/0268396032000150799

Alba, J., Lynch, J., Weitz, B., Janiszewski, C., Lutz, R., Sawyer, A., et al. (1997).Interactive home shopping: consumer, retailer, and manufacturer incentivesto participate in electronic marketplaces. J. Mark. 61, 38–53. doi: 10.2307/1251788

Anderson, R., and Srinivasan, S. S. (2003). E–satisfaction and e–loyalty: acontingency framework. Psychol. Mark. 20, 123–138. doi: 10.1002/mar.10063

Animesh, A., Pinsonneault, A., Yang, S.-B., and Oh, W. (2011). An odysseyinto virtual worlds: exploring the impacts of technological and spatialenvironments on intention to purchase virtual products. MIS Q. 35,789–810.

Argyriou, E. (2012). Consumer intentions to revisit online retailers: a mentalimagery account. Psychol. Mark. 29, 25–35. doi: 10.1002/mar.20405

Baron, R. M., and Kenny, D. A. (1986). The moderator-mediator variabledistinction in social psychological research: conceptual, strategic, and statisticalconsiderations. J. Pers. Soc. Psychol. 51, 1173–1182. doi: 10.1037/0022-3514.51.6.1173

Bigné, E., and Andreu, L. (2004). Emociones, satisfacción y lealtad del consumidoren entornos comerciales [Consumer’s emotions, satisfaction, and loyalty inshopping environments]. Distrib. Consum. 76, 77–87.

Bitner, M. (1992). Servicescapes: the impact of physical surroundings on customersand employees. J. Mark. 56, 57–71. doi: 10.2307/1252042

Celsi, R. L., and Olson, J. C. (1988). The role of involvement in attention andcomprehension processes. J. Consum. Res. 15, 210–224. doi: 10.1086/209158

Chang, M.-L., and Wu, W.-Y. (2012). Revisiting perceived risk in the context ofonline shopping: an alternative perspective of decision-making styles. Psychol.Mark. 29, 378–400. doi: 10.1002/mar.20528

Chang, H.-J., Eckman, M., and Yan, R.-N. (2011). Application of the stimulus-organism-response model to the retail environment: the role of hedonic

motivation in impulse buying behavior. Int. Rev. Retail Distrib. Consum. Res.21, 233–249. doi: 10.1080/09593969.2011.578798

Chen, M.-Y., and Teng, C.-I. (2013). A comprehensive model of theeffects of online store image on purchase intention in an e-commerceenvironment. Electron. Commer. Res. 13, 1–23. doi: 10.1007/s10660-013-9104-5

Childers, T. L., Carr, C. L., Peck, J., and Carson, S. (2001). Hedonic and utilitarianmotivations for online retail shopping behavior. J. Retail. 77, 511–535. doi:10.1016/S0022-4359(01)00056-2

Choudhary, F. (2016). “Online visual merchandising affects customer responsebehaviour- a conceptual model,” in Proceedings of the International Conferenceon Recent Trends in Engineering Science and Management. Jammu Taw: TheBusiness School, University of Jammu, 1–19.

Choura, F., and Saber, M. (2005). “Typologies des animations des bannièrespublicitaires sur internet: etude exploratoire,” in Proceedings of MarketingTrends Conference, Paris.

Cristóbal, E. (2005). Efectos del diseño de la tienda virtual en el comportamientode compra: tipificación del consumidor virtual [Effects of virtual store designon shopping behavior: profiling the online consumer]. Invest. Eur. Dir. Econ.Empresa 11, 203–221.

Cronbach, L. J. (1970). Essentials of Psychological Testing. New York, NY: Harperand Row.

Csikszentmihalyi, M. (1975). Beyond Boredom and Anxiety. San Francisco: Jossey-Bass.

Dahan, E., and Srinivasan, V. (2000). The predictive power of Internet–based product concept testing using visual depiction and animation.J. Prod. Innov. Manage. 17, 99–109. doi: 10.1016/S0737-6782(99)00029-6

Dailey, L. (2004). Navigational web atmospherics. Explaining the influence ofrestrictive navigation cues. J. Bus. Res. 5754, 1–9.

Davino, C., and Romano, R. (2014). Assessment of composite indicators usingthe ANOVA model combined with multivariate methods. Soc. Indic. Res. 119,627–646. doi: 10.1007/s11205-013-0532-3


https://doi.org/10.1080/0268396032000150799

https://doi.org/10.2307/1251788

https://doi.org/10.2307/1251788

https://doi.org/10.1002/mar.10063


https://doi.org/10.1037/0022-3514.51.6.1173

https://doi.org/10.1037/0022-3514.51.6.1173

https://doi.org/10.2307/1252042

https://doi.org/10.1086/209158


https://doi.org/10.1080/09593969.2011.578798

https://doi.org/10.1007/s10660-013-9104-5

https://doi.org/10.1007/s10660-013-9104-5

https://doi.org/10.1016/S0022-4359(01)00056-2

https://doi.org/10.1016/S0022-4359(01)00056-2

https://doi.org/10.1016/S0737-6782(99)00029-6

https://doi.org/10.1016/S0737-6782(99)00029-6

https://doi.org/10.1007/s11205-013-0532-3






Donovan, R. J., Rossiter, J. R., Marcoolyn, G., and Nesdale, A. (1994). Storeatmosphere and purchasing behavior. J. Retail. 70, 283–294. doi: 10.1016/0022-4359(94)90037-X

Eroglu, S. A., Machleit, K. A., and Davis, L. M. (2001). Atmospheric qualities ofonline retailing: a conceptual model and implications. J. Bus. Res. 54, 177–184.

Eroglu, S. A., Machleit, K. A., and Davis, L. M. (2003). Empirical testing of amodel of online store atmospherics and shopper responses. Psychol. Mark. 20,139–150. doi: 10.1002/mar.10064

Flavián, C., Guinalíu, M., and Gurrea, R. (2004). “Análisis empírico de lainfluencia ejercida por la usabilidad percibida, la satisfacción y la confianzadel consumidor sobre la lealtad a un sitio web [Empirical analysis about theinfluence of user’s percevied usability, satisfaction, and trust on loyalty towardsa website],” in Proceedings of XVI Encuentros de Profesores Universitarios deMarketing, Alicante, 209–226.

Flavián, C., Guinalíu, M., and Gurrea, R. (2005). The role placed by perceivedusability, satisfaction and consumer trust on website loyalty. Inf. Manag. 43,1–14. doi: 10.1016/j.im.2005.01.002

Fortin, D. R., and Dholakia, R. R. (1999). “Examining the effects of interactivityon social presence and involvement with a virtual advertisement. In marketingcommunications and consumer behavior,” in Proceedings of the ThirdInternational Research Seminar, France.

Gallarza, M. G., and Gil, I. (2006). Value dimensions, perceived value, satisfactionand loyalty: an investigation of university students’ travel behavior. Tour.Manag. 27, 437–452. doi: 10.1016/j.tourman.2004.12.002

Gani, M. O., Reza, S. S., Rabi, M. R. I., and Reza, S. S. (2015). “Neuromarketing:methodologies of marketing science,” in Proceedings of The 3rd InternationalConference On Advances in Economics, Management and Social Study, KualaLumpur.

Gazzaniga, M. S., Ivry, R. B., and Mangun, G. R. (2002). The methods of cognitiveneuroscience. Cogn. Neurosci. 2, 96–147.

Goldsmith, R. E., and Emmert, J. (1991). Measuring product category involvement:a multitraitmultimethod study. J. Bus. Res. 23, 363–371. doi: 10.1016/0148-2963(91)90021-O

Gregg, D. G., and Walczak, S. (2010). The relationship between website quality,trust and price premiums at online auctions. Electron. Commer. Res. 10, 1–25.doi: 10.1007/s10660-010-9044-2

Grobelny, J., and Michalski, R. (2015). The role of background color,interletter spacing, and font size on preferences in the digital presentationof a product. Comput. Hum. Behav. 43, 85–100. doi: 10.1016/j.chb.2014.10.036

Hair, J., Anderson, R., Tatham, R., and Black, W. (1999). Multivariant Analysis.Madrid: Prentice-Hall.

Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., and Tatham, R. L. (2006).Multivariate Data Analysis. Upper Saddle River, NJ: Pearson Prentice Hall.

Hernández, B., Jiménez, J., and Martín, M. J. (2010). Customer behavior inelectronic commerce: the moderating effect of e-purchasing experience. J. Bus.Res. 63, 964–971. doi: 10.1016/j.jbusres.2009.01.019

Hirschman, E. C., and Holbrook, M. B. (1982). Hedonic consumption: emergingconcepts, methods and propositions. J. Mark. 46, 92–101. doi: 10.2307/1251707

Hoffman, D. L., and Novak, T. P. (1996). Marketing in hypermedia computer-mediated environments: conceptual foundations. J. Mark. 60, 50–68. doi: 10.2307/1251841

Hofstede, G., Garibaldi de Hilal, A. V., Malvezzi, S., Tanure, B., andVinken, H. (2010). Comparing regional cultures within a country: lessonsfrom Brazil. J. Cross Cult. Psychol. 41, 336–352. doi: 10.1177/0022022109359696

Hong, W., Thong, J. Y. L., and Yan, K. (2004). Does animation attract virtualusers’ attention? The effects of flash on information search performance andperceptions. Infr. Syst. Res. 15, 60–86. doi: 10.1287/isre.1040.0017

Hsu, C.-L., Chang, K.-C., and Chen, M.-C. (2011). The impact of website qualityon customer satisfaction and purchase intention: perceived playfulness andperceived flow as mediators. Infr. Syst. Bus. Manag. 10, 549–570. doi: 10.1007/s10257-011-0181-5

Huang, E. (2013). Interactivity and identification influences on virtual shopping.Int. J. Electron. Commer. Stud. 4, 305–312. doi: 10.7903/ijecs.1123

Huang, M. H. (2006). Flow, enduring, and situational involvement in theweb environment: a tripartite second-order examination. Psychol. Mark. 23,383–411. doi: 10.1002/mar.20118

Iacobucci, D. (1994). “Analysis of experimental data,” in Principles of MarketingResearch, ed. R. Bagozzi (Cambridge, MA: Blackwell), 224–278.

Ivonin, L., Chang, H.-M., Chen, W., and Rauterberg, M. (2013). Unconsciousemotions: quantifying and logging something we are not aware of.Pers. Ubiquitous Comput. 17, 663–673. doi: 10.1007/s00779-012-0514-5

Javor, A., Koller, M., Lee, N., Chamberlain, L., and Ransmayr, G. (2013).Neuromarketing and consumer neuroscience: contributions to neurology. BMCNeurol. 13:13. doi: 10.1186/1471-2377-13-13

Jeong, S. W., Fiore, A. M., Niehm, L. S., and Lorenz, F. O. (2009). The role ofexperiential value in online shopping: the impacts of product presentation onconsumer responses towards an apparel web site. Internet Res. 19, 105–124.doi: 10.1108/10662240910927858

Kapferer, J. N., and Laurent, G. (1985). Consumers’ involvement profile: newempirical results. Adv. Consum. Res. 12, 290–295.

Keng, C. J., Liao, T. H., and Yang, Y. I. (2012). The effects of sequentialcombinations of virtual experience, direct experience, and indirect experience:the moderating roles of need for touch and product involvement. Electron.Commerc. Res. 12, 177–199. doi: 10.1007/s10660-012-9093-9

Khakimdjanova, L., and Park, J. (2005). Online visual merchandising practiceof apparel e-merchants. J. Retail. Consum. Serv. 12, 307–318. doi: 10.1016/j.jretconser.2004.10.005

Kim, J., and Lennon, S. J. (2013). Effects of reputation and website quality ononline consumers’ emotion, perceived risk and purchase intention: based onthe stimulus-organism-response model. J. Res. Interact. Mark. 7, 33–56. doi:10.1108/17505931311316734

Kim, J. B. (2012). An empirical study on consumer first purchase intention inonline shopping: integrating initial trust and TAM. Electron. Commer. Res. 12,125–150. doi: 10.1007/s10660-012-9089-5

Kim, S., and Stoel, L. (2004). Apparel retailers: website quality dimensions andsatisfaction. J. Retail. Consum. Serv. 11, 109–117. doi: 10.1016/S0969-6989(03)00010-9

Koernig, S. (2003). E-Scapes: the electronic physical environment and servicetangibility. Phychol. Mark. 20, 151–167. doi: 10.1002/mar.10065

Koo, D.-M., and Ju, S.-H. (2010). The interactional effects of atmospherics andperceptual curiosity on emotions and online shopping intention. Comput. Hum.Behav. 26, 377–388. doi: 10.1016/j.chb.2009.11.009

Kotler, P. (1973). Atmospherics as a marketing tool. J. Retail. 49, 48–64.Lallemand, C., Gronier, G., and Koenig, V. (2015). User experience: a

concept without consensus? Exploring practitioners’ perspectives through aninternational survey. Comput. Hum. Behav. 43, 35–48. doi: 10.1016/j.chb.2014.10.048

Lian, S. (2011). Innovative internet video consuming based on media analysistechniques. Electron. Commer. Res. 11, 75–89. doi: 10.1007/s10660-010-9066-9

Liu, H., Chu, H., Huang, Q., and Chen, X. (2016). Enhancing the flowexperience of consumers in China through interpersonal interaction insocial commerce. Comput. Hum. Behav. 58, 306–314. doi: 10.1016/j.chb.2016.01.012

Lorenzo, C., Gómez-Borja, M. A., and Mollá-Descals, A. (2007). Website designand e-consumer: effects and responses. Int. J. Internet Mark. Adv. 4, 114–141.doi: 10.1504/IJIMA.2007.014800

Manganari, E. E., Siomkos, G. J., Rigopoulou, I. D., and Vrechopoulos, A. P. (2011).Virtual store layout effects on consumer behaviour: applying an environmentalpsychology approach in the online travel industry. Internet Res. 21, 326–346.doi: 10.1108/10662241111139336

Manganari, E. E., Siomkos, G. J., and Vrechopoulos, A. P. (2009). Store atmospherein web retailing. Eur. J. Mark. 43, 1140–1153. doi: 10.1108/03090560910976401

Martin, B. A., Sherrard, M. J., and Wentzel, D. (2005). The role of sensationseeking and need for cognition on web-site evaluations: a resource-matchingperspective. Psychol. Mark. 22, 109–126. doi: 10.1002/mar.20050

McKinney, L. N. (2004). Creating a satisfying Internet shopping experience viaatmospheric variables. Int. J. Consum. Stud. 28, 268–283. doi: 10.1111/j.1470-6431.2004.00368.x

McMahan, E. A., Ryu, S., and Choi, I. (2014). Lay conceptions of well-being amongundergraduate students from the United States and South Korea: culture-leveldifferences and correlates. Soc. Indic. Res. 119, 321–339. doi: 10.1007/s11205-013-0476-7


https://doi.org/10.1016/0022-4359(94)90037-X

https://doi.org/10.1016/0022-4359(94)90037-X


https://doi.org/10.1016/j.im.2005.01.002

https://doi.org/10.1016/j.tourman.2004.12.002

https://doi.org/10.1016/0148-2963(91)90021-O

https://doi.org/10.1016/0148-2963(91)90021-O

https://doi.org/10.1007/s10660-010-9044-2

https://doi.org/10.1016/j.chb.2014.10.036


https://doi.org/10.1016/j.jbusres.2009.01.019

https://doi.org/10.2307/1251707

https://doi.org/10.2307/1251841

https://doi.org/10.2307/1251841

https://doi.org/10.1177/0022022109359696

https://doi.org/10.1177/0022022109359696

https://doi.org/10.1287/isre.1040.0017

https://doi.org/10.1007/s10257-011-0181-5

https://doi.org/10.1007/s10257-011-0181-5

https://doi.org/10.7903/ijecs.1123


https://doi.org/10.1007/s00779-012-0514-5

https://doi.org/10.1007/s00779-012-0514-5

https://doi.org/10.1186/1471-2377-13-13

https://doi.org/10.1108/10662240910927858

https://doi.org/10.1007/s10660-012-9093-9

https://doi.org/10.1016/j.jretconser.2004.10.005

https://doi.org/10.1016/j.jretconser.2004.10.005

https://doi.org/10.1108/17505931311316734

https://doi.org/10.1108/17505931311316734

https://doi.org/10.1007/s10660-012-9089-5

https://doi.org/10.1016/S0969-6989(03)00010-9

https://doi.org/10.1016/S0969-6989(03)00010-9





https://doi.org/10.1007/s10660-010-9066-9



https://doi.org/10.1504/IJIMA.2007.014800

https://doi.org/10.1108/10662241111139336

https://doi.org/10.1108/03090560910976401

https://doi.org/10.1108/03090560910976401


https://doi.org/10.1111/j.1470-6431.2004.00368.x

https://doi.org/10.1111/j.1470-6431.2004.00368.x

https://doi.org/10.1007/s11205-013-0476-7

https://doi.org/10.1007/s11205-013-0476-7






McMurray, B., Dennhardt, J. L., and Struck-Marcell, A. (2008). Context effectson musical chord categorization: different forms of top-down feedbackin speech and music? Cogn. Sci. 32, 893–920. doi: 10.1080/03640210802222021

Mehrabian, A., and Russell, J. A. (1974). An Approach to Environmental Psychology.Cambridge, MA: MIT Press.

Milliman, R. E. (1986). The influence of background music on the behavior ofrestaurant patron. J. Consum. Res. 13, 286–289. doi: 10.1086/209068

Nath, C. K. (2009). Behaviour of customers in retail store environment- anempirical study. Vilakshan 6, 63–74.

Nesset, E., Nervik, B., and Helgesen, O. (2011). Satisfaction and image as mediatorsof store loyalty drivers in grocery retailing. Int. Rev. Retail Distrib. Consum. Res.21, 267–292. doi: 10.1080/09593969.2011.588716

Nielsen, J. (2000). Designing Web Usability: The Practice of Simplicity. Indianapolis:New Riders.

Oakes, S. (2003). Musical tempo and waiting perceptions. Psychol. Mark. 20,685–705. doi: 10.1002/mar.10092

O’Cass, A., and Fenech, T. (2003). Web retailing adoption: exploring the natureof internet users web retailing behavior. J. Retail. Consum. Serv. 10, 81–94.doi: 10.1016/S0969-6989(02)00004-8

Oliver, R. L. (1997). Satisfaction. A Behavioral Perspective on the Consumer.Singapore: McGraw-Hill.

Ozok, A. A., and Wei, J. (2010). An empirical comparison of consumer usabilitypreferences in online shopping using stationary and mobile devices: resultsfrom a college student population. Electron. Commer. Res. 10, 111–137. doi:10.1007/s10660-010-9048-y

Park, J., and Stoel, L. (2005). Effect of brand familiarity, experience and informationon online apparel purchase. Int. J. Retail Distrib. Manag. 33, 148–160. doi:10.1108/09590550510581476

Ranganathan, C. S., and Ganapathy, S. (2002). Key dimensions of business toconsumer web sites. Infr. Manag. 39, 457–465. doi: 10.1016/S0378-7206(01)00112-4

Schlosser, A. E. (2003). Experiencing products in the virtual world: the role of goaland imagery in influencing attitudes versus purchase intentions. J. Consum. Res.30, 184–198. doi: 10.1086/376807

Schneiderman, B. (1998). Designing the user interface. Reading, MA: Addison–Wesley.

Shankar, V., Smith, A., and Rangaswamy, A. (2003). Customer satisfaction andloyaty in online and offline environments. Int. J. Res. Mark. 20, 153–175. doi:10.1016/S0167-8116(03)00016-8

Sherman, E., Mathur, A., and Smith, R. B. (1997). Store environment and consumerpurchase behavior: mediating role of consumer emotions. Psychol. Mark.14, 371–378. doi: 10.1002/(SICI)1520-6793(199707)14:4<361::AID-MAR4>3.0.CO;2-7

Slovic, P., Fischhoff, B., and Lichtenstein, S. (1980). Facts and Fears: UnderstandingPerceived Risk. Societal Risk Assessment: How Safe is Safe Enough? New York,NY: Plenum.

Smith, D. N., and Sivakumar, K. (2004). Flow and Internet shopping behavior.A conceptual model and research propositions. J. Bus. Res. 57, 1199–1208.doi: 10.1016/j.fsc.2010.06.003

Solaymani, S., Sohaili, K., and Yazdinejad, E. A. (2012). Adoption and use ofe-commerce in SMEs. A case study. Electron. Commer. Res. 1, 249–263. doi:10.1007/s10660-012-9096-6

Sundar, S. S., and Kalyanaraman, S. (2004). Arousal, memory and impressionformation effects of animation speed in web advertising. J. Adv. 33, 7–17.doi: 10.1080/00913367.2004.10639152

Titus, P. A., and Everett, P. B. (1995). The consumer retail search process: aconceptual mode and research agenda. J. Acad. Mark. Sci. 23, 106–119. doi:10.1177/0092070395232003

Vanhamme, J. (2000). The link between surprise and satisfaction: an exploratoryresearch. J. Mark. Manag. 16, 565–582. doi: 10.1362/026725700785045949

Vrechopoulos, A. P., Doukidis, G. I., and O’Keefe, R. M. (2000). “Virtual storeatmosphere in Internet retailing,” in Proceedings of the 13th International BledElectronic Commerce Conference, Bled.

Wang, Y. J., Hernandez, M. D., and Minor, M. S. (2010). Web aesthetics effectson perceived online service quality and satisfaction in an e-tail environment:the moderating role of purchase task. J. Bus. Res. 63, 935–942. doi: 10.1016/j.jbusres.2009.01.016

Yoh, E., Damhorst, M. L., Sapp, S., and Laczniac, R. (2003). Consumer adoptionof the internet: the case of apparel shopping. Psychol. Mark. 20, 1095–1118.doi: 10.1002/mar.10110

Zaichkowsky, J. L. (1986). Conceptualizing involvement. J. Adv. 15, 4–14. doi:10.1080/00913367.1986.10672999

Zeithaml, V. A., Berry, L. L., and Parasuraman, A. (1996). The behavioralconsequences of service quality. J. Mark. 60, 31–46. doi: 10.2307/1251929

Conflict of Interest Statement: The authors declare that the research wasconducted in the absence of any commercial or financial relationships that couldbe construed as a potential conflict of interest.

Copyright © 2016 Lorenzo-Romero, Alarcón-del-Amo and Gómez-Borja. This is anopen-access article distributed under the terms of the Creative Commons AttributionLicense (CC BY). The use, distribution or reproduction in other forums is permitted,provided the original author(s) or licensor are credited and that the originalpublication in this journal is cited, in accordance with accepted academic practice.No use, distribution or reproduction is permitted which does not comply with theseterms.


https://doi.org/10.1080/03640210802222021

https://doi.org/10.1080/03640210802222021

https://doi.org/10.1086/209068

https://doi.org/10.1080/09593969.2011.588716


https://doi.org/10.1016/S0969-6989(02)00004-8

https://doi.org/10.1007/s10660-010-9048-y

https://doi.org/10.1007/s10660-010-9048-y

https://doi.org/10.1108/09590550510581476

https://doi.org/10.1108/09590550510581476

https://doi.org/10.1016/S0378-7206(01)00112-4

https://doi.org/10.1016/S0378-7206(01)00112-4

https://doi.org/10.1086/376807

https://doi.org/10.1016/S0167-8116(03)00016-8

https://doi.org/10.1016/S0167-8116(03)00016-8

https://doi.org/10.1002/(SICI)1520-6793(199707)14:4<361::AID-MAR4>3.0.CO;2-7

https://doi.org/10.1002/(SICI)1520-6793(199707)14:4<361::AID-MAR4>3.0.CO;2-7

https://doi.org/10.1016/j.fsc.2010.06.003

https://doi.org/10.1007/s10660-012-9096-6

https://doi.org/10.1007/s10660-012-9096-6

https://doi.org/10.1080/00913367.2004.10639152

https://doi.org/10.1177/0092070395232003

https://doi.org/10.1177/0092070395232003

https://doi.org/10.1362/026725700785045949

https://doi.org/10.1362/026725700785045949




https://doi.org/10.1080/00913367.1986.10672999

https://doi.org/10.1080/00913367.1986.10672999

https://doi.org/10.2307/1251929









Analyzing the User Behavior toward Electronic Commerce … · · 2017-04-13S-O-R model) refer to...

Documents

Transcript of Analyzing the User Behavior toward Electronic Commerce … · · 2017-04-13S-O-R model) refer to...