- Email: [email protected]
Discernible impact of augmented reality on retail customer's experience, satisfaction and willingness to buy
Journal of Retailing and Consumer Services 34 (2017) 229–234
Contents lists available at ScienceDirect
Journal of Retailing and Consumer Services journal homepage: www.elsevier.com/locate/jretconser
Discernible impact of augmented reality on retail customer's experience, satisfaction and willingness to buy
crossmark
⁎
Atieh Poushneh, Arturo Z. Vasquez-Parraga
College of Business and Entrepreneurship, University of Texas Rio Grande Valley, 1201 W. University Drive, Edinburg, TX 78539, USA
A R T I C L E I N F O
A BS T RAC T
Keywords: Augmented reality Retail user experience User satisfaction User's Willingness to buy
User technology has an agreeable impact on consumer decisions; yet the way such impact takes place may be little known. This study attempts to examine the impact of augmented reality (AR) on retail user experience (UX) and its subsequent influence on user satisfaction and user's willingness to buy. Five hypotheses are tested using a lab experiment. The results show that AR significantly shapes UX, by impinging on various characteristics of product quality, and that UX subsequently influences user satisfaction and user's willingness to buy. UX is captured as a third-order formative construct derived from four user experience characteristics: pragmatic quality, aesthetic quality, hedonic quality by stimulation and hedonic quality by identification. Except for the latter, these characteristics are second-order constructs. Important implications for researchers and managers follow.
1. Introduction Customers intending to buy a toy walk into a store. An unassembled 3D puzzle catches their eye, but they are not quite sure what the final assembly will look like. Then, they are told about augmented reality, a collection of viewing features that helps customers visualize the assembled toy in three dimensions (3D), which enables them to observe the puzzle from every angle. This example illustrates how AR helps customers/users make purchase decisions. The literature on AR has emphasized the technological aspects of AR, but it has neglected the end user's needs and problems (Swan and Gabbard, 2005). Yet, AR is increasingly employed in designing and delivering products, even though research has not been able to catch up with the trend from a marketing perspective (Kozick and Gettliffe, 2010; Swan and Gabbard, 2005), especially the growing impact of AR on user experience (UX). This study attempts to understand the way AR influences UX and, at the outset, user satisfaction (US) and user willingness to buy (UWB). Although prior literature has studied some UX dimensions, no mutual agreement about measuring UX has been reached (Vermeeren et al., 2010). Earlier UX studies focused on such cognitive dimensions of UX as usability (e.g., Butler, 1996), but they have ignored UX's affective dimensions. To correct such narrow focus, a user-centered design (UCD) that involved users in the design process emerged (Karat, 1996) and embraced the cognitive and affective dimensions of UX (Alben, 1996). Thus, this study attempts to measure a unified measure of UX and answer the following research questions.
⁎
RQ1: How does augmented reality improve retail user experience? Why is it important that augmented reality enhance the user experience? RQ2: Which and how do key factors moderate the relationship between augmented reality and expected retail user experience, if any? RQ3: What are the effects of retail user experiences on two main consumer outcomes, user satisfaction and user's willingness to buy? The remainder of this paper is organized as follows: first, a brief literature review of the main concepts used and relationships in the study will be discussed. Next, hypotheses, methodology, and results will be explained. Finally, conclusions, managerial implications, limitations, and proposals for future research will be discussed. 2. Conceptual framework 2.1. Augmented Reality (AR) AR is a series of technologies that integrate real world and virtual information, thereby enhancing a specific reality (Lamantia, 2009). Some customers do not make online purchase because such deficiencies make the process risky (Kim and Forsythe, 2008a). AR can produce meaningful experiences for online shoppers (MacIntyre et al., 2001) by providing sufficient product information (Lu and Smith, 2007) that enables them to evaluate the targeted products
Corresponding author. E-mail addresses: [email protected] (A. Poushneh), [email protected] (A.Z. Vasquez-Parraga).
http://dx.doi.org/10.1016/j.jretconser.2016.10.005 Received 18 September 2016; Accepted 6 October 2016 0969-6989/ © 2016 Elsevier Ltd. All rights reserved.
Journal of Retailing and Consumer Services 34 (2017) 229–234
A. Poushneh, A.Z. Vasquez-Parraga
3. Relationships and hypotheses
(Kim and Forsythe, 2008a) and make decisions with more certainty (Oh et al., 2008).
3.1. AR effect on UX as reflected in product pragmatic quality (PQ) PQ is also called usability when it relates to the effectiveness, efficiency, and satisfaction of the UX (Butler, 1996). Since the usability aspect of UX covers a narrow scope of UX, it is not examined as a criterion to evaluate UX (Norman, 2004). The many features of a product, including usability, functions, size, weight, symbols, aesthetic, and usefulness may influence UX. PQ involves a portion of those interactions that emphasize the utility and usability of a product in relation to its potential tasks (Hassenzahl et al., 2003). AR enhances UX by revealing more product information than products without AR, which results in higher UX at the time of purchase, reduces users’ anxiety (Huang and Hsu-Liu, 2014), and facilitates decision-making (Kim and Forsythe, 2008a, 2008b).
2.1.1. AR is reflected in the level of interactivity AR is a stimulus in this study, and the level of interactivity was chosen to reflect AR. Interactivity refers to the “extent to which users can participate in modifying the form and content of a mediated environment in real time” (Steuer, 1992, p. 84). Interactivity entertains users and enables them to personalize information in a 3D virtual model (Fiore, Kim and Lee, 2005), and they enjoy interacting with virtual objects more than they do handling or looking at physical objects (Li et al., 2001). In this study, three levels of interactivity are examined: high, middle, and low. It is assumed that a high level of interactivity will generate a greater UX and subsequently higher user satisfaction and user willingness to buy. Conversely, a low level of interactivity will generate weaker UX and subsequently weaker user satisfaction and user willingness to buy. In this study, high and the middle level of interactivity were examined as AR treatments, and low level of interactivity was examined as non-AR treatment.
3.2. AR effect on UX as reflected in product hedonic quality (HQ) PQ is essential to UX, but it does not exhaust UX. UX also involves emotional reactions (Hassenzahl and Tractinsky, 2006; Norman, 2004). Consequently, AR may influence UX as well by affecting HQ and thus facilitating several emotional benefits. AR facilitates user involvement and thereby enhances the hedonic value of experience (Kim and Forsythe, 2008b), which provides users the ability to share personalized experiences on social networks, thus enhancing playfulness (Huang and Hsu-Liu, 2014). Yet, the effect of AR on HQ can vary depending on whether the experience reflects enjoyment or social reference. Hassenzahl et al. (2003) distinguish three types of effects in HQ: effects by stimulation (HQ-S), effects by identification (HQ-I), and effects by evocation (HQ-E). HQ-S is related to the fulfillment of human needs for novelty and challenge. HQ-I refers to the fulfillment of human needs as self-expressions. HQ-E refers to the human fulfillment needs for symbolic meanings of an object.
2.1.2. Retail user experience (UX) UX is holistic and subjective (McCarthy and Wright, 2004), and varies across time (Law et al., 2009). It is also defined as: “All the aspects of how people use an interactive product: the way it feels in their hands, how well they understand how it works, how they feel about it while they are using it, how well it serves their purposes, and how well it fits into the entire context in which they are using it” (Alben, 1996, p. 5). UX is a complex construct that encompasses a user's inner state, product characteristics, and the context of use (Hassenzahl and Tractinsky, 2006). Product attributes include pragmatic quality (PQ), aesthetic quality (AQ), and hedonic quality (HQ). To measure UX, this study focuses on these product attributes (See Fig. 1)
3.3. AR effect on UX as reflected in product aesthetic quality (AQ) The AQ of UX involves pleasurable experiences. Jordan (2002)
Fig. 1. Conceptual framework: The impact of augmented reality on user experience and its outcomes.
230
Journal of Retailing and Consumer Services 34 (2017) 229–234
A. Poushneh, A.Z. Vasquez-Parraga
Table 1 (continued)
Table 1 First-order constructs: EFA results. Constructs
Hedonic Quality by Identification (α=.848; AVE=.44) Unpresentable-presentable Separates me from people-bring me closer to people Alienating-integrating Cheap-expensive Tacky-stylish Isolating-connective Decreases one's self image-augments one's self-image Loneliness-the sense of belonging to the community
.773 .510 .786 .372 .807 .820 .549 .542
Hedonic Quality by Emotional Stimulation (α=.882; AVE=.54) Repelling-appealing Discouraging-motivating Not absorbed-over absorbed Not immerse-immerse
.930 .712 .454 .759
Hedonic Quality by Rational Stimulation (α=.922; AVE=.669) Ordinary-novel Dull-captivating Conservative-innovative Cautious-bold Unimaginative-creative Conventional-inventive
.720 .828 .809 .749 .868 .918
Aesthetic Quality - Cognitive (α=.955; AVE=.729) Ugly-beautiful Unattractive-attractive Asymmetric-symmetric Unclean-clean Aesthetically unpleasing-aesthetically pleasing Rigid design-artistic design Static-vivid Artificial-Realistic
.926 .930 .711 .797 .833 .892 .904 .816
Aesthetic Quality - Affective (α=.894; AVE=.753) Unfriendly-friendly Annoying-enjoyable Unpleasant-pleasant
.722 .882 .980
Pragmatic Quality - Practical (α=.902; AVE=.61) Unruly-manageable Confusing-clearly structured Impractical-practical Complicated-simple Difficult to learn-easy to learn Effortful-effortless
.583 .853 .911 .834 .821 .652
Pragmatic Quality - Reliable (α=.864; AVE=.52) Unpredictable-predictable Cumbersome-straightforward Unprofessional-professional Insecure-secure Irrelevant-relevant Unreliable-reliable
.750 .868 .611 .641 .619 .809
Pragmatic Quality - Informative (α=.733; AVE=.42) Too few information-too much information Shady-trustworthy Not personalized-personalized Highly decreases one's awareness-highly augments one's awareness Pragmatic Quality - Useful (α=.974; AVE=.81) Highly decrease one's capabilities-highly augments one's capabilities Risky to use-safe to use User Satisfaction (α=.938; AVE=.836) Overall, I am satisfied with the Ray-Ban website. Being a user of this website has been a satisfying experience.
Constructs
Factor loadings
Having experienced this website was pleasurable.
.890
User's Willingness to Buy (α=.954; AVE=.874) I intend to buy my eyeglasses/sunglasses via the Ray-Ban website.
.911
Factor loadings
I would be willing to buy my eyeglasses/sunglasses via the Ray-Ban website. In future, I would buy my eyeglasses/sunglasses via the RayBan website. Trade-off Price and Value (α=.905; AVE=.779) The product offered in the website of Ray-Ban app is reasonably priced. The product offered in the website of Ray-Ban is a good value for the money. At the current price, the product offered in the website of Ray-Ban provides a good value. User's information privacy Control (α=.956; AVE=.84) I was informed about the personal information that Ray-Ban website would collect about me, such as email, name, and location. This website explained the reasons why my personal information is being collected. This website informed the way my personal information would be used. This website gave me a clear choice before using personal information about me.
.929
.726 .982 .920
.934
.979 .915 .838
Note: The following items were eliminated: PQ1 (human-technical), HS2 (challengingundemanding), and ASC6 (bad-good), as explained in the text.
identifies four types of pleasure. Physio-pleasure is related to the sensual UX (e.g. touch, smell, taste). Socio- pleasure is related to the relationship of the user with others (e.g. status, connection). Psychopleasure is related to people's cognitive and emotional reactions (e.g. satisfaction of instrumental needs), and ideo-pleasure is related to the values of people (e.g. aesthetics, taste, personal aspirations). It is possible that not all types of pleasure are sought at once, but some are. Thus, H1. Augmented reality positively and significantly impacts user experience as reflected in four product characteristics: pragmatic quality, aesthetic quality, hedonic quality by stimulation, and hedonic quality by identification. 3.4. Moderating Variables of the ARUX Relationship: Trade-off between Price and Value (PV) and User’s Information Privacy Control (UIPC) Dodds et al. (1991) explain that perceived value functions as a trade-off between perceived quality and monetary sacrifice. Consistent with Equity Theory (Adams et al., 1976), shoppers compare what they gain against what lose when they buy a product. UX may be strengthened if users find the trade-off between price and value to be acceptable, and it may be weakened if users do not consider the tradeoff between price and value acceptable. Thus,
.736 .602 .529 .694
H2. The relationship between AR and UX is stronger when users find a proper fitting trade-off between value and price, and it is weaker when users do not find a fitting trade-off between value and price. Additionally, this study hypothesizes that user's information privacy control (UIPC) moderates the impact of AR on UX. UIPC refers to individuals’ ability to control their personal information (Metzger, 2004), and AR has the potential to collect and personalize user information. Users are sensitive about their privacy and expect to have
.900 .900
.898 .954 (continued on next page)
231
Journal of Retailing and Consumer Services 34 (2017) 229–234
A. Poushneh, A.Z. Vasquez-Parraga
Table 2 Constructs, path coefficients, R-squared, and p-values. Construct Order
Constructs
Mean
Standard deviation
Path coefficients
Beta
R-squared
P-value
Third-order First-order First-order
UX US UWB
.842 .825 .676
.034 .032 .057
AR→UX UX→User Satisfaction UX→User's Willingness to Buy
.30 .87 .76
.867 .760 .580
.003 .000 .000
Fig. 2. Results of SmartPLS: path coefficients, R-squared, p-values.
4. Methodology
access to AR applications that do not require them to share personal information (Olsson et al., 2013). Consistent with Equity Theory (Adams et al., 1976), the amount, exactness, and sensitivity of information that the user is willing to share with a service rely on the extent of value gained from the feature that requires user information (Olsson et al., 2013). If users have control over their personal information while using AR, then they may be genuinely motivated to engage in an experience using AR and vice versa. Thus,
The study was conducted in a laboratory environment located at a large South West University in the US. It utilized a true experiment (Kerlinger and Lee, 2000) in which participants were randomly assigned to three treatments. AR was considered as a stimulus in terms of two levels of interactivity, namely high and middle. For the experimental groups (high and middle level of interactivity), two AR treatments, Ray-Ban sunglasses and Virtual Model were examined. For the control group (very low level of interactivity), traditional online shopping was examined. Quantifying UX is important because it provides a guideline for product designers to choose effective design strategies (Law and Van Schaik, 2010). To measure PQ and HQ, an AttrakDiff 2 questionnaire from Hassenzahl et al. (2003) was used. Since AR is interactive, simulated sensory experience, it was necessary to add more properties to reflect AR. To measure PQ, eight items were taken from Haasenzahl et al. (2003) and one item from Laugwitz et al., (2008). Ten new items were added. To measure HQ-I, six items were taken from Hassenzahl and two new items were added. To measure HQ-S, nine items were taken from Hassenzahl et al., (2003) and two new items were added. And to measure AQ, three items were taken from Lavie and Tractinsky (2004), five items from Laugwitz et al., (2008), and four new items were added. Overall, all UX items were measured using a bipolar semantic differential 7-scale method. To measure user satisfaction (US), items were taken from Taylor and Baker (1994); to measure UWB, three items from Engel et al. (1995); to measure PV, three items were taken from (Dodds et al., 1991); and to measure UIPC, four items were taken from Liu et al. (2004). These items were measured using a 7-point Likert scale with the anchors being “strongly disagree” and “strongly agree.”
H3. The relationship between AR and UX is strengthened or weakened when the user's information privacy control is either empowered or diminished by AR.
3.5. Outcome variables: user satisfaction (US) and user's willingness to buy (UWB) Customer “satisfaction is not [only] the pleasurableness of the [consumption] experience, it is the evaluation rendered that the experience was at least as good as it was supposed to be” (Hunt, 1977, p. 459). AR generates US through experiential value (Yuan and Wu, 2008). Additionally AR is also able to influence US before the buying process (Bulearca and Tamarjan, 2010). Thus, H4. An AR-enriched user experience positively and significantly impacts user satisfaction. User Willingness to Buy (UWB) refers to consumers’ tendency to purchase targeted products in the future and may predict actual purchase behavior (Morrison, 1979). UWB is also impacted by an AR-enriched UX. AR, such as virtual image technology (Verton, 2001) can offer a simulated experience to users with the purpose of encouraging them to buy the product (Huang and Hsu-Liu, 2014). Virtual objects and the information contributed by AR may heighten user's enjoyment and mental imagery (Schlosser, 2003), which in turn may stimulate UWB (Huang and Hsu-Liu, 2014; Kim and Forsythe, 2008a). Thus,
4.1. Participants, prescreening questions and procedures Purposive sampling was used to recruit the participants. The sample for the study consisted of 99 mostly young consumers (45 male, and 54 female) at a U.S. Southern Metropolitan City. The age of the consumers ranged from 20 to 60. All questionnaires were used in the analyses. Missing data was minor and did not exert any undue effect. Before conducting the experiment, the researcher told partici-
H5. An AR-enriched user experience positively and significantly impacts user's willingness to buy. 232
Journal of Retailing and Consumer Services 34 (2017) 229–234
A. Poushneh, A.Z. Vasquez-Parraga
pants that “this study is trying to evaluate the impact of AR on UX and its outcomes.” Before exposing them to the treatments, they answered prescreening questions that asked about their technology use and online shopping (Jin, 2001; Olsson et al., 2013), including the extent to which they like to buy eyeglasses/sunglasses online. The prescreening questions’ Mean (M) and standard deviations (SD) are as follow. “I am familiar with using the Internet” (M=6.9; SD=.35); “I frequently use the Internet to shop online” (M=6.5; SD=.9); “I think that technology is necessary for my daily works” (M=6.6; SD=.57); “I visit the Internet retail websites to collect product information” (M=5.6; SD=1.7); “I visit the Internet retail websites for purchasing products” (M=5.3; SD=1.6); “I am a user of eyeglasses/sunglasses” (M=5.8; SD=1.4); “I would like to wear eyeglasses/sunglasses” (M=6.6; SD=.7). A few studies have considered the level of interactivity as a stimulus (e.g., Lee et al., 2006). This research includes manipulation check questions. Each question is measured using a 7-point Likert scale from 1=strongly disagree to 7=strongly agree. The questions are “The website provides a variety of ways for viewing product image,” “The website provides personalized product,” “The website is interactive,” “The website allows the user to interact with the products shown on the screen,” “The website allows the user to adjust the product,” and “The website shows dynamic product images.” All participants in both the experimental and control groups were asked to answer the prescreening questions, and then the researcher explained the purpose of study to them, which was to understand the impact of augmented reality on user experience. For the high level of interactivity stimulus, the researcher asked the participants to log in to their computers and enter “http://www.rayban.com/international/ virtual-mirror”. They could personalize, zoom in, zoom out on their favorite eyeglasses and sunglasses, and they could take a photo with the virtual product and share it on their social networks. For the middle level of interactivity, the experimenter asked the participants to enter http://www.ray-ban.com/international/virtual-mirror. This virtual model does not have all the features included in high level of interactivity of AR described above. The virtual model includes two models for men and women. For the low level of interactivity (control group), the researcher asked the participants to enter “http://www.rayban.com/usa.” This treatment did not have novel features. After ten minutes, the researcher asked all participants to log off their computers and then answer the survey.
threshold value of .5 for convergent validity (McDonald and Ringo-Ho, 2002). AVEs above .5 and square roots of AVEs above inter-factor correlations show discriminant validity (Fornell and Larcker, 1981). All constructs had discriminant validity. The data obtained from the high and middle level of interactivity of AR was considered AR data, whereas the data obtained from the low level of interactivity was considered non-AR data. The structural model was tested using SmartPLS 3.0 because SmartPLS works well with small sample sizes, and it is appropriate for both reflective and formative constructs. Moreover, it also works well for the aims of exploratory or theory development research (Hair et al., 2012). Measuring UX is challenging, and no studies as yet have attempted to measure a unified UX (Law and Van Schaik, 2010). In this study, UX was measured as a formative third-order construct because the direction of causality is from the indicators toward the construct. To validate the formative nature of the UX, the weights of the formative construct's indicators were checked, and only indicators with significant weights were retained (Hair et al., 2012). The weights of PQ, HQS, HQ-I, and AQ were significant; therefore, they were also retained. The results of SmartPLS 3.0 demonstrated that the AR condition is positively and significantly associated with UX (β=.3; R2=.867; t=2.37; p=.003), indicating that H1was supported. Regarding the outcome variables, the results showed that H4 and H5 are also supported, that is, UX is positively and significantly associated with US (β=.873; R2=.763; t=49.90; p=.000) and UWB (β=.761; R2=.58; t=21.89; p=.000). Table 2 shows the constructs, path coefficients, R squared coefficients, and p-values associated with H1, H4, and H5. To test the impact of the moderators of the AR-UX relationship, SmartPLS was applied. The results showed that PV (β=.85, p=.45, t=1.46) and UIPC did not moderate the impact of AR on UX (β=.032, p=.2974, t=1.055). To test the mediation impact of UX on US, and UWB, SmartPLS 3.0 by 5000 bootstrap was used. The results indicated that UX partially mediates the relationship between AR and UWB (R2=.058, β=.246, p=.008, t=2.655) and that UX is only a partial mediator of the relationship between AR and US (R2=.183, β=.427, p=.000, t=6.667). Fig. 2 shows the path coefficients, R-square, and pvalues.
5. Analysis and results
6. Discussion and conclusion
SPSS was conducted to obtain descriptive statistics and reliability results. Table 1 shows the results of reliability, AVE results, and factor loadings. Cronbach Alphas range from .733 to .991, thus demonstrating construct internal consistencies (Nunnally and Bernstein, 1994), and factor loadings range from .372 to .982. After checking reliability, exploratory factor analysis (EFA) with the sixty-three items using the maximum likelihood method (MLE) and Varimax rotation was conducted to check the uni-dimensionality of constructs. Thirteen factors emerged with acceptable default eigenvalues. All factor loadings were higher than the minimal level, .3 (Hair et al., 2006). Overall, four factors from PQ emerged: practical, informative, reliable, and useful. Additionally, two factors emerged from HQ-S, rational stimulation and emotional stimulation. Two factors emerged from AQ, cognitive aesthetic and affective aesthetic. Other constructs, HQ-I, US, UWB, UIPC, and PV were uni-dimension. After running a reliability test and EFA, three items with low alpha levels, PQ1 (human-technical), HS2 (challenging-undemanding), and ASC6 (bad-good), were eliminated. The final EFA using the sixty retained items confirmed the dimensionality of the scales and returned theoretically and empirically acceptable solutions (see Table 1). To check convergent and discriminant validity, the average variances extracted (AVE) and composite reliabilities (CR) were checked. The AVEs ranged from .59 to .87, thus satisfying the recommended
The study investigated the impact of AR on UX and its subsequent effects on US and UWB. AR significantly and positively influences UX (H1). AR provides more 3D product information, in different colors and styles, which enhances users’ perception of reality. AR also provides the user with enriched product information gained from a physical store as well as online store. For example, AR allows the user to simulate the product's features on a website in online shopping (Fiore et al., 2005). In addition, AR empowers users to share their personalized experiences on social networks, which enhances playfulness (Huang and Hsu-Liu, 2014). In relation to the impact of an AR-enriched UX on US, the results indicated that an AR-enriched UX empowers users to better perform their tasks and appreciate the functionality of the product more. An AR-enriched UX is more entertaining and enables potential customers to have endless interaction with virtual information. Thus, AR-enriched UX produces higher user satisfaction (H4) and user willingness to buy (H5). AR can enhance hedonic values, leading to an increase in UWB (Huang and Hsu-Liu, 2014) by adding virtual information to real information and offering 3D pictures of products in different shapes, colors, and styles. Further, UX partially mediated the impact of AR on US and UWB. In relation to the moderator effects, the results indicated that PV and UIPC did not moderate the impact of AR on UX (H2 and H3). 233
Journal of Retailing and Consumer Services 34 (2017) 229–234
A. Poushneh, A.Z. Vasquez-Parraga
Hassenzahl, M., Burmester, M. Koller, F., 2003. AttrakDiff: Ein fragebogen zur messung wahrgenommener hedonischer und pragmatischer Qualität, Mensch & Computer, Vieweg Teubner Verlag, pp. 187–196 Huang, T.-L., Hsu-liu, F., 2014. Formation of augmented reality interactive technology’s persuasive effects from the perspective of experiential value. Internet Res. 24 (1), 82–109. Conceptualization and measurement of consumer satisfaction and dissatisfaction. In: Hunt, H.K. (Ed.), . Marketing Science Institute, Cambridge, MA, 77–103. Jin, H.-J., 2001. Effects of Interactive Web Site Features on Approach Responses Towards An Online Apparel Retailer. Iowa State University, Ames, IA. Jordan, P.W., 2002. Human factors for pleasure seekers, Design and the Social Sciences: Making Connections, 2(9) Karat, J., 1996. User centered design: quality or quackery? Interactions 3 (4), 18–20. Kim, J., Forsythe, S., 2008a. Adoption of virtual try-on technology for online apparel shopping. J. Interact. Mark. 22 (2), 45–59. Kim, J., Forsythe, S., 2008b. Sensory enabling technology acceptance model (SE-TAM): a multiple-group structural model comparison. Psychol. Mark. 25 (9), 901–922. LamantiaJ., 2009. Inside out: interaction design for augmented reality. UX Matters, August 17. Laugwitz, B., Held, T., Schrepp, M., 2008. Construction and Evaluation of a User Experience Questionnaire. Springer, Berlin Heidelberg, 63–76. Lavie, T., Tractinsky, N., 2004. Assessing dimensions of perceived visual aesthetics of web sites. Int. J. Hum.-Comput. Stud. 60 (3), 269–298. Law, E., L.-C., Roto, V., Hassenzahl, M., Vermeeren, A.P., Kort, J., 2009. Understanding, scoping and defining user experience: a survey approach. In: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. ACM, pp. 719–728. Law, E.L.-C., Van Schaik, P., 2010. Modelling user experience–an agenda for research and practice. Interact. Comput. 22 (5), 313–322. Lee, H.-H., Fiore, A.M., Kim, J., 2006. The role of the technology acceptance model in explaining effects of image interactivity technology on consumer responses. Int. J. Retail Distrib. Manag. 34 (8), 621–644. Li, H., Daugherty, T., Biocca, F., 2001. Characteristics of virtual experience in electronic commerce: a protocol analysis. J. Interact. Mark. 15 (3), 13–30. Liu, C., Marchewka, J.T., Ku, C., 2004. American and Taiwanese perceptions concerning privacy, trust, and behavioral intentions in electronic commerce. J. Glob. Inf. Manag. 12 (1), 18–40. MacIntyre, B., Bolter, J.D., Moreno, E., Hannigan, B., 2001. Augmented reality as a new media experience. In: Augmented Reality, Proceedings: IEEE and ACM International Symposium on, pp. 197–206. McCarthy, J., Wright, P., 2004. Technology as experience. Interactions 11 (5), 42–43. McDonald, R.P., Ringo-Ho, M.-H, 2002. Principles and practice in reporting structural equation analyses. Psychol. Methods 7 (1), 64. Morrison, D.G., 1979. Purchase intentions and purchase behavior. J. Mark., 65–74. Norman, D.A., 2004. Emotional Design: Why We Love (or Hate) Everyday Things. Basic Books, New York. Nunnally, J.C., Bernstein, I.-H., 1994. The assessment of reliability. Psychom. Theory 3 (1), 248–292. Oh, H., Yoon, S.-Y., Shyu, C.-R., 2008. How can virtual reality reshape furniture retailing? Cloth. Text. Res. J. 26 (2), 143–163. Olsson, T., et al., 2013. Expected user experience of mobile augmented reality services: a user study in the context of shopping centers. Pers. Ubiquitous Comput. 17 (2), 287–304. Schlosser, A.E., 2003. Experiencing products in the virtual world: the role of goal and imagery in influencing attitudes versus purchase intentions. J. Consum. Res 30 (2), 184–198. Steuer, J., 1992. Defining virtual reality: dimensions determining telepresence. J. Commun. 42 (4), 73–93. Swan, J.E., Gabbard, J.L., 2005. Survey of user-based experimentation in augmented reality. In: Proceedings of the 1st International Conference on Virtual Reality. Taylor, S.A., Baker, T.L., 1994. An assessment of the relationship between service quality and customer satisfaction in the formation of consumers' purchase intentions. J. Retail. 70 (2), 163–178. Vermeeren, A. P. et al., 2010. In: User experience evaluation methods: Current state and development needs. In:Proceedings of the 6th Nordic Conference on HumanComputer Interaction: Extending Boundaries. ACM, pp. 521–530. Verton, D., 2001. Churn. Computer World. 35(6), pp. 50 Yuan, Y.-H., Wu, K.W., 2008. Relationships among experiential marketing, experiential value, and customer satisfaction. J. Hosp. Tour. Res 32 (3), 387–410.
7. Managerial implications, limitations, and future research AR has been developed as a usable and artistic tool to draw users’ attention. The number and quality of AR applications have been increasing, and it seems that only recently AR developers are paying attention to user's actual needs and desires. AR should be developed to satisfy the various characteristics of UX, in particular PQ, HQ-S, HQ-I, and AQ. A well-designed AR should be practical, easy to use, easy to learn, organized, symmetric, attractive, and pleasant if it is to provide effective and relevant information to users, and it should contemplate the emotional aspects of UX. AR should be novel and pleasant so that it will empower and encourage users to express themselves on social media. Recognizing the power of AR in UX and its mediated role in customer's satisfaction and purchase intention, retailers and AR companies can collaborate to develop marketing strategies that effectively enrich and enhance customers’ shopping experience. For instance, retailers can implement promotions that enhance the participation of customers using AR apps such as Blippar apps that are used in smartphones to reveal surrounding objects by just pointing the smartphone to the goods. Bountiful virtual information that relate to real products will appear on the smart phone's screen as if the viewer would be in a fantasy world. As a result, customers will be motivated to download AR apps that enhance reality along with promotions. In turn, retailer can make more sales and AR companies receive additional orders for the technology. This study has some limitations. First, this study did not include the impact of such other variables as culture and ethnicity. Additionally, only two consequences of UX were considered in this research. For future research, hedonic quality by evocation may be included as another product characteristic of UX. Moreover, it would be useful and illuminating to evaluate the impact of AR in other user contexts, such as health care. References Adams, J.S., Berkowitz, L., Walster, E., 1976. Advances in Experimental Social Psychology, Equity Theory: Toward a General Theory of Social Interaction. Academic Press, 9. Alben, L., 1996. Defining the criteria for effective interaction design. Interactions 3 (3), 11–15. Bulearca, M., Tamarjan, D., 2010. Augmented reality: a sustainable marketing tool? Glob. Bus. Manag. Res.: Int. J. 2 (2–3), 237–252. Butler, K.A., 1996. Usability engineering turns 10. Interactions 3 (1), 58–75. Dodds, W.B., Monroe, K.B., Grewal, D., 1991. Effects of price, brand, and store information on buyers' product evaluations. J. Mark. Res., 307–319. Engel, J.F., Blackwell, R.D., Miniard, P.W., 1995. Consumer Behavior. Dryder, New York. Fiore, A.M., Kim, J., Lee, H.-H., 2005. Effect of image interactivity technology on consumer responses toward the online retailer. J. Interact. Mark. 19 (3), 38–53. Fornell, C., Larcker, D.F., 1981. Evaluating structural equation models with unobservable variables and measurement error. J. Mark. Res., 39–50. Hair, J.F., Black, W.C., Babin, B.J., Anderson, R.E., Tatham, R.L., 2006. Multivariate Data Analysis 6. Pearson Prentice Hall, Upper Saddle River, NJ. Hair, J.F., Sarstedt, M., Ringle, C.M., Mena, J.A., 2012. An assessment of the use of partial least squares structural equation modeling in marketing research. J. Acad. Mark. Sci. 40 (3), 414–433. Hassenzahl, M., Tractinsky, N., 2006. User experience – a research agenda (Editorial). Behav. Info Tech. 25 (2), 91–97.
234
Contents lists available at ScienceDirect
Journal of Retailing and Consumer Services journal homepage: www.elsevier.com/locate/jretconser
Discernible impact of augmented reality on retail customer's experience, satisfaction and willingness to buy
crossmark
⁎
Atieh Poushneh, Arturo Z. Vasquez-Parraga
College of Business and Entrepreneurship, University of Texas Rio Grande Valley, 1201 W. University Drive, Edinburg, TX 78539, USA
A R T I C L E I N F O
A BS T RAC T
Keywords: Augmented reality Retail user experience User satisfaction User's Willingness to buy
User technology has an agreeable impact on consumer decisions; yet the way such impact takes place may be little known. This study attempts to examine the impact of augmented reality (AR) on retail user experience (UX) and its subsequent influence on user satisfaction and user's willingness to buy. Five hypotheses are tested using a lab experiment. The results show that AR significantly shapes UX, by impinging on various characteristics of product quality, and that UX subsequently influences user satisfaction and user's willingness to buy. UX is captured as a third-order formative construct derived from four user experience characteristics: pragmatic quality, aesthetic quality, hedonic quality by stimulation and hedonic quality by identification. Except for the latter, these characteristics are second-order constructs. Important implications for researchers and managers follow.
1. Introduction Customers intending to buy a toy walk into a store. An unassembled 3D puzzle catches their eye, but they are not quite sure what the final assembly will look like. Then, they are told about augmented reality, a collection of viewing features that helps customers visualize the assembled toy in three dimensions (3D), which enables them to observe the puzzle from every angle. This example illustrates how AR helps customers/users make purchase decisions. The literature on AR has emphasized the technological aspects of AR, but it has neglected the end user's needs and problems (Swan and Gabbard, 2005). Yet, AR is increasingly employed in designing and delivering products, even though research has not been able to catch up with the trend from a marketing perspective (Kozick and Gettliffe, 2010; Swan and Gabbard, 2005), especially the growing impact of AR on user experience (UX). This study attempts to understand the way AR influences UX and, at the outset, user satisfaction (US) and user willingness to buy (UWB). Although prior literature has studied some UX dimensions, no mutual agreement about measuring UX has been reached (Vermeeren et al., 2010). Earlier UX studies focused on such cognitive dimensions of UX as usability (e.g., Butler, 1996), but they have ignored UX's affective dimensions. To correct such narrow focus, a user-centered design (UCD) that involved users in the design process emerged (Karat, 1996) and embraced the cognitive and affective dimensions of UX (Alben, 1996). Thus, this study attempts to measure a unified measure of UX and answer the following research questions.
⁎
RQ1: How does augmented reality improve retail user experience? Why is it important that augmented reality enhance the user experience? RQ2: Which and how do key factors moderate the relationship between augmented reality and expected retail user experience, if any? RQ3: What are the effects of retail user experiences on two main consumer outcomes, user satisfaction and user's willingness to buy? The remainder of this paper is organized as follows: first, a brief literature review of the main concepts used and relationships in the study will be discussed. Next, hypotheses, methodology, and results will be explained. Finally, conclusions, managerial implications, limitations, and proposals for future research will be discussed. 2. Conceptual framework 2.1. Augmented Reality (AR) AR is a series of technologies that integrate real world and virtual information, thereby enhancing a specific reality (Lamantia, 2009). Some customers do not make online purchase because such deficiencies make the process risky (Kim and Forsythe, 2008a). AR can produce meaningful experiences for online shoppers (MacIntyre et al., 2001) by providing sufficient product information (Lu and Smith, 2007) that enables them to evaluate the targeted products
Corresponding author. E-mail addresses: [email protected] (A. Poushneh), [email protected] (A.Z. Vasquez-Parraga).
http://dx.doi.org/10.1016/j.jretconser.2016.10.005 Received 18 September 2016; Accepted 6 October 2016 0969-6989/ © 2016 Elsevier Ltd. All rights reserved.
Journal of Retailing and Consumer Services 34 (2017) 229–234
A. Poushneh, A.Z. Vasquez-Parraga
3. Relationships and hypotheses
(Kim and Forsythe, 2008a) and make decisions with more certainty (Oh et al., 2008).
3.1. AR effect on UX as reflected in product pragmatic quality (PQ) PQ is also called usability when it relates to the effectiveness, efficiency, and satisfaction of the UX (Butler, 1996). Since the usability aspect of UX covers a narrow scope of UX, it is not examined as a criterion to evaluate UX (Norman, 2004). The many features of a product, including usability, functions, size, weight, symbols, aesthetic, and usefulness may influence UX. PQ involves a portion of those interactions that emphasize the utility and usability of a product in relation to its potential tasks (Hassenzahl et al., 2003). AR enhances UX by revealing more product information than products without AR, which results in higher UX at the time of purchase, reduces users’ anxiety (Huang and Hsu-Liu, 2014), and facilitates decision-making (Kim and Forsythe, 2008a, 2008b).
2.1.1. AR is reflected in the level of interactivity AR is a stimulus in this study, and the level of interactivity was chosen to reflect AR. Interactivity refers to the “extent to which users can participate in modifying the form and content of a mediated environment in real time” (Steuer, 1992, p. 84). Interactivity entertains users and enables them to personalize information in a 3D virtual model (Fiore, Kim and Lee, 2005), and they enjoy interacting with virtual objects more than they do handling or looking at physical objects (Li et al., 2001). In this study, three levels of interactivity are examined: high, middle, and low. It is assumed that a high level of interactivity will generate a greater UX and subsequently higher user satisfaction and user willingness to buy. Conversely, a low level of interactivity will generate weaker UX and subsequently weaker user satisfaction and user willingness to buy. In this study, high and the middle level of interactivity were examined as AR treatments, and low level of interactivity was examined as non-AR treatment.
3.2. AR effect on UX as reflected in product hedonic quality (HQ) PQ is essential to UX, but it does not exhaust UX. UX also involves emotional reactions (Hassenzahl and Tractinsky, 2006; Norman, 2004). Consequently, AR may influence UX as well by affecting HQ and thus facilitating several emotional benefits. AR facilitates user involvement and thereby enhances the hedonic value of experience (Kim and Forsythe, 2008b), which provides users the ability to share personalized experiences on social networks, thus enhancing playfulness (Huang and Hsu-Liu, 2014). Yet, the effect of AR on HQ can vary depending on whether the experience reflects enjoyment or social reference. Hassenzahl et al. (2003) distinguish three types of effects in HQ: effects by stimulation (HQ-S), effects by identification (HQ-I), and effects by evocation (HQ-E). HQ-S is related to the fulfillment of human needs for novelty and challenge. HQ-I refers to the fulfillment of human needs as self-expressions. HQ-E refers to the human fulfillment needs for symbolic meanings of an object.
2.1.2. Retail user experience (UX) UX is holistic and subjective (McCarthy and Wright, 2004), and varies across time (Law et al., 2009). It is also defined as: “All the aspects of how people use an interactive product: the way it feels in their hands, how well they understand how it works, how they feel about it while they are using it, how well it serves their purposes, and how well it fits into the entire context in which they are using it” (Alben, 1996, p. 5). UX is a complex construct that encompasses a user's inner state, product characteristics, and the context of use (Hassenzahl and Tractinsky, 2006). Product attributes include pragmatic quality (PQ), aesthetic quality (AQ), and hedonic quality (HQ). To measure UX, this study focuses on these product attributes (See Fig. 1)
3.3. AR effect on UX as reflected in product aesthetic quality (AQ) The AQ of UX involves pleasurable experiences. Jordan (2002)
Fig. 1. Conceptual framework: The impact of augmented reality on user experience and its outcomes.
230
Journal of Retailing and Consumer Services 34 (2017) 229–234
A. Poushneh, A.Z. Vasquez-Parraga
Table 1 (continued)
Table 1 First-order constructs: EFA results. Constructs
Hedonic Quality by Identification (α=.848; AVE=.44) Unpresentable-presentable Separates me from people-bring me closer to people Alienating-integrating Cheap-expensive Tacky-stylish Isolating-connective Decreases one's self image-augments one's self-image Loneliness-the sense of belonging to the community
.773 .510 .786 .372 .807 .820 .549 .542
Hedonic Quality by Emotional Stimulation (α=.882; AVE=.54) Repelling-appealing Discouraging-motivating Not absorbed-over absorbed Not immerse-immerse
.930 .712 .454 .759
Hedonic Quality by Rational Stimulation (α=.922; AVE=.669) Ordinary-novel Dull-captivating Conservative-innovative Cautious-bold Unimaginative-creative Conventional-inventive
.720 .828 .809 .749 .868 .918
Aesthetic Quality - Cognitive (α=.955; AVE=.729) Ugly-beautiful Unattractive-attractive Asymmetric-symmetric Unclean-clean Aesthetically unpleasing-aesthetically pleasing Rigid design-artistic design Static-vivid Artificial-Realistic
.926 .930 .711 .797 .833 .892 .904 .816
Aesthetic Quality - Affective (α=.894; AVE=.753) Unfriendly-friendly Annoying-enjoyable Unpleasant-pleasant
.722 .882 .980
Pragmatic Quality - Practical (α=.902; AVE=.61) Unruly-manageable Confusing-clearly structured Impractical-practical Complicated-simple Difficult to learn-easy to learn Effortful-effortless
.583 .853 .911 .834 .821 .652
Pragmatic Quality - Reliable (α=.864; AVE=.52) Unpredictable-predictable Cumbersome-straightforward Unprofessional-professional Insecure-secure Irrelevant-relevant Unreliable-reliable
.750 .868 .611 .641 .619 .809
Pragmatic Quality - Informative (α=.733; AVE=.42) Too few information-too much information Shady-trustworthy Not personalized-personalized Highly decreases one's awareness-highly augments one's awareness Pragmatic Quality - Useful (α=.974; AVE=.81) Highly decrease one's capabilities-highly augments one's capabilities Risky to use-safe to use User Satisfaction (α=.938; AVE=.836) Overall, I am satisfied with the Ray-Ban website. Being a user of this website has been a satisfying experience.
Constructs
Factor loadings
Having experienced this website was pleasurable.
.890
User's Willingness to Buy (α=.954; AVE=.874) I intend to buy my eyeglasses/sunglasses via the Ray-Ban website.
.911
Factor loadings
I would be willing to buy my eyeglasses/sunglasses via the Ray-Ban website. In future, I would buy my eyeglasses/sunglasses via the RayBan website. Trade-off Price and Value (α=.905; AVE=.779) The product offered in the website of Ray-Ban app is reasonably priced. The product offered in the website of Ray-Ban is a good value for the money. At the current price, the product offered in the website of Ray-Ban provides a good value. User's information privacy Control (α=.956; AVE=.84) I was informed about the personal information that Ray-Ban website would collect about me, such as email, name, and location. This website explained the reasons why my personal information is being collected. This website informed the way my personal information would be used. This website gave me a clear choice before using personal information about me.
.929
.726 .982 .920
.934
.979 .915 .838
Note: The following items were eliminated: PQ1 (human-technical), HS2 (challengingundemanding), and ASC6 (bad-good), as explained in the text.
identifies four types of pleasure. Physio-pleasure is related to the sensual UX (e.g. touch, smell, taste). Socio- pleasure is related to the relationship of the user with others (e.g. status, connection). Psychopleasure is related to people's cognitive and emotional reactions (e.g. satisfaction of instrumental needs), and ideo-pleasure is related to the values of people (e.g. aesthetics, taste, personal aspirations). It is possible that not all types of pleasure are sought at once, but some are. Thus, H1. Augmented reality positively and significantly impacts user experience as reflected in four product characteristics: pragmatic quality, aesthetic quality, hedonic quality by stimulation, and hedonic quality by identification. 3.4. Moderating Variables of the ARUX Relationship: Trade-off between Price and Value (PV) and User’s Information Privacy Control (UIPC) Dodds et al. (1991) explain that perceived value functions as a trade-off between perceived quality and monetary sacrifice. Consistent with Equity Theory (Adams et al., 1976), shoppers compare what they gain against what lose when they buy a product. UX may be strengthened if users find the trade-off between price and value to be acceptable, and it may be weakened if users do not consider the tradeoff between price and value acceptable. Thus,
.736 .602 .529 .694
H2. The relationship between AR and UX is stronger when users find a proper fitting trade-off between value and price, and it is weaker when users do not find a fitting trade-off between value and price. Additionally, this study hypothesizes that user's information privacy control (UIPC) moderates the impact of AR on UX. UIPC refers to individuals’ ability to control their personal information (Metzger, 2004), and AR has the potential to collect and personalize user information. Users are sensitive about their privacy and expect to have
.900 .900
.898 .954 (continued on next page)
231
Journal of Retailing and Consumer Services 34 (2017) 229–234
A. Poushneh, A.Z. Vasquez-Parraga
Table 2 Constructs, path coefficients, R-squared, and p-values. Construct Order
Constructs
Mean
Standard deviation
Path coefficients
Beta
R-squared
P-value
Third-order First-order First-order
UX US UWB
.842 .825 .676
.034 .032 .057
AR→UX UX→User Satisfaction UX→User's Willingness to Buy
.30 .87 .76
.867 .760 .580
.003 .000 .000
Fig. 2. Results of SmartPLS: path coefficients, R-squared, p-values.
4. Methodology
access to AR applications that do not require them to share personal information (Olsson et al., 2013). Consistent with Equity Theory (Adams et al., 1976), the amount, exactness, and sensitivity of information that the user is willing to share with a service rely on the extent of value gained from the feature that requires user information (Olsson et al., 2013). If users have control over their personal information while using AR, then they may be genuinely motivated to engage in an experience using AR and vice versa. Thus,
The study was conducted in a laboratory environment located at a large South West University in the US. It utilized a true experiment (Kerlinger and Lee, 2000) in which participants were randomly assigned to three treatments. AR was considered as a stimulus in terms of two levels of interactivity, namely high and middle. For the experimental groups (high and middle level of interactivity), two AR treatments, Ray-Ban sunglasses and Virtual Model were examined. For the control group (very low level of interactivity), traditional online shopping was examined. Quantifying UX is important because it provides a guideline for product designers to choose effective design strategies (Law and Van Schaik, 2010). To measure PQ and HQ, an AttrakDiff 2 questionnaire from Hassenzahl et al. (2003) was used. Since AR is interactive, simulated sensory experience, it was necessary to add more properties to reflect AR. To measure PQ, eight items were taken from Haasenzahl et al. (2003) and one item from Laugwitz et al., (2008). Ten new items were added. To measure HQ-I, six items were taken from Hassenzahl and two new items were added. To measure HQ-S, nine items were taken from Hassenzahl et al., (2003) and two new items were added. And to measure AQ, three items were taken from Lavie and Tractinsky (2004), five items from Laugwitz et al., (2008), and four new items were added. Overall, all UX items were measured using a bipolar semantic differential 7-scale method. To measure user satisfaction (US), items were taken from Taylor and Baker (1994); to measure UWB, three items from Engel et al. (1995); to measure PV, three items were taken from (Dodds et al., 1991); and to measure UIPC, four items were taken from Liu et al. (2004). These items were measured using a 7-point Likert scale with the anchors being “strongly disagree” and “strongly agree.”
H3. The relationship between AR and UX is strengthened or weakened when the user's information privacy control is either empowered or diminished by AR.
3.5. Outcome variables: user satisfaction (US) and user's willingness to buy (UWB) Customer “satisfaction is not [only] the pleasurableness of the [consumption] experience, it is the evaluation rendered that the experience was at least as good as it was supposed to be” (Hunt, 1977, p. 459). AR generates US through experiential value (Yuan and Wu, 2008). Additionally AR is also able to influence US before the buying process (Bulearca and Tamarjan, 2010). Thus, H4. An AR-enriched user experience positively and significantly impacts user satisfaction. User Willingness to Buy (UWB) refers to consumers’ tendency to purchase targeted products in the future and may predict actual purchase behavior (Morrison, 1979). UWB is also impacted by an AR-enriched UX. AR, such as virtual image technology (Verton, 2001) can offer a simulated experience to users with the purpose of encouraging them to buy the product (Huang and Hsu-Liu, 2014). Virtual objects and the information contributed by AR may heighten user's enjoyment and mental imagery (Schlosser, 2003), which in turn may stimulate UWB (Huang and Hsu-Liu, 2014; Kim and Forsythe, 2008a). Thus,
4.1. Participants, prescreening questions and procedures Purposive sampling was used to recruit the participants. The sample for the study consisted of 99 mostly young consumers (45 male, and 54 female) at a U.S. Southern Metropolitan City. The age of the consumers ranged from 20 to 60. All questionnaires were used in the analyses. Missing data was minor and did not exert any undue effect. Before conducting the experiment, the researcher told partici-
H5. An AR-enriched user experience positively and significantly impacts user's willingness to buy. 232
Journal of Retailing and Consumer Services 34 (2017) 229–234
A. Poushneh, A.Z. Vasquez-Parraga
pants that “this study is trying to evaluate the impact of AR on UX and its outcomes.” Before exposing them to the treatments, they answered prescreening questions that asked about their technology use and online shopping (Jin, 2001; Olsson et al., 2013), including the extent to which they like to buy eyeglasses/sunglasses online. The prescreening questions’ Mean (M) and standard deviations (SD) are as follow. “I am familiar with using the Internet” (M=6.9; SD=.35); “I frequently use the Internet to shop online” (M=6.5; SD=.9); “I think that technology is necessary for my daily works” (M=6.6; SD=.57); “I visit the Internet retail websites to collect product information” (M=5.6; SD=1.7); “I visit the Internet retail websites for purchasing products” (M=5.3; SD=1.6); “I am a user of eyeglasses/sunglasses” (M=5.8; SD=1.4); “I would like to wear eyeglasses/sunglasses” (M=6.6; SD=.7). A few studies have considered the level of interactivity as a stimulus (e.g., Lee et al., 2006). This research includes manipulation check questions. Each question is measured using a 7-point Likert scale from 1=strongly disagree to 7=strongly agree. The questions are “The website provides a variety of ways for viewing product image,” “The website provides personalized product,” “The website is interactive,” “The website allows the user to interact with the products shown on the screen,” “The website allows the user to adjust the product,” and “The website shows dynamic product images.” All participants in both the experimental and control groups were asked to answer the prescreening questions, and then the researcher explained the purpose of study to them, which was to understand the impact of augmented reality on user experience. For the high level of interactivity stimulus, the researcher asked the participants to log in to their computers and enter “http://www.rayban.com/international/ virtual-mirror”. They could personalize, zoom in, zoom out on their favorite eyeglasses and sunglasses, and they could take a photo with the virtual product and share it on their social networks. For the middle level of interactivity, the experimenter asked the participants to enter http://www.ray-ban.com/international/virtual-mirror. This virtual model does not have all the features included in high level of interactivity of AR described above. The virtual model includes two models for men and women. For the low level of interactivity (control group), the researcher asked the participants to enter “http://www.rayban.com/usa.” This treatment did not have novel features. After ten minutes, the researcher asked all participants to log off their computers and then answer the survey.
threshold value of .5 for convergent validity (McDonald and Ringo-Ho, 2002). AVEs above .5 and square roots of AVEs above inter-factor correlations show discriminant validity (Fornell and Larcker, 1981). All constructs had discriminant validity. The data obtained from the high and middle level of interactivity of AR was considered AR data, whereas the data obtained from the low level of interactivity was considered non-AR data. The structural model was tested using SmartPLS 3.0 because SmartPLS works well with small sample sizes, and it is appropriate for both reflective and formative constructs. Moreover, it also works well for the aims of exploratory or theory development research (Hair et al., 2012). Measuring UX is challenging, and no studies as yet have attempted to measure a unified UX (Law and Van Schaik, 2010). In this study, UX was measured as a formative third-order construct because the direction of causality is from the indicators toward the construct. To validate the formative nature of the UX, the weights of the formative construct's indicators were checked, and only indicators with significant weights were retained (Hair et al., 2012). The weights of PQ, HQS, HQ-I, and AQ were significant; therefore, they were also retained. The results of SmartPLS 3.0 demonstrated that the AR condition is positively and significantly associated with UX (β=.3; R2=.867; t=2.37; p=.003), indicating that H1was supported. Regarding the outcome variables, the results showed that H4 and H5 are also supported, that is, UX is positively and significantly associated with US (β=.873; R2=.763; t=49.90; p=.000) and UWB (β=.761; R2=.58; t=21.89; p=.000). Table 2 shows the constructs, path coefficients, R squared coefficients, and p-values associated with H1, H4, and H5. To test the impact of the moderators of the AR-UX relationship, SmartPLS was applied. The results showed that PV (β=.85, p=.45, t=1.46) and UIPC did not moderate the impact of AR on UX (β=.032, p=.2974, t=1.055). To test the mediation impact of UX on US, and UWB, SmartPLS 3.0 by 5000 bootstrap was used. The results indicated that UX partially mediates the relationship between AR and UWB (R2=.058, β=.246, p=.008, t=2.655) and that UX is only a partial mediator of the relationship between AR and US (R2=.183, β=.427, p=.000, t=6.667). Fig. 2 shows the path coefficients, R-square, and pvalues.
5. Analysis and results
6. Discussion and conclusion
SPSS was conducted to obtain descriptive statistics and reliability results. Table 1 shows the results of reliability, AVE results, and factor loadings. Cronbach Alphas range from .733 to .991, thus demonstrating construct internal consistencies (Nunnally and Bernstein, 1994), and factor loadings range from .372 to .982. After checking reliability, exploratory factor analysis (EFA) with the sixty-three items using the maximum likelihood method (MLE) and Varimax rotation was conducted to check the uni-dimensionality of constructs. Thirteen factors emerged with acceptable default eigenvalues. All factor loadings were higher than the minimal level, .3 (Hair et al., 2006). Overall, four factors from PQ emerged: practical, informative, reliable, and useful. Additionally, two factors emerged from HQ-S, rational stimulation and emotional stimulation. Two factors emerged from AQ, cognitive aesthetic and affective aesthetic. Other constructs, HQ-I, US, UWB, UIPC, and PV were uni-dimension. After running a reliability test and EFA, three items with low alpha levels, PQ1 (human-technical), HS2 (challenging-undemanding), and ASC6 (bad-good), were eliminated. The final EFA using the sixty retained items confirmed the dimensionality of the scales and returned theoretically and empirically acceptable solutions (see Table 1). To check convergent and discriminant validity, the average variances extracted (AVE) and composite reliabilities (CR) were checked. The AVEs ranged from .59 to .87, thus satisfying the recommended
The study investigated the impact of AR on UX and its subsequent effects on US and UWB. AR significantly and positively influences UX (H1). AR provides more 3D product information, in different colors and styles, which enhances users’ perception of reality. AR also provides the user with enriched product information gained from a physical store as well as online store. For example, AR allows the user to simulate the product's features on a website in online shopping (Fiore et al., 2005). In addition, AR empowers users to share their personalized experiences on social networks, which enhances playfulness (Huang and Hsu-Liu, 2014). In relation to the impact of an AR-enriched UX on US, the results indicated that an AR-enriched UX empowers users to better perform their tasks and appreciate the functionality of the product more. An AR-enriched UX is more entertaining and enables potential customers to have endless interaction with virtual information. Thus, AR-enriched UX produces higher user satisfaction (H4) and user willingness to buy (H5). AR can enhance hedonic values, leading to an increase in UWB (Huang and Hsu-Liu, 2014) by adding virtual information to real information and offering 3D pictures of products in different shapes, colors, and styles. Further, UX partially mediated the impact of AR on US and UWB. In relation to the moderator effects, the results indicated that PV and UIPC did not moderate the impact of AR on UX (H2 and H3). 233
Journal of Retailing and Consumer Services 34 (2017) 229–234
A. Poushneh, A.Z. Vasquez-Parraga
Hassenzahl, M., Burmester, M. Koller, F., 2003. AttrakDiff: Ein fragebogen zur messung wahrgenommener hedonischer und pragmatischer Qualität, Mensch & Computer, Vieweg Teubner Verlag, pp. 187–196 Huang, T.-L., Hsu-liu, F., 2014. Formation of augmented reality interactive technology’s persuasive effects from the perspective of experiential value. Internet Res. 24 (1), 82–109. Conceptualization and measurement of consumer satisfaction and dissatisfaction. In: Hunt, H.K. (Ed.), . Marketing Science Institute, Cambridge, MA, 77–103. Jin, H.-J., 2001. Effects of Interactive Web Site Features on Approach Responses Towards An Online Apparel Retailer. Iowa State University, Ames, IA. Jordan, P.W., 2002. Human factors for pleasure seekers, Design and the Social Sciences: Making Connections, 2(9) Karat, J., 1996. User centered design: quality or quackery? Interactions 3 (4), 18–20. Kim, J., Forsythe, S., 2008a. Adoption of virtual try-on technology for online apparel shopping. J. Interact. Mark. 22 (2), 45–59. Kim, J., Forsythe, S., 2008b. Sensory enabling technology acceptance model (SE-TAM): a multiple-group structural model comparison. Psychol. Mark. 25 (9), 901–922. LamantiaJ., 2009. Inside out: interaction design for augmented reality. UX Matters, August 17. Laugwitz, B., Held, T., Schrepp, M., 2008. Construction and Evaluation of a User Experience Questionnaire. Springer, Berlin Heidelberg, 63–76. Lavie, T., Tractinsky, N., 2004. Assessing dimensions of perceived visual aesthetics of web sites. Int. J. Hum.-Comput. Stud. 60 (3), 269–298. Law, E., L.-C., Roto, V., Hassenzahl, M., Vermeeren, A.P., Kort, J., 2009. Understanding, scoping and defining user experience: a survey approach. In: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. ACM, pp. 719–728. Law, E.L.-C., Van Schaik, P., 2010. Modelling user experience–an agenda for research and practice. Interact. Comput. 22 (5), 313–322. Lee, H.-H., Fiore, A.M., Kim, J., 2006. The role of the technology acceptance model in explaining effects of image interactivity technology on consumer responses. Int. J. Retail Distrib. Manag. 34 (8), 621–644. Li, H., Daugherty, T., Biocca, F., 2001. Characteristics of virtual experience in electronic commerce: a protocol analysis. J. Interact. Mark. 15 (3), 13–30. Liu, C., Marchewka, J.T., Ku, C., 2004. American and Taiwanese perceptions concerning privacy, trust, and behavioral intentions in electronic commerce. J. Glob. Inf. Manag. 12 (1), 18–40. MacIntyre, B., Bolter, J.D., Moreno, E., Hannigan, B., 2001. Augmented reality as a new media experience. In: Augmented Reality, Proceedings: IEEE and ACM International Symposium on, pp. 197–206. McCarthy, J., Wright, P., 2004. Technology as experience. Interactions 11 (5), 42–43. McDonald, R.P., Ringo-Ho, M.-H, 2002. Principles and practice in reporting structural equation analyses. Psychol. Methods 7 (1), 64. Morrison, D.G., 1979. Purchase intentions and purchase behavior. J. Mark., 65–74. Norman, D.A., 2004. Emotional Design: Why We Love (or Hate) Everyday Things. Basic Books, New York. Nunnally, J.C., Bernstein, I.-H., 1994. The assessment of reliability. Psychom. Theory 3 (1), 248–292. Oh, H., Yoon, S.-Y., Shyu, C.-R., 2008. How can virtual reality reshape furniture retailing? Cloth. Text. Res. J. 26 (2), 143–163. Olsson, T., et al., 2013. Expected user experience of mobile augmented reality services: a user study in the context of shopping centers. Pers. Ubiquitous Comput. 17 (2), 287–304. Schlosser, A.E., 2003. Experiencing products in the virtual world: the role of goal and imagery in influencing attitudes versus purchase intentions. J. Consum. Res 30 (2), 184–198. Steuer, J., 1992. Defining virtual reality: dimensions determining telepresence. J. Commun. 42 (4), 73–93. Swan, J.E., Gabbard, J.L., 2005. Survey of user-based experimentation in augmented reality. In: Proceedings of the 1st International Conference on Virtual Reality. Taylor, S.A., Baker, T.L., 1994. An assessment of the relationship between service quality and customer satisfaction in the formation of consumers' purchase intentions. J. Retail. 70 (2), 163–178. Vermeeren, A. P. et al., 2010. In: User experience evaluation methods: Current state and development needs. In:Proceedings of the 6th Nordic Conference on HumanComputer Interaction: Extending Boundaries. ACM, pp. 521–530. Verton, D., 2001. Churn. Computer World. 35(6), pp. 50 Yuan, Y.-H., Wu, K.W., 2008. Relationships among experiential marketing, experiential value, and customer satisfaction. J. Hosp. Tour. Res 32 (3), 387–410.
7. Managerial implications, limitations, and future research AR has been developed as a usable and artistic tool to draw users’ attention. The number and quality of AR applications have been increasing, and it seems that only recently AR developers are paying attention to user's actual needs and desires. AR should be developed to satisfy the various characteristics of UX, in particular PQ, HQ-S, HQ-I, and AQ. A well-designed AR should be practical, easy to use, easy to learn, organized, symmetric, attractive, and pleasant if it is to provide effective and relevant information to users, and it should contemplate the emotional aspects of UX. AR should be novel and pleasant so that it will empower and encourage users to express themselves on social media. Recognizing the power of AR in UX and its mediated role in customer's satisfaction and purchase intention, retailers and AR companies can collaborate to develop marketing strategies that effectively enrich and enhance customers’ shopping experience. For instance, retailers can implement promotions that enhance the participation of customers using AR apps such as Blippar apps that are used in smartphones to reveal surrounding objects by just pointing the smartphone to the goods. Bountiful virtual information that relate to real products will appear on the smart phone's screen as if the viewer would be in a fantasy world. As a result, customers will be motivated to download AR apps that enhance reality along with promotions. In turn, retailer can make more sales and AR companies receive additional orders for the technology. This study has some limitations. First, this study did not include the impact of such other variables as culture and ethnicity. Additionally, only two consequences of UX were considered in this research. For future research, hedonic quality by evocation may be included as another product characteristic of UX. Moreover, it would be useful and illuminating to evaluate the impact of AR in other user contexts, such as health care. References Adams, J.S., Berkowitz, L., Walster, E., 1976. Advances in Experimental Social Psychology, Equity Theory: Toward a General Theory of Social Interaction. Academic Press, 9. Alben, L., 1996. Defining the criteria for effective interaction design. Interactions 3 (3), 11–15. Bulearca, M., Tamarjan, D., 2010. Augmented reality: a sustainable marketing tool? Glob. Bus. Manag. Res.: Int. J. 2 (2–3), 237–252. Butler, K.A., 1996. Usability engineering turns 10. Interactions 3 (1), 58–75. Dodds, W.B., Monroe, K.B., Grewal, D., 1991. Effects of price, brand, and store information on buyers' product evaluations. J. Mark. Res., 307–319. Engel, J.F., Blackwell, R.D., Miniard, P.W., 1995. Consumer Behavior. Dryder, New York. Fiore, A.M., Kim, J., Lee, H.-H., 2005. Effect of image interactivity technology on consumer responses toward the online retailer. J. Interact. Mark. 19 (3), 38–53. Fornell, C., Larcker, D.F., 1981. Evaluating structural equation models with unobservable variables and measurement error. J. Mark. Res., 39–50. Hair, J.F., Black, W.C., Babin, B.J., Anderson, R.E., Tatham, R.L., 2006. Multivariate Data Analysis 6. Pearson Prentice Hall, Upper Saddle River, NJ. Hair, J.F., Sarstedt, M., Ringle, C.M., Mena, J.A., 2012. An assessment of the use of partial least squares structural equation modeling in marketing research. J. Acad. Mark. Sci. 40 (3), 414–433. Hassenzahl, M., Tractinsky, N., 2006. User experience – a research agenda (Editorial). Behav. Info Tech. 25 (2), 91–97.
234