Effect of pictures in positive online consumer reviews on consumers’ shopping intention

!

MASTER THESIS

Effect of Pictures in Positive Online Consumer Reviews on Consumers’ Shopping Intention.

University of Amsterdam Faculty of Economics and Business

Master of Science in Business Administration Track: Marketing


Under supervision of: Zerres Alfred

By: Student: Jiang Lyu Student Number: 10918841

ABSTRACT

In recent years, the sharing of electronic word-of-mouth has become a major information source for consumer purchase decisions (Park, Gu, & Lee, 2012). Online consumer reviews, one type of eWOM, affect consumers’ online shopping intentions by offering positive or negative statements made by previous consumers about a product or service in online stores. Because people assume that pictures are always worth a thousand words (Hoffman & Daugherty, 2013), it is popular to include pictures in online consumer reviews. However, the influence of pictures in positive online consumer reviews on consumers’ online shopping intentions remains unclear. In order to better understand the perceived usefulness of online reviews, this study involved a within-subjects experiment on 200 Chinese online shopping consumers. The study results suggest that there is a positive correlation between pictures in positive online reviews and consumers’ online shopping intentions. Perceived usefulness and perceived credibility play a mediating role in this positive relationship. Pictures can increase perceived usefulness and perceived credibility of positive online consumer reviews and increase consumers’ shopping intentions as a result. Furthermore, the number of pictures, the quality of the pictures and the content of the pictures are shown to be associated with consumers’ online shopping intention. This study concludes by explaining the managerial implications of the results and offering suggestions for directions for future research.

Keywords: Online shopping intention, eWOM, online consumer reviews, perceived usefulness, perceived credibility


ACKNOWLEDGEMENT

I would like to thank several individuals for their guidance and help to support me in writing this thesis. First, I would like to acknowledge Professor Mr. Dr. Zerres Alfred for all his support and guidance that he provided during this thesis development process. Furthermore, I would like to thank

my friend Randi Pei and my family who always supported me and gave me a helping hand when needed. This thesis would not have been possible without their support. At last, my thesis is written.

I hope you enjoy reading this thesis!

Statement of Originality

This document is written by Jiang Lyu who declares to take full responsibility for the contents of this document.

!I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

TABLE OF CONTENT

1. Introduction………10

2. Literature review……….13

2.1 Online consumer behavior………..13

2.2 Electronic Word-of-Mouth………..15

2.3 Online Consumer Review………17

2.4 Theory of Planned Behavior and Online shopping intention………..20

2.5 Perceived usefulness and perceived credibility………..20

2.6 Picture characteristics……….23 2.7 Conceptual framework……….………..24 3. Methodology………..26 3.1 The sample……….26 3.2 Research design………..28 3.3 The procedure……….28 3.3.1 Pilot study………..28 3.3.2 Main study……….……28 4. Result………..29 4.1 Reliability………29 4.2 Correlation check………30 4.3 Model testing..………32 5. Discussion………38

5.1 Theoretical and practical implications………39

5.3 Further research..………43

6. Conclusions……….43

Reference………..46

INDEX OF TABLES

Table 1: Descriptive analysis of subjects………..26 Table 2: Reliability of scales……….……20 Table 3: Descriptive analysis of subjects………..31 Table 4: H1. Difference in consumer’s shopping intentions of positive online consumer reviews without pictures and of positive online reviews with one picture………..………..32 Table 5: H2. Difference in consumer’s perceived usefulness of positive online consumer reviews without pictures and of positive online reviews with one picture………..…..34 Table 6: H3. Difference in consumer’s perceived credibility of positive online consumer reviews without pictures and of positive online reviews with one picture ……….…..34 Table 7: H4. Difference in consumer’s online shopping intentions, perceived credibility and perceived usefulness of different number of pictures in positive online consumer reviews……….……36 Table 8: H5. Difference in consumer’s online shopping intentions and perceived usefulness of different content of pictures in positive online consumer reviews………..….37 Table 9: H6. Difference in consumer’s online shopping intentions, perceived credibility and perceived usefulness of different quality of pictures in positive online consumer reviews………..…38

INDEX OF FIGURES

Figure 1: Conceptual framework……….………..……25

1. INTRODUCTION

Since the development of World Wide Web (WWW) on the Internet in the early 1990s, an increasing number of companies have been trying to carry out electronic commerce (EC) (Park & Lee, 2009). Although electronic commerce expend a major effort in maximizing customer satisfaction and minimizing customer dissatisfaction, they face critical limitations. Online customers cannot see/ touch/smell/hear the actual products via online transactions (Cho et al., 2003). In the meantime, the rapid proliferation of the Internet has also brought about a transition in customer-to-customer communication, away from traditional word-of-mouth (WOM) networks and toward digital networks, which is a form of communication called electronic word of mouth (eWOM) (Hsieh et al., 2012). Recently, the WWW is used as a new marketing channel to show recommendations from previous consumers. Through the WWW, consumers can share their opinions to a wider public than ever before through its accessibility, reach and transparency (Hennig, Gwinner, Walsh, & Gremler, 2004). In addition, this trend also reduces time and space constraints, enabling WOM recommendations to spread faster, farther, and more efficiently (Hsu & Liu, 2009). The sharing of eWOM has become a major information source for consumer purchase decisions (Park et al., 2012) because eWOM enables consumers to evaluate alternatives, reduce uncertainty and provides consumers with indirect experiences (Park & Lee, 2009).

Online consumer reviews, one type of eWOM, refers to positive or negative statement made by previous consumers about a product or service in online shopping malls. An online consumer review as a route for social influence plays two roles – as an informant and a recommender. As an informant, online consumer reviews deliver additional user-oriented information. As a recommender, they provide either a positive or negative signal of product popularity (Park&lee, 2009). The online consumer review is vital for making online shopping reliable. For example, customers who have prior experiences with specific brands and/or products might not hesitate to purchase them online because their familiarity has accorded them full information about the products. However, without those prior familiarities, it is not easy to determine the quality of products such as clothes, shoes, and cosmetics on the Web. But, online consumer reviews also entail several risks. On the Internet, because consumers and online consumer reviews authors do not know each other, consumers are often unsure

about the credibility and truthfulness of the reviews. Online consumer reviews authors may exaggerate the product quality and efficiency; evaluate the product based on their personal experience, expectations, and needs; be paid to positively review the product; or deceive others with malicious intent. In a word, consumers bear risks when they make purchasing decisions based on online consumer reviews. Many online shopping sellers realize both the importance and risks of the online consumer reviews and therefore use various methods to reduce the risks. In recent years, it is common that many online shopping sellers encourage and reward their previous consumers to add online reviews with pictures, e.g., consumers of Taobao.com are encouraged to upload pictures together with their online reviews in exchange for a cash refund.

Since these people assume that pictures are always worth a thousand words (Hoffman & Daugherty, 2013). However, whether the online consumer reviews with pictures will lead to a higher shopping intention than those without pictures has yet to be decided. Previous researches mainly mentioned the characteristics of verbal contents or the different effects of online reviews in different product categories. Willemsen et al. (2011) did research on online reviews content characteristics. Susan M. Mudambi (2010) analyzed 1587 reviews from Amazon.com across six electronic products and indicates that review extremity, review depth, and product type affect the perceived usefulness of the review. Only few studies focused on visual information (e.g., pictures or videos) rather than verbal information of eWOM. In addition, these studies were all involved in investigations of the effects of visual information in social media (e.g., Facebook, YouTube, Pinterest). Hoffman(2013) did research on Pinterest to compare the effectiveness of images and text in capturing consumer attention. The current literature about eWOM fails to discuss online consumer reviews in online shopping malls or the impacts of pictures in online consumer reviews. Therefore, this thesis aims to analyze how pictures in the online consumer reviews affect consumers’ online shopping intention to bridge this gap. Depending on their valence, online consumer reviews can be classified into positive and negative forms. Positive reviews elicit more positive responses (i.e., attitude or intention to use) than negative ones (Vermeulen & Seegers, 2009) and this thesis focuses on studying the positive reviews.

This study assumes that there is a positive relationship between the pictures in positive online consumer reviews and consumers’ shopping intentions. In the detailed process, the perceived

usefulness of online consumer reviews (Yulihasri et al., 2011) and the perceived credibility of online consumer reviews (Hino, 2015) might be the mediators in this positive relationship. In addition, the number of pictures, the quality of the pictures and the content of the pictures are three possible moderators in this relationships. This quantitative research employs experimental research design a quasi-experiment and within-subjects design. Every subject would experience all the conditions that help reduce the random individual differences in different groups, as individual, consumers’ feelings are different and their criteria of scale level are subjective. This experiment was achieved by online questionnaires that contain six conditions. Condition 1 works as the control condition, Condition 2, 3 work as both control condition and experiment condition and Condition 4, 5, 6 are the experiment conditions. This experiment uses men’s white shirt as a sample product and the target population of this study is all the online shopping consumers. However, only Chinese online shopping malls enable uploading pictures with online reviews. Since Chinese online shopping consumers are quite representative of the world’s online shopping population and there is no significant difference among consumers from different nations, this study chose Chinese customers as the research target. Samples were chosen to stratify by age and gender in order to control the real online shopping consumer situation. 200 randomly chosen, experienced Chinese online shopping consumers were invited to this experiment. Experiment sample was consisted of 72 subjects aged from 14-24 (36 male and 36 female), 76 subjects aged from 25 to 34 (38 male and 38 female), 40 subjects aged from 35 to 44 (20 male and 20 female), 8 subjects aged from 45 to 54 (4 male and 4 female) and 4 aged 55-64 years old consumers (1 male and 3 female). This questionnaire contains six different online consumer reviews and questions about shopping intentions, perceived credibility and perceived usefulness of these online consumer reviews. MacKenzie, Lutz, and Belch (1986) used a three-item, seven-point scale (likely/unlikely, probable/improbable, and possible /impossible) to measure shopping intention and this research used the same scale. Perceived credibility and perceived usefulness were measured by multidimensional measure which was appeared in the Communication literature in 1966. Incredible/ useless 1 2 3 4 5 6 7 credible/useful ( Numbers 1 and 7 indicate a very strong feeling. Numbers 2 and 6 indicate a strong feeling. Numbers 3 and 5 indicate a fairly weak feeling. Number 4 indicates you are undecided). Then, six hypotheses were tested to make sure the complicated relationships between

independent variable, dependent variable, mediators and moderators. This research also provide important implications for online marketing strategies. Only by understanding how the online shopping intention is affected by pictures in positive online consumer reviews can marketers most effectively appeal to those intentions in their marketing campaigns.

In order to reach a comprehensive conclusion, the remainder of this study is divided into five chapters. The next chapter describes a review of the literature on the key concepts about the consumer online shopping behavior, electronic word-of-mouth(eWOM), online consumer reviews, perceived usefulness, perceived credibility and online shopping intentions. Furthermore, a theoretical framework with hypotheses about pictures in positive online consumer reviews’ effect on shopping intention are developed. This framework is examined in the field research, for which the research method is explained in chapter three. The model is presented in chapter four. Finally, there is a discussion of the empirical findings, the managerial implications are given, the limitations and suggestions for further research are discussed and the conclusion is given.

2. LITERATURE REVIEW

In order to have a clear view of the previous study about how pictures in positive online consumer reviews affect consumers’ shopping intention, a comprehensive review of key articles on the research topic field is analyzed as following. First, some basic information about online consumer behavior is given. Second, previous studies about eWOM and the online consumer reviews, reasons why picture is chosen to be the independent variable are stated. Then, some detailed analysis of the importance of online shopping intentions is given. In addition, the possible variables worked in the relationship between the pictures in positive online consumer reviews and shopping intentions are explained. Finally, based on this review, hypotheses and a conceptual framework are developed in this study.

2.1 Online Consumer Behavior

Recently, the Internet as an online shopping channel has boomed. The continuous growth of electronic commerce has stimulated great interest in studying online consumer behaviors. Given the significant growth in online shopping, better understanding of customers allows better marketing

strategies to be designed (Nalchigar and Weber, 2012). Thus, studying consumer online shopping behavior has been one of the most important research agendas in e-commerce during the past decade (Chen, 2009).

Consumer online shopping behavior is a complicated socio-technical phenomenon, which involves too many factors. These various factors play different roles in online consumer behaviors. Some previous studies are about the effects consumers characteristics: Garbarino and Strahilevitz (2004) found that women perceive a higher level of risk in online purchasing than do men and having a site recommended by a friend lends to both a great reduction in perceived risk and a greater increase in the willingness to buy online among women than among men. Andrews, Kiel, Drennan, Boyle, and Weerawardena (2007) concluded that male online purchasers were discriminated from female purchasers by social value and from male non-purchasers by conditional value. Female purchasers were discriminated from male purchasers by functional value and from female non-purchasers by social value. Schiffman, Sherman, and Long (2003) stated that differences were observed in behavior and feelings about the online shopping based on personal values. Respondents who scored high on self-fulfillment were more likely to use the Internet for learning or gathering online shopping information. Seock and Yu (2007) found that consumers with different shopping orientations differ in their Website evaluation criteria, online information searches, and online purchases.

Besides, some researchers focused on analyzing more objective parts, i.e., the product type, company reputation and website difference. Various categories of Web-design elements reinforce Web customers’ beliefs, which in turn positively impact attitudinal constructs that lead to changes in their purchase intentions (Song and Zahedi, 2005). Rewards, privacy notices, and reputation greatly influence consumers’ intention to provide accurate personal information over the Internet (Xie, Teo, and Wan, 2006): Rewards have a positive impact on decision to provide accurate personal information but not for demographic data. Privacy notices boost decision to provide personal and demographic information. A company’s reputation and willingness to provide accurate personal information are highly correlated.

In addition to these findings, consumers’ feelings and attitudes have been another interesting topic for researchers. Consumers’ attitude towards online grocery buying is positively affected by

perceived offline physical effort and negatively affected by offline shopping enjoyment (Hansen, 2006). Attitude towards online shopping positively affected online shopping behavior of consumers (Javadi et al., 2012). Cho (2006) did a survey on adult consumers and concludes that trust and distrust are shaped by different dimensions of trustworthiness, and trust affects behavioral intentions differently than distrust. Therefore, changing consumers attitudes is the main approach for marketers to better online shopping behaviors. Therefore, electronic word-of-mouth (eWOM) has a significant impact on consumer choice, and on post-purchase product perception (Bone, 1995), as well as the company reputation, and should be a valuable phenomenon for analysis. “How does eWOM affect consumers’ online purchasing behavior?” is an interesting question. The next chapter elaborates on eWOM.

2.2 Electronic Word-of-Mouth (eWOM)


The arrival and expansion of the Internet has extended consumers' options for gathering product information by including other consumers' comments, posted on the Internet, and has provided consumers opportunities to offer their own consumption-related advice by engaging in electronic word-of-mouth (eWOM) (Thurau et al. 2004). Over the past few years, numerous studies have examined eWOM (Tsao& Hsieh, 2015).

Most studies on eWOM prior to 2010 focused on quality, quantity, and valence, which are associated with the impact of the content on eWOM. Park & Lee, (2007) examined how the eWOM information direction (positive vs. negative) and a website's reputation (established vs. unestablished) contribute to the eWOM effect. The product type (search vs. experience) always play a moderating role. The eWOM effect is greater for negative eWOM than for positive eWOM, greater for established websites than for unestablished websites, and greater for experience goods than for search goods. The results support the moderating effects of product type on the eWOM information direction-website reputation-eWOM effect relationship. The impact of negative eWOM on the eWOM effect is greater for experience goods than for search goods. Similarly, the impact of website reputation on the eWOM effect is greater for experience goods than for search goods.

Recently, some studies have focused on how credible consumers perceive the eWOM information is, and compare the word-of-mouth(WOM) with eWOM. In eWOM however, unlike the

case of WOM from interpersonal sources, recommendations are typically from unknown individuals and in a text-based format. For online consumer reviews to serve as decision-making aids, users have to first trust the reviewer. However, unlike face-to-face communication, consumers have to rely on personal profile information to establish confidence in the reviewer in an online shopping context. The reviewer profile on most online consumer review websites includes both self-created cues, such as profile picture, and system-generated cues, like the reputation among other users. These personal profile characteristics may serve as cues of source credibility. Reputation cue and profile picture cue contributed differently to users’ affective trust and cognitive trust towards the reviews (Xu, 2014). Reputation cue, generated by the system, was found to influence both affective and cognitive dimensions of trust, whereas the self-generated cue of profile picture affected only affective trust. Reputation cue had a direct influence on perceived review credibility, whereas the influence of profile picture on perceived review credibility was dependent upon review valence (Xu, 2014).

With the emergence of new forms of social media, such as online forums, blogs, Facebook, Line, and Twitter, the electronic word-of-mouth (eWOM) platforms used by consumers are becoming increasingly diverse. Much of the previous research on eWOM platforms has focused on independent platforms (e.g., blogs, online forums, and online communities) (Casteleyn et al. 2009), and excluded corporate platforms. Tsao & Hsieh (2015) investigated how the persuasiveness of eWOM varies according to the type of eWOM platform, and addressed the issue of eWOM related to credence goods. The influence of eWOM on consumers’ purchase decisions varies with the source of the eWOM and product type. The type of eWOM platform moderates the influence of eWOM quality on eWOM credibility and purchase intention, and this phenomenon is particularly significant in search goods. eWOM credibility is a partial mediator between eWOM quality and purchase intention. So, establishing user-oriented online reply systems is vital.

Though eWOM in social media can affect consumers’ purchasing intentions, it is time-wasting and ridiculous for consumers to search for these eWOMs before buying every product online. During their daily online shopping, consumers mainly rely on online consumer reviews in the internet shopping malls. Online consumer reviews are part of the consumer-created information by website users who have already bought the product. Online consumer reviews contain information and

recommendations of the products from the consumer’s perspective (Park et al., 2007). Previous studies have mainly investigated online consumer reviews as electronic word-of-mouth (eWOM) that influences consumer behavior (Brown and Reingen, 1987; Chatterjee, 2001; Chen and Xie, 2008), and seldom did special research on online consumer reviews. As the most extensive type of eWOM information, online consumer reviews have higher credibility, empathy and relevance to the customer than marketer-created sources of information on the Web (Bickart and Schindler, 2001), and are playing an increasingly important role in consumers’ purchasing decisions. So, the next chapter gives an overall review of the research on online consumer reviews.

2.3 Online Consumer Reviews

To reduce uncertainty and perceived risks, consumers often search for online consumer reviews when making online shopping decisions. According to findings from a joint research by PowerReviews and the E-tailing Group, about 22% of respondents said that they “always” read consumer reviews before making a purchase, 43% of respondents said that they check consumer ratings and reviews “most of the time”, and about 68% read “at least four reviews” before making a purchase (Kee, 2008). Online reviews enable people to obtain detailed information with high trustworthiness and credibility compared to information provided by marketers. However, the influx of online consumer reviews has caused information overload, making it difficult for consumers to choose reliable reviews (Baek et.al. 2012).

Based on this situation, a number of researchers in marketing and information systems have concerned the characteristics of reviews and reviewers to estimate the effect of online reviews (Park& Nicolau, 2015). An analysis by Susan M. Mudambi (2010) of 1,587 reviews from Amazon.com across six electronic products indicates that review extremity, review depth, and product type affect the perceived helpfulness of the review. Both peripheral cues, including review rating and reviewer’s credibility, and central cues, such as the content of reviews, influence the helpfulness of reviews. Based on dual process theories, consumers focus on different information sources of reviews, depending on their purposes for reading reviews: online reviews can be used for information search or for evaluating alternatives (Baek et.al. 2012). In addition, the present of reviewers’ photo makes the online reviews higher quality (Lee and Shin, 2013). Park &Nicolau (2015) concluded that consumers

tend to seek heuristic information cues to simplify the amount of information involved in tourist decisions. Accordingly, star ratings in online reviews are a critical heuristic element of the perceived evaluation of online consumer information and consumers perceive extreme ratings (positive or negative) as more useful and enjoyable than moderate ratings, giving rise to a U-shaped line, with asymmetric effects: the size of the effect of online reviews depends on whether they are positive or negative.

Besides the review content characteristics and the number of reviews, the awareness of persuasive intent, perceived interactivity also play a role in affecting consumer attitudes. Online consumer reviews play a recommender role as well as an informant role, so a large number of reviews deliver the signal of product popularity with a variety of product information (Park and Lee, 2008). Interactivity in the form of a blog significantly influences attitude toward the website, but not attitudes toward the candidate or voting intention (Thorson and Rodgers, 2006). Awareness of persuasive intent exerts a negative influence, whereas the humor and multimedia effects have positive influences on both attitude toward a received online video and forwarding intentions (Hsieh et al., 2012). The lack of an interaction effect between the wording and expectation manipulations shows that the pattern of results may be attributed to Verbal Politeness: wordings like ‘not bad’ convey a weakened meaning as compared to ‘good’, whereas the use of ‘not good’ instead of ‘bad’ is interpreted as expressing the same evaluation, albeit in a more polite way (Kamoen et al., 2014).

For the review information to be sufficiently informative, the seller-created product attribute information and the buyer-created review information will interact with each other. For example, when the product cost is low and/or there is a sufficient number of expert (more sophisticated) product users, the two types of information are complementary, and the seller’s best response is to increase the amount of product attribute information conveyed via its marketing communications after the reviews become available. However, when the product cost is high and there are sufficient novice (less sophisticated) product users, the two types of information are substitutes, and the seller’s best response is to reduce the amount of product attribute information it offers, even if it is cost-free to provide such information. Second, product/market conditions under which the seller benefits from facilitating such buyer-created information (e.g., by allowing consumers to post user-based product

reviews on the seller’s website). Finally, the importance of the timing of the introduction of consumer reviews available as a strategic variable and show that delaying the availability of consumer reviews for a given product can be beneficial if the number of expert (more sophisticated) product users is relatively large and cost of the product is low.

Studies in the last three years have done more practical research (e.g. link the online consumer reviews with offline sales), and usually use hotels as their subject. Xie et al. (2014), presented a panel data analysis of online consumer reviews and management responses for 843 hotels on a hotel review website, and found that overall rating, attribute ratings of purchase value, location and cleanliness, variation and volume of consumer reviews, and the number of management responses are significantly associated with hotel performance. Besides, variation and volume of consumer reviews moderate the relationship between overall rating and hotel performance. In addition to the first study, Xie et al. (2015) found empirical evidence which shows the relative effect of online consumer review factors on offline hotel popularity when controlling for other hotel characteristics. In particular, the effect of review quality lasts for at least a couple of quarters, whereas that of other online consumer review factors remains short-term. The findings provide a managerial basis to improve the online presence of hotels on social media platforms by strategically utilizing important review factors. Cssalo et al. (2015) investigated perceived usefulness of online hotel reviews. Study results suggest that high risk-averse travelers find negative online reviews more useful than positive reviews. For positive online reviews, high-risk averse travelers feel expert reviewers' postings, travel product pictures, and well-known brand names enhance usefulness of the positive online reviews.

To sum up, these studies primarily focused on three main parts: the content quality, number of online consumer reviews, the characteristics of reviewers; the awareness of persuasive intent, perceived interactivity; and practical researches on hotel booking websites. However, none of these previous studies paid attention to the pictures in online consumer reviews, even there are some relevant findings about visual information in online consumer reviews, e.g., subjects rated eWOM articles in blogs with visual information significantly higher than identical articles without visual information (Lin et al. 2012), and Xu, (2014) found that the reviewer profile pictures on most online consumer review websites make these reviewers more credible. For an online market to succeed, it is

important to lead product reviewers to write more helpful reviews, and for consumers to get helpful reviews more easily by figuring out the factors determining the helpfulness of online reviews (Baek et.al., 2012). Therefore, the pictures in online consumer reviews is a novel research area which is worthy to be studied.

2.4 Theory of Planned Behavior (TPB) and Online shopping Intention

Ajzen (1991) mentioned that the central factor in the theory of planned behavior (TPB) is the person’s intention to perform a behavior. The intention is proposed to explain factors that influencing behavior. That is to indicate, how people will try, about many efforts to plan in using, in executing behavior (Ajzen, 1991). Ajzen and Maden (1985) found that perception of control, like attitude towards the behavior and subjective norm can have an important impact on an individual’s behavioral motivation. Perceived behavior control the importance of behavior control is self evident and refers to people’s perception of the ease or difficulty performs the interest behavior (Ajzen & Maden, 1985). According to the theory of planned behavior that the behavior is a joint function with intentions and perceived control behavioral. The two perceptions and intention conduct behavior can make important contribution to behavior prediction. Intention is assumed to capture the motivational variables that influence a behavior. They are indications of how hard people are willing to perform the behavior. In addition, the stronger the intention to engage in a behavior, the more likely must be its performance (Yulihasri, 2011). Therefore, in this study, the independent variable is online shopping intentions and the following hypothesis is tested:

H1: There is a direct positive relationship between pictures in positive online consumer reviews and consumers’ online shopping intentions, which means online consumer reviews with pictures can raise stronger consumers’ online shopping intentions than those without pictures.

2.5 Perceived Usefulness and Perceived Credibility Perceived Usefulness

Despite the recognized relationship between pictures in consumer online reviews and online shopping intentions, there is little understanding of how consumers process them. Given the abundance of online reviews and the multitude of problems associated with them (such as anonymity and the potential for manipulation) (Chevalier and Mayzlin 2006), it is important to understand how

consumers manage to alleviate uncertainty and get the information they want and are willing to trust. Assuming that consumers are consulting online reviews because they have the intention to buy a certain product or service (Goldsmith and Horowitz, 2006), some reviews will be categorized as useful (i.e., helpful in making a decision about whether or not to buy or use the reviewed product or service) (Cheung, Lee, and Rabjohn, 2008), and others less so. Consequently, the question that arises is ‘‘What types of reviews do consumers find useful?’’ Understanding this issue will help both marketing researchers and managers come to terms with the reputation-based online information mechanisms currently used by most information providers and online retailers (Libai et al. 2010). To what extent can different picture characteristics in positive online reviews influence their perceived usefulness and how does the perceived usefulness of these reviews influence the impact of reviews in the decision-making process? When consumers are exposed to a large amount of information on the Internet, how do they decide which piece of information is useful and therefore, will be used in their purchase decision process (Purnawirawan, et al. 2012)?


Classical consumer behavior theories such as the hierarchy of effects models (e.g., Bruner and Kumar 2000; Lavidge and Steiner 1961) and the theory of reasoned action (e.g., Ajzen and Fishbein 2000; Hansen, Möller Jensen, and Stubbe Solgaard 2004) postulate that attitudes and behavioral intentions are formed on the basis of beliefs (i.e. a link between an object and some attributes that is accessible in memory). In this perspective, recall of information and attitude are intrinsically related (e.g., Babin and Carder 1996; Gupta and Lord 1998; Hovland, Janis, and Kelley 1953; Krugman 1965) because a person's attitude towards the object is a function of his evaluations of these attributes. Vijayasarathy (2003) furthermore mentioned a set of variables in Technology Acceptance Model possible for explaining technology adoption at work, where usage at the time of technology in most cases, compulsory its user, otherwise was compelled.

Technology Acceptance Model (TAM) was introduced to explain acceptance of information technology (IT) originally. It was relied on TRA (the Theory of Reasoned Action) and its contents intention and behavior to use an information system which depends on two salient beliefs, namely perceived usefulness and ease of use. A key purpose of TAM is to provide a basis for tracing the impact of external factors on internal beliefs, attitude and intention, more TAM consist that two

particular beliefs, perceived usefulness and perceived ease of use are primary relevances for computer acceptance behaviors (Davis, Bagozzi & Warshaw, 1989). TAM also figures in a harmony link between ease of use and usefulness, proposed individual’s perception how easy or difficult to use that system will influence their perception about usefulness of that system (Vijayasarathy, 2003). David, Rodney and Allison (1989) found three main insights concerning the determinant of managerial computer use which are: firstly, people’s computer use can be predicted reasonably well from their intention, secondly, perceived usefulness is a major determinant of people’s intention to use computer, thirdly, perceived ease of use as a significant secondary determinant of people’s intention to use computer. Fishbein and Ajzen (1975) the Theory of Reasoned Action (TRA) and Davis (1989) the Technology Acceptance Model (TAM) provided theoretical context to measure beliefs, forecast future behavior. TAM model accommodate the Theory of Reasoned Action (TRA) (Fishbein and Ajzen 1975) to show acceptance information technology. TAM express that two beliefs of specific behavior, ease of use and usefulness, determine the individual intention and behavior to use a technology, where attitudes toward use directly influence the intention for the actual usage. Adam, Nelson and Todd (1992) also noted that perceived ease of use has a direct effect on both perceived usefulness and technology usage (Yulihasri, 2011).

Many researchers adopted the TAM model in e-commerce studies. Basyir (2000) adopted this model to study the various factors associated with acceptance of internet shopping behavior. Teck (2002) used TAM model for research about the impact perceived web security, perceived privacy, perceived usefulness, and perceived ease of use on the based online transaction intent. Aulvin (2000) modified the TAM model to study the individual differences such as prior web experience, shopping orientation and demographic factors that will influence the individuals’ intention to shop on the web. Fok (2001) adopted TAM to study on self-efficacy and its determinant as factors that are affecting perceived ease of use, perceived usefulness, and the use of the internet. Choong (2003) used TAM to assess owners/managers intention to adopt Web-based Supply Chain Management in SMI organizations. More recently Ramayah, Aafaqi and Jantan (2003) used TAM to predict students acceptance of a course website toward e-learning in Malaysia. Finally, Ramayah, Dahlan, Teck and

Aafaqi (2003) used TAM to predict perceived web security that influence web-based online transaction intent.

Perceived Credibility

Many previous studies on behavior also shown that trust is an important factor in online shopping. Online transactions, either through a debit card, credit card or PayPal transfer and the likes, involve a trust factor. The technology of the Internet itself has to be considered as an object of trust (Shankar, Urban, &Sultan, 2002). Prior research has suggested that online consumer reviews are likely to be more credible than seller-created information (Dellarocas, 2003; Wilson and Sherrell, 1993). On the other hand, several researchers suggested that there exist some challenges related to customer-created information (Dellarocas, 2003). These challenges are related to online identity and feedback operating perspectives. In the online identity perspective, online identity can be changed easily in an online environment. This leads to various forms of strategic manipulation. Dellarocas (2003) provided some examples in this regard: community members can build good reputations, take advantage of this by cheating other members, and then disappear and reappear under new online identities and clean records (Friedman and Resnick, 2001). Moreover, they can use fake online identities to post dishonest feedback; thus, attempting to improve their reputation or tarnish that of their competitors (Dellarocas, 2000).

The greater the perceived credibility of online consumer reviews among potential consumers, the higher is the purchase intention (Park&Kim, 2008). For instance, consumer endorsement in advertisements is of three types: celebrity, expert, and typical consumer endorsements (Fireworker and Friedman, 1977; Kelman, 1961; Friedman and Friedman, 1979). The endorsers are persuaded through the informal influence of the process of internalization (Fireworker and Friedman, 1977; Kelman, 1961; McGuire, 1969). Internalization occurs when the receiver adopts an attitude because it is useful for the solution of a problem or is demanded by his or her value system. Further, internalization occurs if reference groups are considered credible (McGuire, 1969). Consistent with this view, information from high-credibility sources is likely to be more easily accepted (Bearden and Etzel, 1982). Credibility includes expertise and trustworthiness (Hovland et al., 1953). Expertise is defined as the perceived ability of the source to make valid assertions, and trustworthiness is defined

as the perceived willingness of the source to make valid assertions (Hovland and Weiss, 1951-1952). The model indicates that sources exhibiting expertise and trustworthiness are credible and persuasive (Atkin and Block, 1983; Kamen et al., 1975; Klebba and Unger, 1983). Further, trustworthiness positively influences consumer attitudes toward a brand, consumer intentions, and their purchase behaviors (Lascu et al., 1995; Petty and Wegener, 1998). Perceived credibility is a partial mediator between eWOM quality and purchase intention (Tsao & Hsieh, 2015).

Many of these previous studies on internet acceptance were done by using the Technology Acceptance Model (TAM) developed by Davis (1989). It has been well established that online purchase behavior model can be developed from TAM. Perceived usefulness has however direct influence on intention to use. It is also the fact that behavioral intention influences the actual behavior. Therefore, this study use the theory of reasoned action (TRA) and technology acceptance model (TAM) to explain how perceived usefulness play a mediating role in affecting consumers’ online shopping intention. The research model was adapted from Vijayasarathy (2003), for two reasons: firstly, this model is an extended TAM model (Vijayasarathy, 2003), and second, this model integrates three models: TAM, TRA, and TPB. This indicates that in addition to perceived usefulness, perceived credibility should be relevant to online shopping intentions. This study investigates the mediating role of perceived usefulness and perceived credibility, as these concepts pertain to consumers’ online shopping intentions.

The hypotheses for the study are as follows:

H2: Pictures in positive online consumer reviews have an indirect positive relationship with shopping intention mediated by perceived usefulness of positive online consumer reviews, so that pictures can increase perceived usefulness of positive online consumer reviews and then increase consumers’ shopping intentions.

H3: Pictures in positive online consumer reviews have an indirect positive relationship with shopping intention mediated by perceived credibility of positive online consumer reviews, so that pictures can increase perceived credibility of positive online consumer reviews and then increase consumers’ shopping intentions.

2.6 Picture characteristics

Consumers do not follow a structured format when posting their product evaluations on the web (Park & Kim, 2008; Pollach, 2008). As a consequence, reviews range from simple recommendations that are accompanied by extremely positive or negative statements, to nuanced product evaluations that are supported by pictures taken by consumers themselves. Willemsen et al., (2011) considered content characteristics of the online consumer review are paramount to understanding the perceived usefulness and credibility of reviews. However, hardly any research has been conducted in order to catalogue differences in the pictures in online consumer reviews, or study the impact of such differences on the perceived usefulness and credibility of reviews.

To fill this research gap, this research aims at gaining a better understanding of the picture characteristics that make online consumer reviews a useful source of information. More specifically, we seek to understand how three types of review characteristics—the number of pictures, the quality of the pictures, and the content of the pictures—affect the perceived usefulness of reviews. For this objective, the study is based on dual process theories, which distinguish between two types of information processing, one of which takes relatively more effort and is more extensive than the other (Higgins, E.T. , 1999). In this research, two prominent dual process theories, namely, the heuristic systematic model (HSM) by Chaiken,(1980) and the elaboration likelihood model (ELM) by Petty and Cacioppo, (1986) are used to classify the factors that influence the helpfulness of a review into peripheral cues for heuristic information and central cues for systematic information processing. ELM distinguishes between the central route, wherein a subject considers an idea logically, and the peripheral route, wherein a subject uses preexisting ideas and superficial qualities to be persuaded (Petty& Cacioppo, 1986). ELM has been widely applied to understand how information processing by individuals leads to their decision outcomes in online environments (Lee, et. al., 2008). Based on ELM, prior studies have examined the effect of online consumer review depending on consumer skepticism (Sher, P.J., and Lee, S.H, 2009), consumer involvement(Lee, et. al., 2008), and consumer expertise (Park, D., and Kim, S, 2008).

Peripheral cues are noncontent cues used in a subjective manner in heuristic information processing (Chaiken, S., 1980). Peripheral route attitude changes are based on a variety of attitude

change processes that require less cognitive effort (Petty&Wegener, 1999). As for pictures in positive online consumer reviews, these peripheral cues are the number of pictures and the quality of the pictures.

Central cues are the arguments contained in a message and used in systematic information processing in an objective manner (Chaiken, S., 1980). Systematic information processing emphasizes detailed processing of message content and the role of message-based cognitions in deciding to accept a message’s conclusion (Chaiken, S., 1980). As for the pictures in positive online consumer reviews, the central cues are the content of the pictures, such as no relationship pictures and negative content pictures.

These following hypotheses will be tested:

H4: The relationship between pictures in positive online consumer reviews and perceived credibility and relationship between pictures in positive online consumer reviews and perceived usefulness are moderated by the number of pictures. The greater the number of pictures in positive online consumer reviews, the greater the perceived credibility and perceived usefulness and then the greater shopping intention.

H5: The relationship between pictures in positive online consumer reviews and perceived usefulness is moderated by the content of the pictures. The no relationship content of pictures in positive online consumer reviews decrease perceived usefulness and then decrease shopping intention. The negative content of pictures in positive online consumer reviews decrease the perceived usefulness and then decrease shopping intention.

H6: The relationship between pictures in positive online consumer reviews and perceived credibility and the relationship between pictures in positive online consumer reviews and perceived usefulness are moderated by the quality of the pictures. The higher the quality of the pictures in positive online consumer reviews, the greater the perceived credibility and the perceived usefulness and then the greater shopping intention.

2.7 Conceptual framework

This study is aimed at solving a practical problem in regards to how pictures in positive online reviews affect consumers’ shopping intentions. Since pictures are always worth a thousand

words (Hoffman & Daugherty, 2013), and many online shopping merchants encourage their consumers to upload pictures, this study predicts that there is a positive relationship between the pictures in online reviews and consumers’ shopping intentions. The perceived usefulness of online reviews and the perceived credibility of online reviews might be the mediators in this positive relationship. In addition, the number of pictures, the quality of the pictures, and the content of the pictures are three possible moderators in this relationships.

Figure 1: Conceptual framework

3. METHODOLOGY

To be able to test the proposed hypotheses, data were collected. This chapter describes how data have been collected. In the first part of this chapter, the sample is described. In the second part of

this chapter, the research design is described. In the third part of this chapter, the measures, the measure items and variables are clarified. In the last part of this chapter, the procedure about how data were collected is described.

3.1 The Sample

The target population for this study is all the online shopping consumers. However, only Chinese online shopping malls enable uploading pictures in online consumer reviews. Since Chinese online shopping consumers are quite representative of the world’s online shopping population and there is no significant difference among consumers from different nations, this study chose Chinese customers as research target. Samples were chosen to stratify by age and gender in order to control the real online shopping consumer situation_{（CMMS，2015). 200 randomly chosen, experienced} Chinese online shopping consumers were invited to this experiment. Experiment sample was consisted of 72 subjects aged from 14-24 (36 male and 36 female), 76 subjects aged from 25 to 34 (38 male and 38 female), 40 subjects aged from 35 to 44 (20 male and 20 female), 8 subjects aged from 45 to 54 (4 male and 4 female) and 4 aged 55-64 years old consumers (1 male and 3 female). On the 20th of November 2015, all these 200 subjects (see Table 1) filled in questionnaires.

Table 1: Descriptive analysis of subjects

3.2 Research Design

To be able to identify and explain how pictures in positive online reviews affect consumers’

Age Male Female Sum (Percentage)

14-14 36 36 72 (36%) 25-34 38 38 76 (38%) 35-44 20 20 40 (20%) 45-54 4 4 8 (4%) 55-64 1 3 4 (2%) Sum (Percentage) 99 (49.5%) 101 (50.5%) 200

experimental research is the best research design to control for other variables, to manipulate the independent variables to easily find the cause-and-effect relationship, and to help the experimenters get better results. This is a quasi-experimental, within-subjects design. The first page of the questionnaire was a statement about the study’s purpose. The second page is the product information page, consisting of an advertisement from the online store, some pictures of the product and a detailed description of the product’s features. Online consumer reviews were placed after this product information page. Six different sets of online consumer reviews were developed for each condition. Every subject read all six online consumer reviews, which can reduce the random individual difference in different groups on condition that every consumers’ feelings are different and consumers’ criteria of the scale level are subjective. Each review was presented on a separate page, and subjects could not return to previous pages to ensure that they would read the reviews in the intended sequence. More specifically, Condition 1 contained online consumer review without pictures, Condition 2 contained online consumer review with one high quality picture, Condition 3 involved online consumer review with three high quality pictures, Condition was online consumer review with three low quality pictures, Condition was online consumer review with no relationship content pictures Condition 6 shown online consumer review with negative pictures. Condition 1 worked as the control condition, Condition 2, 3 worked as both control conditions and experiment conditions and Condition 4,5,6 were as the experiment conditions.

This experiment used a men’s white shirt as the sample product for two reasons: first, according to the IMMS survey, clothing is the most common product in the online shopping market, and second, men’s white shirts are the products that all ages may purchase. After reading each online consumer review, subjects indicated their shopping intentions, perceived credibility and/or usefulness by answering 2 or 3 questions. All questions are shown in appendix I Questionnaire.

Online shopping intentions

The consumer’s online shopping intentions, perceived credibility and perceived usefulness were the three main test variables in this study. First, in order to test how pictures in a positive online consumer review affect online shopping intentions, every condition contained questions about the consumer’s shopping intentions. MacKenzie, Lutz, and Belch (1986) used a three-item, seven-point

scale (likely/unlikely, probable/improbable, and possible /impossible) to measure shopping intention, and this study adapted this same scale. This questionnaire used a seven-point Likert-type scale ranging from 1(no purchase intention/lowest degree/no desire) to 7 (strong purchase intention/highest degree/strong desire), three different questions (“How would you describe your intention regarding the purchase of his shirt?”, “To what degree of desire do you want to own this shirt?” and “How likely is it that you will buy this shirt?”) to measure the consumer’s shopping intentions after reading each online consumer reviews. Perceived credibility and perceived usefulness were measured by multidimensional measures which appeared in communication literature in 1966. This questionnaire measured perceived credibility by this method: after subjects read every review, subjects were asked to choose three words to describe their feeling about the reviews. A seven-point scale will be used to stand different descriptive words (1 being fake/undependable/not credible and 7 being true/ dependable/credible). This questionnaire tested perceived usefulness in the same way, only changing the descriptive words into (1 being useless/unnecessary/unhelpful and 7 being useful/beneficial/ helpful). Numbers 1 and 7 indicate a very strong feeling. Numbers 2 and 6 indicate a strong feeling. Numbers 3 and 5 indicate a fairly weak feeling. Number 4 indicates the respondent is undecided. And to make sure all of the respondents can understand all of the questions and express their real feelings,

this questionnaire was translated into Chinese. 3.3 The procedure

3.3.1 Pilot Study

Before distributing the questionnaires to the subjects, a pilot study was conducted on the 28th of October 2015, inviting five male online shopping consumers and five female online shopping consumers in different ages to fill the questionnaires. These ten consumers participated voluntary, no incentives were offered. All ten questionnaires were completely finished, and all the volunteers said they could easily understand this questionnaire and consider it to be clear and short. SPSS 22.0 was used to calculate the reliability for the scales. All questions were clear and therefore no amendment was made before the main study.

Experiments were done through an online survey, which was created and organized on www.sojump.com. Before starting distribution, I have invited 200 Chinese online shopping consumers from different ages and sexes to form the sample. To be more precise, the sample is formed by 72 subjects aged from 14-24 (36 male and 36 female), 76 subjects aged from 25 to 34 (38 male and 38 female), 40 subjects aged from 35 to 44 (20 male and 20 female), 8 subjects aged from 45 to 54 (4 male and 4 female) and 4 aged 55-64 years old subjects (1 male and 3 female). On 30th October, 2015, I send the questionnaires to these 200 subjects through my personal Wechat friend circle. Within ten days, on 10th November, 2015 all the subjects filled out the questionnaire. Participation was completely voluntary and responses were strictly confidential.

4.RESULTS

In this chapter, the reliability of scales and hypotheses were tested by doing analyses in SPSS. 22.0. For a period of ten days, the survey was distributed online using Sojump. This resulted in a database of 200 subjects. In order to find if there were any errors in the data, frequency of all these data has been checked. No error has been examined. Second step is checking missing data, no missing data has been found which means all the 200 subjects answered every question in the questionnaire.

4.1

Reliability

In order to make sure all these answers are meaningful, constructs with multi-item scales had to be computed. Therefore, the Cronbach’s alpha of these variables was tested, in order to verify if all the items in one scale measure the same, or if some questions should not be used for analysis. Normally, a Cronbach’s alpha of >0.70 is considered acceptable, and if a Cronbach’s alpha of >0.80 means the multiple scale is accurate. According to the Table 2, Cronbach’s alpha of shopping intention when no picture, perceived credibility when no picture, perceived usefulness when no picture, shopping intention when one picture, perceived credibility when one picture, perceived usefulness when one picture, shopping intention when three pictures, perceived credibility when three pictures, perceived usefulness when three pictures, shopping intention when three low quality pictures,perceived credibility when three low quality pictures, perceived usefulness when three low quality pictures, shopping intention when three no relationship pictures, perceived usefulness when

three no relationship content pictures, shopping intention when three negative content pictures and perceived usefulness when three negative content pictures were all above 0.80, and most of them were above 0.90. Therefore, all these scales were reliable and this answers could be used to analyze the hypotheses.

Table2: Reliability of scales

4.2 Correlation check

In addition to the reliable analyze, Pearson’s correlation coefficient was calculated to determine if these variables were related. Table 3 shows that almost all these variables are related and some relationships are especially significant (p<0.05). Besides, the means and standard deviations of all variables was calculated and are exhibited in table 3 as well.

Construct N  of  Items Cronbach’s  Alpha*

Shopping    Inten:on  when  no  picture 3 0.924 Perceived  credibility  when  no  picture 3 0.949 Perceived  usefulness  when  no  picture 3 0.881 Shopping    Inten:on  when  one  picture 3 0.962 Perceived  credibility  when  one  picture 3 0.979 Perceived  usefulness  when  one  picture 3 0.971 Shopping    Inten:on  when  three  pictures 3 0.987 Perceived  credibility  when  three  good    pictures 3 0.891 Perceived  usefulness  when  three  good  pictures 3 0.971 Shopping    Inten:on  when  three  low  quality  pictures 3 0.961 Perceived  credibility  when  three  low  quality  pictures 3 0.959 Perceived  usefulness  when  three  low  quality  pictures 3 0.988 Shopping    Inten:on  when  three  no  rela:onship  pictures 3 0.891 Perceived  usefulness  when  three  no  rela:onship  pictures 3 0.979 Shopping    Inten:on  when  three  nega:ve  pictures 3 0.969 Perceived  usefulness  when  three  nega:ve  pictures 3 0.959 Cronbach’s  Alpha*  should>0.80

4.3 Model testing

In order to determine whether pictures in positive online consumer reviews affect consumers’ online shopping intentions, whether perceived credibility and perceived usefulness play a mediating role in this relationship, and whether the number of pictures, the quality of the pictures and the content of the pictures are three moderators in this relationship, the hypotheses were tested. Statistical tests in SPSS were used to test whether the expected associations in the hypotheses were significant. In this study, hypotheses were tested at a significance level of p<0.05, meaning that results for the tested hypotheses with p-values higher than 0.05 were rejected.

To test Hypothesis 1, the dependent variable was pictures in positive online consumer reviews (positive online reviews without picture=0, positive online reviews with pictures=1) and the independent variable was consumers’ shopping intentions (measured by seven-point Likert scale, 1 means no shopping intention and 7 means strong shopping intention). A Paired-Sample T-Test analysis was conducted. As suggested by H1, consumers’ shopping intention of positive online reviews with one picture (Mean=4.085) is higher than those without pictures (Mean=3.375). As H1 was a directional hypothesis, the difference was statistically significant at 95% level of confidence(t=-4.669, p=0.000). Table 4 provided an overview over the T-Test analysis. What’s more, according to Frequency description, 25% subjects chosen 3 point when seeing positive online reviews without pictures, and 28% subjects chosen 5 point when seeing online reviews with one picture. The difference of the two mode was 2 point, which indicated a significant difference in consumers’ shopping intention. Therefore, H1 was supported.

Table 4: H1. Difference in consumer’s shopping intentions of positive online consumer reviews without pictures and of positive online reviews with one picture

** Statistically significant at the 0.05 level.

Without Pictures With One Picture Difference t-value p Mean Mode SD Mean Mode SD Mean Mode

Shopping

Note. Mean and Mode values on a 7-point scale, where 1indicated “no shopping intention”

and 7 indicated “strong shopping intention”.

After making sure the relationship between independent variable and dependent variable was existed correctly, mediators would be analyzed in the next step. Hypothesis 2 stated that pictures in positive online consumer reviews have an indirect positive relationship with shopping intention mediated by perceived usefulness of online reviews (measured by seven-point Likert scale, 1 means incredible and 7 means credible), so that pictures can increase perceived usefulness of online reviews and then increase consumer’s shopping intentions. Perceived usefulness is the mediator which explains why independent variable pictures in online reviews and dependent variable shopping intention are related. In addition to the perceived usefulness, perceived credibility is another mediator in this relationship. Hypothesis 3 assumed that pictures in positive online consumer reviews have an indirect positive relationship with shopping intention mediated by perceived credibility of positive online consumer reviews, so that pictures can increase perceived credibility of positive online consumer reviews and then increase consumer’s shopping intentions.

Hypothesis 2 and Hypothesis 3 were tested using Paired Sample T-Tests. This test was most appropriate to test the difference in perceived usefulness and perceived credibility between positive online consumer reviews without pictures and positive online consumer reviews with one picture. In Table 5, the differences in perceived usefulness are shown. Significant differences in perceived usefulness between positive online consumer reviews without pictures and positive online consumer reviews with one picture were shown. Positive online consumer reviews with one picture were perceived as more useful than positive online consumer reviews without pictures (p<0.05). Specifically, subjects perceived positive online consumer reviews with one picture (Mean=4.350, SD=1.469) as more useful (Mean Difference=0.975) than positive online consumer reviews without pictures (Mean=3.375, SD=1.738), t(199)=-7.358, p=0.000. While comparing the modes in these two conditions, 18% subjects choose 2 point when seeing positive online consumer reviews without pictures, and 31.5% subjects choose 5 point when seeing online reviews with one picture. The difference in consumers’ shopping intention was significant (Mode Difference=3). Therefore, hypothesis 2 is supported. Besides, results in Table 6 show that, on average, subjects perceived

positive online consumer reviews with one picture (Mean=4.315, SD=1.454) as more credible (Mean Difference=1.045) than positive online consumer reviews without pictures (Mean=3.270, SD=1.571), t(199)=-8.564, p=0.000. As for the Mode difference, 25.5% subjects choose 3 point when seeing positive online reviews without pictures, and 28% subjects choose 5 point when seeing online reviews with one picture. The difference of the modes was 2 points, which indicates a significant difference in consumers’ shopping intention. So, hypothesis 3 was proven. To sum up, pictures in positive online consumer reviews have an indirect positive relationship with shopping intention mediated by the perceived usefulness as well as the perceived credibility.

Table 5: H2. Difference in consumer’s perceived usefulness of positive online consumer reviews without pictures and of positive online reviews with one picture

** Statistically significant at the 0.05 level.

Note. Mean and Mode values on a 7-point scale, where 1indicated “useless” and 7 indicated “useful”.

Table 6: H3. Difference in consumer’s perceived credibility of positive online consumer reviews without pictures and of positive online reviews with one picture

** Statistically significant at the 0.05 level.

Note. Mean and Mode values on a 7-point scale, where 1indicated “incredible” and 7 indicated “credible”.

Hypothesis 4 was tested using a One-way ANOVA to test whether the relationship between the pictures in positive online consumer reviews and the perceived credibility and the relationship between pictures in positive online consumer reviews and the perceived usefulness are moderated by

Without Pictures With One Picture Difference t-value p Mean Mode SD Mean Mode SD Mean Mode

Perceived

usefulness 3.375 3 1.738 4.350 4 1.469 0.975 1 -7.358 0.000**

Without Pictures With One Picture Difference t-value p Mean Mode SD Mean Mode SD Mean Mode

Perceived

the number of pictures. According to Hypothesis 4, the greater the number of pictures in positive online consumer reviews, the greater the perceived credibility and perceived usefulness and then the greater shopping intentions. Table 7 shows the differences in scores for shopping intentions, perceived credibility and perceived usefulness of different number of pictures in positive online consumer reviews. Significant differences between subjects in scores for shopping intention of online consumer reviews with one picture (Mean=4.085, SD=1.438) and of online consumer reviews with three picture (Mean=4.785, SD=1.431) were shown. There was a statistically significant effect of the number of pictures in positive online consumer reviews on consumer’s shopping intention, F(1,398) = 23.803, p<0.05. Furthermore, subjects perceived online consumer reviews with three pictures more credible than online consumer reviews with one picture (with three pictures: M=4.740, SD=1.602, with one picture: M=4.315, SD=1.455). The difference was significant as well (F(1.398)=7.716, P<0.05). On the other hand, subjects perceived online consumer reviews with three pictures more useful than online consumer reviews with one picture (with three pictures: M=4.690, SD=1.639, with one picture: M=4.350, SD=1.469). The difference was also significant (F(1.398)=4.771, P<0.05). On the basis of the above statistics, the average scores of both consumer’s shopping intentions, perceived credibility and perceived usefulness of online consumer reviews with three pictures were significant higher than the scores of online consumer reviews with one picture. Therefore, Hypothesis 4 was supported, the relationship between pictures in positive online consumer reviews and perceived credibility and perceived usefulness are moderated by the number of pictures.