What makes a useful online review?
Implication for travel product websites
Beijing Blasacapital Ltd
Hospitality and Food Management
School of Hospitality and Tourism Management
University of Surrey
Guildford, Surrey, GU2 7XH United Kingdom
Liu, Z., & Park, S. (2015). What makes a useful online review? Implication for travel product
websites. Tourism Management, 47, 140-151.
While the proliferation of online review websites facilitate travellers’ ability to obtain information
(decrease in search costs), it makes it difficult for them to process and judge useful information
(increase in cognitive costs). Accordingly, this study attempts to identify the factors affecting the
perceived usefulness of online consumer reviews by investigating two aspects of online
information: (1) the characteristics of review providers, such as the disclosure of personal identity,
the reviewer’s expertise and reputation, and (2) reviews themselves including quantitative (i.e.,
star ratings and length of reviews) and qualitative measurements (i.e., perceived enjoyment and
review readability). The results reveal that a combination of both messenger and message
characteristics positively affect the perceived usefulness of reviews. In particular, qualitative
aspects of reviews were identified as the most influential factors that make travel reviews useful.
The implications of these findings contribute to tourism and hospitality marketers to develop more
effective social media marketing.
Keywords: Travel products; online reviews; usefulness of online reviews
Online reviews have become an important information source that allow consumers to search
for detailed and reliable information by sharing past consumption experiences (Yoo & Gretzel,
2008; Gretzel, Fesenmaier, Lee, & Tussyadiah, 2011). According to the report by Vlachos (2012),
about 87 percent of international travellers have used the Internet for planning their trips and 43
percent of them have read reviews by other travellers. More specifically, nearly half of online
consumers indicated that they actively read and post reviews after experiencing service products
(Santos, 2014). This study argues that consumer reviews are particularly important in purchasing
experiential goods (e.g., destinations, hotels, and restaurants) because people find it difficult to
assess the quality of the intangible products before consumption. Hence consumers tend to rely on
online comments (a form of e-word of mouth) that allow them to obtain sufficient information and
have indirect purchasing experiences so as to reduce their level of perceived uncertainty (Ye, Law,
Gu & Chen, 2011).
Given the recognition of the importance of online reviews, several online communities
(e.g., Tripadvisor, Yelp, Citysearch, and Virtualtour) that provide a platform showcasing
consumer reviews have gained popularity and in turn have become the leading information source
in tourism and hospitality. In this vein, many studies have investigated the effect of online reviews
on travel behaviours (Vermeulen & Seegers, 2009) and product sales (Duverger, 2013; Racherla,
Connolly & Christodoulidou, 2012; Sparks & Browning, 2011). It is important to recognize that
while the abundance of online consumer reviews in travel-related social communities makes it
easy for travellers to find information, it is difficult for them to process and judge useful
information. More specifically, the extensive travel information available through social media
enables people to spend lower costs/efforts that stimulate the search for information online.
However, many individuals have a limited capability to process a substantial amount of
information, which may bring about information overload (Frias, Rodriguez, & Castaneda, 2008).
In other words, the tendency is to decrease in search costs but increase in cognitive costs (Bellman,
Johnson, Lohse & Mandel, 2006). Thus identifying the factors that generate the perceived
usefulness of online reviews is a crucial issue in online tourism marketing as online sites with
more useful reviews offer greater potential value to customers and contribute to building their
confidence in making a purchase decision (Sussman & Siegal, 2003).
This study explores two key elements in online reviews including the characteristics of review
providers and of consumer reviews themselves in order to predict perceived usefulness. Within
the online environment, offering limited cues of peer recognition and a disclosure of personal
information (e.g., real photo, name and address) and online reputation in the community have a
large influence on the way consumers respond to messages (Forman, Ghose, & Wiesenfeld, 2008).
Furthermore, the extant literature has discussed the importance of the numerical ratings of the
reviews assigned by readers and examined their effects on the purchase decision process (Poston
& Speier, 2005), search costs (Todd & Benbasat, 1992), and product sales (Duan, Gu & Whinston,
2008). However, the authors of this research argue that such quantitative characteristics of online
reviews can explain a partial aspect of review effectiveness due to limited cognitive cues for
recipients to identify the differences between numerous reviews. This study thus suggests a
research approach combining not only quantitative (i.e., length of reviews and star ratings)
elements of information but also qualitative/textual (i.e., perceived enjoyment and readability of
reviews) aspects to better explain the perceived usefulness of online reviews (Mudambi & Schuff,
2010; Van der Heijden, 2003).
Therefore, the purpose of this research is to identify factors important to reviewers (i.e.,
identity disclosure, expertise, and reputation) and composing online reviews (i.e., quantitative
including star ratings and length of reviews, and qualitative measurements containing perceived
enjoyment and review readability), both of which affect information evaluation online. In order to
address the research purpose, this study analyses over 5,000 online reviews of travel products
posted by online consumers. The findings of this research contribute to the ability of tourism and
hospitality marketers to develop more effective online marketing strategies.
2. Literature Reviews
2.1 Online consumer reviews
The studies regarding online consumer reviews have focused their attention on two aspects:
(1) the consumer decision making process and (2) product sales. The research into the effect of
online reviews on consumer purchasing behaviour has largely discussed the concepts of trust and
credibility in online reviews, based on the assumption that the uncertainty of product quality exists
in the online environment. For example, when online consumers view a product listing on a
shopping website (e.g., auction and retail websites), they may not have easy access to information
about the ‘true’ quality of the product and therefore may not be able to judge precisely a product’s
quality prior to its purchase (Fung & Lee, 1999). The difference between the information sellers
and buyers possess refers to information asymmetry (Pavlou & Dimoka, 2006). Since online
consumers have to rely on the electronic information with a lack of ability to physically investigate
the product, people are likely to involve additional risk along with the incomplete or distorted
information provided by online sellers (Lee, 1998). Ba and Pavlou (2002) identified that the
appropriate feedback mechanisms, including positive and negative ratings, induce trust in online
sellers and mitigates information asymmetry by reducing transaction risks.
Mudambi and Schuff (2010) investigated the helpfulness of reviews based on the statement
that helpfulness as a measure of perceived value in the decision-making process reflects
information (i.e., online review) diagnosticity. They identified the positive effects of review depth
(or elaborateness) on the perceived helpfulness of the review. Baek, Ahn, and Choi (2012)
investigated review credibility by performing sentiment analysis for mining review text. Based on
the dual process theory, they found that consumers tend to focus on different information sources
of reviews. Specifically, peripheral cues (i.e., star ratings and rankings of the reviews) are useful
in the information search stage whereas central information processing (i.e., the number of total
words in a review and the number of negative words) is influential in the stage of the evaluation
of alternatives. Hu, Liu, and Zhang (2008) concluded that online consumer reviews infer product
quality and reduce product uncertainty, in turn aiding the final purchase decision.
A number of previous researchers have examined the relationship between online reviews and
product sales. Chevalier and Mayzlin (2006) found a positive relationship between consumer
reviews on retailing websites and book sales (e.g., Barnes & Nobel and Amazon.com). They also
identified that the valence (average numerical rating) (Dellarocas, Zhang & Awad, 2007) and
number of online consumer reviews (Duan et al, 2008) are vital predictors of box office sales.
Clemons, Gao, and Hitt (2006) showed that not only the variance of ratings but also the strength
of the most positive quartile of reviews has a significant impact on the growth of craft beers. In
addition to the features of online reviews, the effect of the degree to which review writers disclose
their identity in the evaluation of product quality and product sales has been discussed (e.g.,
Forman, et al., 2008). While a number of researchers have concluded that an improvement in the
review score leads to an increase in relative sales online, Chen and Wu (2005) and Duan et al.
(2008) demonstrated that high product ratings do not necessarily lead to increased sales. They
explained that consumers’ tastes could be heterogeneous to the extent that the ways in which
consumers use other consumers’ opinion in their purchase decision are different.
From the tourism and hospitality perspectives, several researchers have investigated the role
of online reviews in the decision making process for general trips (Xiang & Gretzel, 2010), hotels
(Sparks & Browning, 2011), and restaurants (Racherla & Friske, 2012), as well as in estimating
the market shares of travel products, such as hotels (Duverger, 2013; Ogut & Tas, 2012) and
restaurants (Zhang, Ye, Law & Li, 2010). Vermeulen and Seegers (2009) investigated the impact
of online hotel reviews on the formation of consumer consideration, and found out that online
reviews improve hotel awareness, which ultimately helps travellers to develop their consideration
set. Ye et al (2011) estimated the effect of online review features on room sales in hotels based on
the assumption that the number of reviews encompasses the linear function of sales. They
concluded that review ratings and room prices are important elements to predict room sales online.
The studies by Zhang et al (2010) analysed online information reflecting context-specific variables
in restaurants (e.g., food quality, environment and services) as well as the number of reviews and
review ratings of the online popularity of restaurants.
2.2 Perceived usefulness of online reviews
Many review websites have designed peer reviewing systems that allow people to vote on
whether they found a review useful in their decision making. For instance, Amazon.com provides
a service that displays the top two most helpful, favourable, and critical reviews posted by online
users in order to help its customers evaluate each displayed product easily. These useful votes are
generally believed not only to be an indicator of review diagnosticity to separate the useful reviews
from the rest (Mudambi & Schuff, 2010), but also to be a signalling cue for users to filter numerous
reviews efficiently (Ghose & Ipeirotis, 2008). In other words, the useful information in a review
may assist customers to evaluate the attributes of the service so as to build confidence in the source
(Gupta & Harris, 2010). That is, the possibility to make better decisions and experience greater
satisfaction with the online platforms can be increased when information seekers encounter
numerous pieces of useful information that affect their decisions. This implies that online sites
with more useful reviews offer greater potential value to customers and contribute to building
confidence in their purchase decisions.
From the online retailers’ perspective, review usefulness can be used as the primary way of
measuring how consumers evaluate a review (Mudambi & Schuff, 2010). It is beneficial for
marketers to gain knowledge about customers’ perception of information. Kumar and Benbasat
(2006) asserted that the presence of useful reviews on a website can improve its social presence.
That is, those reviews have the potential to attract consumer visits, increase length of time to stay
in the platform, and eventually increase the sales of businesses on the site. Accordingly, online
retailers are encouraged not only to provide easy access to useful messages that can create a source
of differentiation but also establish the strategic potential to improve the quality of customer
reviews to offer greater value to customers.
2.3 Factors affecting the perceived usefulness of online reviews
Previous studies which interpret the perceived usefulness of a review as a key determinant of
consumers’ intent to comply with it (Cheung, Lee & Rabjohn, 2008), have paid attention to
identifying the factors affecting information usefulness. Most studies have mainly investigated the
characteristics of consumer reviews, particularly for quantitative attributes, including review
ratings and elaborateness of reviews (e.g., counting words) (Chevalier & Mayzlin, 2006; Mudami
& Schuff, 2010). Recently, several researchers in information systems suggested the importance
of presenting a messenger’s identity (e.g., identity disclosure) in developing confidence in the
information. For example, Racherla and Friske (2012) investigated the effects of three types of
reviewer characteristics and concluded that the levels of a reviewer’s reputation and expertise
significantly affect the usefulness of online reviews. More importantly, the authors of this study
suggests that considering the qualitative elements of the online information reflects ‘source
effects’ with regard to persuasive communication (Janis & Hovland, 1959). Schindler and Bickart
(2012) stated that the message contents and styles are vital factors which make the reviews
appealing to consumers. Accordingly, the perceived enjoyment suggested by Mathwick and
Rigdon (2004) and readability of reviews proposed by Korfiatis, Garcia-Bariocanal and Sanchez-
Alonso (2012) are analysed in this study as measurements for qualitative elements of online
messages. This research attempts to estimate three aspects of online review related factors: (1)
messenger element (i.e., identity disclosure, expertise, and reputation), (2) quantitative facets of
online messages (i.e., review valence and elaborateness) and (3) the qualitative facet of the
messages (i.e., enjoyment and readability of the reviews). Detailed discussions about the
relationships between these factors and perceived usefulness are presented in the following
[Insert Figure 1 here]
2.3.1 Messenger factors
Reviewers’ identity disclosure
Online identity is defined as a social identity that an individual establishes in online
communities and/or websites. It is a way of presenting oneself that helps others find one’s personal
profile or geographic location. Although many members of the online community are concerned
that people’s identity disclosure is pertinent to privacy issues (discouraging them from allowing
themselves to be named in the text) (Nabeth, 2005), precise information about message providers
can bring salient contributions to recipients’ perception of the message.
Prior studies have demonstrated the importance of identity disclosure in online interaction,
arguing that identifiable sources enhance the efficiency of customers’ information acquisition
(Racherla & Friske, 2012). Source identity also plays a significant role in reducing customers’
uncertainty, which arises from the lack of social cues when they search for information in the
online environment (Tidwell & Walther, 2002). Sussman and Siegal (2003) emphasized that the
disclosure of reviewer identity leads to an increase in the message’s credibility and, as a result, the
source is perceived to be more useful (Kruglanski et al, 2006). Fogg et al (2001) demonstrated that
reviewers’ names and photos in the online information source have a positive relationship with
people’s perception of the credibility of websites on the basis of the information processing theory
(Mackie, Worth & Asuncion, 1990). Forman, et al. (2008) pointed out that message recipients may
use the personal information of the message creator as a heuristic cue, drawing on the evaluation
of the information provider as a cognitive device to assist them to reach judgements and actions.
Thus, reviews developed by online users who disclose their personal identities may increase the
credibility of the information source and thus the reviews could be judged as more useful.
Therefore, it is hypothesized that:
Hypothesis 1a (H1a): Reviewers’ disclosure of their identity has a positive effect on the perceived
usefulness of their reviews.
When consumers search for information to help them make a decision, they are inclined to
follow an expert’s suggestion (Gilly, Graham, Wolfinbarger, & Yale, 1998). Online information
provided by an expert is considered to be more useful and to have more influence on attitudes
toward the product and purchase intentions than non-expert ones (Harmon & Cone, 1982; Lascu,
Bearden & Rose, 1995). Bristor (1990:p.73) referred to expertise as ‘the extent to which the
reviews provided by experts are perceived as being capable of providing correct information and
they are expected to prompt reviewers’ persuasion because of their little motivation to check the
reliability of the source’s declarations by retrieving their own thoughts’. Gotlieb and Sarel (1991)
also asserted that the indicator of an ‘expert’ message is determined by the evaluation of the degree
of competence and knowledge that a message holds regarding specific topics of interests.
However, the limited online setting makes it difficult for people to make such an evaluation given
the availability of personal information (Cheung et al., 2008; Schindler & Bickart, 2005). In other
words, message providers’ social background and attributes cannot be verified for the level of
product knowledge in the online environment with limited cues. Evaluations of messengers’
expertise, therefore, have to rely on their past behaviours (Weiss, Lurie & MacInnis, 2008): for
example, the number of reviews written, information provided in response to others’ queries, and
the opinions of the present message. Cheung, et al. (2008) discovered that when a consumer seeks
comments posted by a high degree of expertise user, he/she tends to have a higher perception of
usefulness of the information. Therefore, the following hypothesis can be formulated:
Hypothesis 1b (H1b): A high level of reviewer expertise has a positive effect on the perceived
usefulness of a review.
In addition to expertise, the reputation of reviewers is a core indicator that differentiates one
reviewer from the others. As defined by Helm and Mark (2007), reputation is the extent to which
recipients believe that a reviewer is honest and concerned about others and constant in the long
term. Prior research has argued that reputation and peer recognition can significantly enhance the
degree to which information sharing influences customers’ perceptions of product value and
likelihood of recommending the product (Gruen, Osmonbekov & Czaplenski, 2006). Online
review providers who provide a substantial amount of effort writing reviews are eager for other
consumers to provide their contributions with positive feedback in the form of friendship requests
or votes (Racherla & Friske, 2012). Many review websites related to tourism and hospitality (e.g.,
Yelp.com and TripAdvisor.com), therefore, have established an online reputation system to assist
users to find the reputation of others by collecting, distributing, and aggregating the feedback
associated with the reviewers’ past behaviour in the online community (Resnick, Zeckhauser,
Friedman, & Kuwabara, 2000).
From the social psychological perspective, reputation effects essentially signal social
validation and improve credibility (Cialdini, 2001). As such, consumers can utilize reputation as
an indication of the quality of information to reduce uncertainties regarding information quality
(Helm & Mark, 2007; Resnick et al., 2000). Due to the features of the online environment
involving limited interaction compared with face-to-face communication, the reviews provided by
high-status messengers can create greater compliance in online groups, so reputation can be used
as a heuristic cue (Guéguen & Jacob, 2002). Hence, the relationship between reputation and review
usefulness is formulated as follows:
Hypothesis 1c (H1c): Reviewers’ high reputation has a positive effect on their reviews’ perceived
2.3.2 Message characteristics
188.8.131.52 Quantitative characteristics
Review star ratings
Review star ratings refer to the number of stars allocated by the reviewers, indicating their
assessment of the products/services used. Peer rating is considered as a useful cue to reflect the
extent of consumers’ attitude and in turn helps consumers to evaluate the quality of the products
(Krosnick et al., 1993). Willemsen, Neijens, Bronner, and de Ridder (2011) defined those ratings
as numeric summary statistics (overall ratings) shown in the form of five-point star
recommendations at the surface level of the review and reviewers’ general assessments of the
product. The extant literature has demonstrated an association between online review ratings and
customer behaviours (Clemons et al., 2006; Park, Lee, & Han, 2007). The findings of previous
studies can be presented in relation to two different arguments. First, a linear relationship has been
identified. Online consumer reviews with positive ratings for search goods may exert a significant
effect on consumers’ attitudes and evaluations of the experience product (Zhang, et al., 2010).
Second, online consumers usually perceive positive or negative ratings (i.e., one- and five-star
ratings) as more useful than moderate ratings (i.e., three-star ratings) (Danescu-Niculescu-Mizil,
Kossinets, Kleinberg, & Lee, 2009). Wei, Miao, and Huang (2013) found that reviews offered by
those who give lower hotel ratings tend to be perceived more helpful. Therefore, the relationship
between review ratings and review usefulness can be assumed as the following:
Hypothesis 2a (H2a): The star ratings of online reviews have a positive relationship with the
perceived usefulness of the reviews.
The concept of review content is usually defined as the extensiveness/depth of the information
offered in the review (Racherla & Friske, 2012). Essentially, online reviews are informational cues
that facilitate customers’ evaluation of specific attributes of products and services. Shelat and
Egger (2002) pointed out that the elaborateness of online reviews represents the length of the
reviews and showed a positive influence on shopping intentions. Mudambi and Schuff (2010) also
found that longer reviews include detailed product information regarding how and where the
product was purchased and used in specific contexts. Recent studies have asserted that the
elaborateness of messages can play a powerful role in the message persuasion process. That is,
online reviews with elaborate information contribute to the alleviation of customers’ uncertainty
about the product quality and lead them to develop confidence in the decision process (Johnson &
Payne, 1985). Hence, the hypothesis is formulated that:
Hypothesis 2b (H2b): Reviews with longer text have a positive effect on the perceived usefulness
of the reviews.
184.108.40.206 Messages’ qualitative characteristics
Customer perceived enjoyment
Perceived enjoyment (PEN) can be defined as the extent to which the reading and
understanding of reviews are perceived to be enjoyable in their own right, apart from any
performance consequences that may be anticipated (Davis, Bagozzi & Warshaw, 1992). It is
considered as a qualitative factor presenting individuals’ feeling of pleasure, depression, disgust,
or hate associated with a particular act (Triandis, 1980). Prior studies discussing motivation theory
have indicated that a certain human behaviour (e.g., the usage of information technology) is
determined by both intrinsic and extrinsic motivation (Davis, et al. 1992; Van der Heijden, 2003).
Perceived enjoyment has been considered as intrinsic motivation, which drives the performance
of an action that is not undertaken for any reason other than the process of performing the activity
interaction, Mattila and Wirtz (2000) stressed that consumers’ affective reaction is important as a
cognitive process to understand consumer behaviour and, in particular, emotion is essential in the
evaluation process of products. Furthermore, intrinsic motivation enhances the deliberation and
thoroughness of cognitive processing, which leads to perceptions of perceived usefulness
(Venkatesh, Speier & Morris, 2002). Based on these findings, it is assumed that the enjoyment
vote in each review represents the recipients’ perception of enjoyment and an expected correlation
with the perceived usefulness of online reviews:
Hypothesis 3a (H3a): The reviews that indicate perceived enjoyment by recipients have a positive
effect on perceived usefulness of the reviews.
Online reviews are information resources that consumers utilize to gain knowledge about
products and services. Zakaluk and Samuels (1988) stated that the extent to which an individual
requires to comprehend the product information can present the level of readability. As noted
earlier, understandability is thought to be an important qualitative factor to display the extent to
which customers accept online information on social media platforms. Thus, a readability test can
be used as a measurement of the degree to which a piece of text is understandable to readers based
on its syntactical elements and style. According to the linguistic characteristics, the method to
calculate readability is considered as a scale-based indication of how difficult a piece of text is for
readers to comprehend (Korfiatis, et al., 2012). Table 1 lists the readability tests used in the present
study. As identified by Gunning (1969), Gunning’s Fog Index (FOG) is an indication of the level
at which people who have an average high school education would be able to understand
a particular text. Similar to FOG, the Flesch–Kincaid Reading Ease Index (FRE) is a linguistic
measurement to calculate the words per sentence and syllables per word in a given text (Kincaid,
Fishburne, Rogers, & Chissom, 1975). FRE can be used to evaluate the complexity of the text in
order to determine the number of years of education that would be needed for someone to
understand the text being assessed. Another well-known readability test, namely the Automated
Readability Index (ARI), focuses on the simple approach of calculating the amount of words and
characters to evaluate the given text’s readability (Senter & Smith, 1967). The fourth method is the
Coleman–Liau Index (Coleman & Liau, 1975), which is similar to the ARI in providing a simpler
evaluation and more careful selection of review characteristics. In summary, the FRE and ARI
scores measure the reading ease of each review, whereas the CLI and FOG indexes present the
complexity of the text. Therefore, a hypothesis that the readability of the review content would
positively affect the review usefulness is formulated:
Hypothesis 3b (H3b): The readability of the review text has a positive effect on the perceived
usefulness of the review.
[Insert Table 1 here]
3. Research Methodology
3.1 Research context
This research collected data from Yelp.com in the form of online customer generated reviews.
First, Yelp.com is one of the most popular online advisory sites dedicated to allowing customers
to share and post information about tourism and hospitality products across major cities in the
world (Yelp.com, 2013). Schein (2011) pointed out that Yelp.com is the largest business-listing
website, which has amassed over 42 million reviews about local businesses, and is more popular
than any of its rivals, like TripAdvisor.com and Citysearch.com. Second, Yelp.com can avoid
language comprehension heterogeneity between informants and recipients, which might have a
negative influence on text comprehension. To measure the qualitative element of the reviews,
Korfiatis, et al. (2012) proposed that the analysis of textual characteristics is pointless when the
reader is a non-native speaker of the written language. Thus, Yelp.com allows the researchers to
obtain sufficient online review information as well as to control the confounding effect during the
3.2 Data collection
3.2.1 Research sampling
In total, 5,090 online reviews were collected during summer 2013. This research examines
online reviews of restaurants because this category constitutes the majority of the reviews on the
website (Yelp.com, 2011) and consumers are inclined to resort to signals (useful votes) for
evaluating these attributes. Moreover, it is highly recommended to obtain data from the leading
restaurant market in this research context: for example, London in the United Kingdom, and New
York City in United States. The London tourism market, including the food and hospitality sectors,
reached a total of £79.7 billion in 2013, which represents one of the best economies in Europe and
has grown twice as fast as the rest of the UK (SagitterOne, 2013). Lushing (2012) reported that
New York City is one of top five cities with the most restaurants as an indicator of the restaurant
market share. Apart from the size of the market, these two cities are also listed in Yelp’s top 100
places to eat in each country (Yelp, 2014). These statistics indicate that two selected cities
encompass large supply and demand of restaurant sectors so that the findings have limited bias
relevant toward the specific context of the research setting. The researchers of this study collected
2,500 (35 restaurants) reviews and 2,590 (10 restaurants) reviews about restaurants located in
London and New York City, respectively, in order to reduce the effect of geographical issues.
Note that, in general, the number of consumer reviews for restaurants located in New York are
relatively larger than the restaurants in London. To extract a consistent number of reviews as an
individual sample in this study, restaurants in London have been selected more than the restaurants
in New York City. Furthermore, the price level of the restaurants ranged from budget to luxury
prices, based on four categories of price groups assigned by the online review website.
The literature about consumer behaviour has suggested that prior knowledge of a
product/service affects the external information search and evaluation as well as the final decision
(Gursoy & McCleary, 2004). Thus, the restaurants selected for this research exclude national and
regional chains; rather, local restaurants are focused on to ensure a valid sample. As suggested by
Racherla et al. (2012), a restaurant’s position on the website page has a profound impact on users’
perception. The authors of this research, therefore, focused on the top-ranked businesses on the
first page due to high level of attention to those restaurants displayed on the first two pages. In
order to control the listing of restaurants in a specific order, the collection process operates in a
random way instead of dictating the targeted data in high/low rankings or alphabetical order.
3.2.2 Operationalization of data variables
This study used reviewer and review information retrieved from the selected sample shown
in Figure 2, presenting all the variables utilized for this research. To be more specific, the
dependent variable is online review usefulness (PU) measured by counting the number of online
users who voted that the reviews were useful in response to the reviews posted (Ghose & Ipeirotis,
2010; Jin, Lu & Shi, 2002). The independent variables, including the messenger’s name, address,
and real photo, are binary variables to be measured as ‘1’ if they disclose information and ‘0’
otherwise. The website requires providing background information during the registration process,
including name, address, and a photo. The researchers of this study manually checked if members
indicate their full names or just provide initials (e.g., M. T.) (see Baek et al., 2013; Forman et al.,
2008; Racherla & Friske, 2012). To determine whether the user has a real photo, we checked if
the reviewers use real photos clear enough to identify their faces or provide no animations and
screen shots (e.g., a scene and status). To verify address, we determined if a reviewer provided
information about their residence in their profiles such as city, state and country.
The number of reviews that messengers have written represents their expertise, and reviewers’
reputation is calculated by the number of their friends, fans and Elite awards (Forman et al, 2008;
Racherla & Friske, 2012; Weiss, et al., 2008). To capture the features of reviews, the number of
words in each review and the rating that the reviewer gave from one to five stars were collected
as quantitative characteristics (Chevalier & Mayzlin, 2006; Mudambi & Schuff, 2010). For
qualitative characteristics, the perceived enjoyment was estimated by checking the number of
readers who clicked the funny and cool voting systems at the end of a review (Van der Heijden,
2003). Last, four types of readability methods were taken into account for each review: ARI, CLI,
FOG, and FRE (Korfiatis, et al., 2012).
[Insert Figure 2 here]
3.3 Data analysis
This study used the TOBIT regression model to analyse the data due to the specific feature of
useful votes (dependent variable) and the censored nature of the sample (Mudambi & Schuff,
2010). The first reason for choosing the TOBIT model is that the dependent variable is bounded
in its range and the distribution is skewed in one direction. More specifically, the distribution of
the data about useful votes (dependent variable) appears to be skewed to the left side: 96 per cent
of the data is in the range between ‘0’ and ‘5’, with ‘65’ as the maximum value. Thus, this study
focused on the reviews voted lower than 6 (valid sample: N = 4,908 reviews) to control the outliers.
The second reason for applying the TOBIT model is that the regression method is a useful analysis
for estimating the relationship between a non-negative dependent variable and a set of independent
variables (Long, 1997). In a similar vein, the TOBIT model has the potential to solve the selection
bias inherent in this research context. For example, counting the number of people who have read
a certain review is hard on an online platform. Rather, researchers can recognize how many of
those online users voted that they perceived the review as useful and the number of enjoyment
votes on a review. As such, it is possible for a review to receive a zero response to the ‘useful’
votes, meaning that readers perceived that the particular review is not useful when it is read. The
OLS model, in general, is an effective method to illustrate the correlations between variables.
However, OLS regression has a potential restriction that regards a zero value as a missing value,
while it actually presents customers’ perception of reviews in this research context (Kennedy,
1994). Compared with the OLS model, the TOBIT model can present the real value of each review,
which has zero useful votes if the case meets the conditions to be regarded as zero. Based on the
censored nature of the dependent variable and the potential selection issues, this current study,
therefore, performed TOBIT regression to analyse the data and measured the fit with the likelihood
ratio and Efron’s pseudo R-square value (Long, 1997).
4.1 Descriptive analysis of the variables
Table 2 provides the descriptive statistics for the main variables of the data set. It is interesting
that most messengers provided real names (N = 4,681, 95.4%), photos (N = 3,499, 71.3%) and
addresses (N = 4,858, 99.0%), indicating that the level of identity disclosure is quite high. With
regard to the other variables of the review providers, expertise (mean = 157.65, SD = 283.45),
friends (mean = 78.68, SD = 245.25) and fans (mean = 10.67, SD = 42.17) show relatively large
variance compared with Elite awards (mean = 1.16, SD = 1.77).
Looking at the variable about useful votes, on average, each review has one useful vote to
represent the probability of evaluating the perceived usefulness of online consumers. In addition,
the average review star rating was 4.28 (SD = 0.88), with an average text length of 138.42 (SD =
119.53) words. In terms of message qualitative characteristics, the review enjoyment shows the
distribution from a minimum of 0 to a maximum of 36. The values of the readability test vary
across the four methods, FOG (mean = 9.23, SD = 3.17), ARI (mean = 6.21, SD = 3.87), CLI
(mean = 7.08, SD = 2.52) and FRE (mean = 81.62, SD = 13.05).
[Insert Table 2 here]
4.2 Hypothesis testing
Initially, a correlation analysis including all of the variables to estimate was conducted.
Correlation values shown at Table 3 indicate acceptable levels with little concerns for variables of
‘expertise’ and readability indexes (around .80). However, Tabachnick and Fidell (2007) stated
that “The statistical problems created by singularity and multicollinearity occur at much higher
correlations (.90 and higher)” (pp. 90). Furthermore, this study conducted a series of statistical
estimations for multicollinearity in regression analysis (see Appendix A).
[Insert Table 3 here]
Model 1, shown below, reflects the hypothetical relationships between messenger factors and
perceived usefulness (PU), referring to H1a, H1b, and H1c.
Model 1: PU = β11*RealName+ β12*RealPhoto+ β13*RealAddress+ β14*Expertise+ β15*Friends+
The results of the regression analysis for Model 1 are shown in Table 4 (log likelihood = -
6642.81). As this study proposed, RealName (b = .30) in a marginal manner (p < .10) and
RealPhoto (b = .66; p < .001) were statistically significant, while RealAddress (b = 1.50) was not
insignificant regarding perceived usefulness. In addition, the number of friends (b = .0007; p
< .001) and Elite awards (b = .18; p < .001) as well as fans (b = .002; p < .10) within the marginal
level positively influenced the dependent variable (perceived usefulness). These variables
explained about 3 per cent of the variance of perceived usefulness (R2 = 0.03).
[Insert Table 4 here]
Model 2 reflects the hypothetical relationships between the quantitative characteristics of review
messages and perceived usefulness, referring to H2a and H2b.
Model 2: PU = β21*RealName+β22*RealPhoto+β23*RealAddress+β24*Expertise+
Model 2 was estimated by adding the variables about the quantitative characteristics of
reviews from Model 1, including the star rating, square of the star rating and word count. All three
factors showed statistically significant results. The relationship of the star rating is interesting: the
star rating (b = .37, p < .001) and the square of the star rating (b = .16, p <.001), which implies a
positive and U-shaped pattern together, were both statistically significant. When comparing
coefficient values, this research emphasized the positive relationship. That is, online reviewers
tend to assign a higher level of usefulness to the reviews that provide positive star ratings. This
finding is consistent with studies investigated by Wei et al. (2013). Word counts (b = .003, p <
.001) positively affected the perceived usefulness of the reviews. This means that as the online
reviews included a larger number of words, consumers were more likely to perceive the usefulness
of the contents. Despite the significant results of the quantitative measurements, the number of
significant reviewer-related factors diminished into two variables: the existence of real photos (b
= 0.6028, p < .001) and the number of friends (b = .0007, p < .01). As a result, Model 2 accounted
for about 4 per cent of the variance of the perceived usefulness (R2 = 0.04) (see Table 5).
[Insert Table 5 here]
Model 3 reflects the hypothetical relationships between the qualitative characteristics of the review
message and the perceived usefulness, referring to H3a and H3b.
Model 3: PU = β31*RealName+β32*RealPhoto+β33*RealAddress+β34*Expertise+
Model 3 additionally contains enjoyment votes and the four types of readability metrics, to test
mainly the relationships between the qualitative characteristics of online reviews and the perceived
usefulness. Overall, the variables reflecting the qualitative characteristics increased considerably
by an additional 11 per cent (Pseudo R2 = 0.13; ∆R2 = 0.11) compared with Model 1. In
particular, perceived enjoyment had a greater coefficient value than any other variables, referring
to one of the most important elements to predict the dependent variable (b = .50, p < .001). Of the
four readability tests, the Flesch Reading Ease Index, positively affected the perceived usefulness
(b = .01, p < .05), whereas the Automated Readiblity Index was marginally significant (b = .02,
p < .01). Referring to the definition of the readability tests, the FOG and CLI scores are the
measurements of text complexity, whereas the FRE and ARI scores reflect the ease of reading the
reviews. It was found that review complexity has no effect on online users’ perceived usefulness
of the information; however, the ease of reading the reviews was regarded as an important element
in judging the usefulness. Interestingly, comparing the coefficient values between FRE and word
counts, it can be said that qualitative characteristics present a greater effect than merely counting
the number of words in the consumer reviews (see Table 6).
[Insert Table 6 here]
5.1 Messengers’ factors
The findings of this research reveal that reviewers’ identity disclosure has a significant impact
on review usefulness. Based on the assumption that message recipients use social information
about the source of a review as a heuristic factor to judge message providers’ reliability (Chaiken,
1987), the result of this research shows that reviews with self-disclosure are evaluated as more
useful. That is, online consumers respond more positively to reviews including social information
than non-identifiable online sources (Jarvenpaa & Leidner, 1999; Xia & Bechwati, 2008).
Moreover, we identified that the variable of expertise has no significant relationship with the
review’s usefulness. This is in contrast to the results of the previous literature that expertise
enhances message persuasiveness and credibility (Belch & Belch, 2011). One possible explanation
for this inconsistent finding may lie in the conceptualization of expertise: the definition of
expertise in this study refers to the source’s level of expertise (i.e., cues provided by a system to
present how many reviews a messenger posts on the website). It reflects a slightly different
approach from previous studies: for example, Biswas, Biswas and Das (2006) asserted that a
perceived ‘expert’ is related to a great deal of knowledge on a particular topic rather than a
generalized level of knowledge. Another explanation is that the role of the information provider’s
reputation varies according to the different goals of information seekers online between a learning
orientation and a decision-making orientation (Weiss, et al., 2008). This implies that
understanding the motivations of online information recipients to check reviews is an important
research subject recommended for future research.
Furthermore, the results regarding the number of friends, fans, and Elite awards strongly
support the notion that the messages written by the information providers with a high reputation
are perceived as more useful than reviews posted by those who have a low reputation. As asserted
in the previous sections, such a reputation mechanism can reduce the uncertainties regarding
service quality and performance because they help customers to identify whom to trust for their
decision making. This finding also emphasizes the importance of a reputation system in
maintaining the valuable consequence of peer recognition for both message creators and recipients
(Jeppesen & Frederiksen, 2006).
5.2. Messages’ quantitative characteristics
The results show that positive reviews are perceived as more useful than either negative or
moderate reviews. As explained by Russo, Meloy, and Medvec (1998), positive information in
valence evaluation that is consistent with a customer’s pre-decisional preference provides
validation for their interests in the tourism industry and is thus perceived to be more useful.
Interestingly, however, the findings of this study indicated both positive and quadratic
relationships. This suggests for the future research to investigate the links between star ratings and
review usefulness/helpfulness. Other than review ratings, the empirical results also imply that
message recipients perceive reviews with longer text to be more useful than those with shorter
text. This result is similar to the findings of recent eWOM research (e.g., Chevalier & Mayzlin,
2006; Mudambi & Schuff, 2010). Racherla and Friske (2012) concluded that longer reviews
relatively contain more information about the product, which helps consumers to obtain indirect
5.3. Messages’ qualitative characteristics
The empirical study supports the notion that perceived enjoyment and readability metrics are
two stronger determinants of customers’ perception of usefulness when evaluating online
information. To be specific, the analysis of the affective factor of enjoyment shows a greater
impact than the other variables, implying that the influence of reviews’ qualitative characteristics
on their usefulness moves beyond the impact of messengers’ social information as well as
messages’ quantitative features. Prior research highlighted that individuals seek hedonic
information in the computer-mediated environment to satisfy their entertainment purposes and
eventually to bring about actual behaviours (Atkinson & Kydd, 1997).
In addition to customer perceived enjoyment, the readability of the message content appears
to be an important predictor of reviews’ usefulness. An interesting finding is that just
measurements estimating the ease of reading reviews (ARI and FRE scores) show a significant
relationship with perceived usefulness. This finding suggests that online consumers would be
likely to search for tourism product reviews that are easy to read, which facilitates them obtaining
the specific information necessary within the overwhelming amount of reviews posted online. This
finding is coincided with the argument of Dillard, Shen and Vail (2007) that consumers would
prefer reviews that are ‘playful, useful and understandable’ rather than reviews with complicated
The implications of this study can be explained through two fields: theoretical and managerial
aspects. This research takes an important step towards identifying the factors that drive customers’
perception of the usefulness of online reviews. First, the findings of this paper highlight that review
messages’ qualitative characteristics (i.e., perceived enjoyment and readability) make greater
contributions to explaining the review usefulness beyond the other characteristics, such as
messengers’ and reviews’ quantitative factors. Moreover, this paper attempts to assess the
combination of all the factors related to online reviews, which allows the researchers to develop a
comprehensive model so as to estimate the relative importance of predicting review usefulness.
Along with the theory of information diagnosticity that refers to the extent to which the consumer
believes the product information is helpful to understand and evaluate products (Herr, Kardes, &
Kim, 1991), the findings of this research sheds light on the multi-aspects of information attributes
to improve the perceived diagnosticity. In the online environment where consumers have limited
capability to diagnose the integrity of the product information, especially for the experience goods,
the information about reviewer’s identity and previous experiences in the online community and
also message related attributes (such as the quantitative elements and features representing the
quality of online reviews) improves the perceived usefulness (i.e., perceived information
diagnosticity) (Mudambi & Schuff, 2010).
With the online environment in which online travellers are now facing a large number of
online consumer communities and reviews competing for a portion of their viewing time, the
findings of this research provide a number of practical implications to develop online marketing
strategy in tourism and hospitality. First, online community should disclose details of the
information providers’ identity, such as their names, addresses and photos as reviewers’ identity
disclosure is likely to spur other readers to perceive the usefulness of the reviews. In the same
vein, online marketers are recommended to develop a system to award the Elite rating to certain
review posters by considering not only the number of times they have left a review but also the
contents of the reviews itself so that online consumers can trust the authorized Elite award. Since
positive reviews of product experiences are more influential than negative reviews, tourism and
hospitality marketers should pay more attention on positive reviews by responding to the reviews
in effective ways. Last, online tourism marketers need to find reviews that make readers joyful
and are easy to read, and regularly update them to be displayed on the front pages.
With the growing availability and popularity of web-based opinion platforms, online
reviews are now an emerging market phenomenon that is playing an important role in consumer
decision making process. People can easily obtain a substantial amount of information from the
consumer review websites. However, with regard to bounded rationality (Payne, Bettman, &
Johnson, 1992), the marketers need to offer selected information useful to their consumers in order
to alleviate cognitive cost in the information evaluation. To do so, this study investigated three
aspects of online reviews in predicting perceived usefulness of the consumer reviews, including
(1) reviewer’s characteristics, (2) quantitative facet, and (3) qualitative facet of review
characteristics. Consequently, this research delineates specific aspects of online reviews that
consumers use to assess the usefulness of consumer generated contents. By investigating online
comments posted by actual consumers (i.e., secondary data), this study aims to provide a blueprint
for analysing the nature and role of online reviews in better understanding travel information
Atkinson, M.A., & Kydd, C. (1997). Individual characteristics associated with World Wide Web
use: an empirical study of playfulness and motivation. Data Base for Advances in Information
Systems, 28(2), 53–62.
Ba, S., & Pavlou, P. (2002). Evidence of the effect of trust building technology in electronic
markets: Price premiums and buyer behavior. MIS Quarterly, 26(3), 243-268.
Baek, H., Ahn, J. H., & Choi, Y. (2013). Helpfulness of online consumer reviews: Readers’
objectives a d review cues. International Journal of Electronic Commerce, 17(2), 99-
Belch, G., & Belch, M. (2011). Advertising and Promotion: An Integrated Marketing
Communications Perspective. 9th edn. McGraw-Hill, New York, NY, Bristor.
Bellman, S., Johnson, E. J., Lohse, G. L., & Mandel, N. (2006). Designing marketplaces of the
artificial with consumers in mind: Four approaches to understanding consumer behavior in
electronic environments. Journal of Interactive Marketing, 20(1), 21-33.
Biswas, D., Biswas, A. & Das, N. (2006). The differential effects of celebrity and expert
endorsements on consumer risk perceptions. The role of consumer knowledge, perceived
congruency, and product technology orientation. Journal of Advertising, 35(2), 17–31.
Bristor, J. (1990). Enhanced explanations of word of mouth communications: the power of
relationships. Research in Consumer Behaviour, 4, 51–83.
Chaiken, S. (1987). The heuristic Model of Persuasion in Social Influence: The Ontario
Symposium, Vol. 4, Mark P. Zanna, E. Tory Higgins, and C. P. Herman, eds. Hillsdale, NJ.
Lawrence Erlbaum Associates.
Chevalier, J.A. & Mayzlin, D. (2006). The effect of word of mouth on sales: Online book reviews.
Journal of Marketing Research, 43, 345–354.
Cheung, C. M. K., Lee, M. K. O., & Rabjohn, N. (2008). The impact of electronic word-of-mouth:
The adoption of online opinions in online customer communities. Internet Research, 18, 229–
Cialdini, R.B. (2001). Harnessing the science of persuasion. Harvard Business Review, 79(9), 72–
Clemons, E., Gao, G., & Hitt, L. (2006). When Online Reviews Meet Hyper differentiation: A
Study of the Craft Beer Industry. Journal of Management Information Systems, 23(2), 149–
Chen, P-Y., & Wu, S-Y. (2005). The impact of online recommendations and consumer feedback
on sales. Working Paper, Carnegie Mellon University, Pittsburgh, PA.
Coleman, M., & Liau, T. L. (1975). A computer readability formula designed for machine scoring.
Journal of Applied Psychology, 60(2), 283–284.
Danescu-Niculescu-Mizil, C., Kossinets, G., Kleinberg, J., & Lee, L. (2009). How opinions are
received by online communities: A case study on Amazon.com. 18th international
conference on the World Wide Web. Madrid, Spain.
Davis, F.D., Bagozzi, R.P., & Warshaw, P.R. (1992). Extrinsic and Intrinsic Motivation to Use
Computers in the Workplace. Journal of Applied Social Psychology, 22(14), 1111–1132.
Dellarocas, C., Zhang, X., & Awad, N. F. (2007). Exploring the value of online product reviews
in forecasting sales: The case of motion pictures. Journal of Interactive Marketing, 21(4),
Dillard, J., Shen, L., & Vail, R. (2007). Does perceived message effectiveness cause persuasion
or vice versa? 17 consistent answers. Human Communication Research, 33(4), 467–488.
Duan, W., Gu, B., & Whinston, A. B. (2008). The dynamics of online word-of-mouth and
product sales—An empirical investigation of the movie industry. Journal of Retailing, 84
Duverger, P. (2013). Curvilinear effects of user-generated content on hotels' market share: A
dynamic panel-data analysis. Journal of Travel Research, 52, 465-478.
Flesch, R. (1951). How to test readability. New York: Harper.
Fogg, B.J. Marshall, J., Laraki, O., Osipovich, A., Varma, C., Fang, N., Paul, J., Rangnekar, A.,
Shon, J., Swani, P., & Treinen, M. (2001). What makes Web sites credible? A report on a
large quantitative study: the SIGCHI Conference on Human Factors in Computing Systems,
Forman, C., Ghose, A., & Wiesenfeld, B. (2008). Examining the relationship between reviews and
sales: The role of reviewer identity disclosure in electronic markets. Information Systems
Research, 19(3), 291–313.
Frias, D. M., Rodriguez, M. A., & Castaneda, J. A. (2008). Internet vs. travel agencies on pre-visit
destination image formation: An information processing view. Tourism Management, 29,
Fung, R., & Lee, M. (1999). EC-Trust: Exploring the antecedent factors. Proceedings of the Fifth
Americas Conference on Information Systems, W. D. Haseman and D. L. Nazareth (eds.),
Milwaukee, WI, pp. 517-519.
Ghose, A., & Ipeirotis, P.G. (2008). Estimating the socio-economic impact of product reviews:
Mining text and reviewer characteristics. CeDER Working paper. Retrieved from
Ghose, A., & Ipeirotis. P.G. (2010). Estimating the helpfulness and economic impact of product
reviews: Mining text and reviewer characteristics: IEEE Trans. Knowledge Data
Engineering. 23(10), 1498-1512.
Gilly, M. C., Graham, J. L., Wolfinbarger, M. F., & Yale, L. J. (1998). A dyadic study of
interpersonal information search. Journal of the Academy of Marketing Science, 26(2), 83–
Gotlieb, J.B., & Sarel, D. (1991). Comparative advertising effectiveness -The role of involvement
and source credibility. Journal of Advertising, 20(1), 38–45.
Guéguen, N., & Jacob, C. (2002). Social presence reinforcement and computer-mediated
communication: the effect of the solicitor’s photography on compliance to a survey request
made by E-mail. Cyber Psychology & Behaviour, 5(2), 139– 142.
Gunning, R. (1969). The fog index after twenty years. Journal of Business Communication, 6(2),
Gupta, P., & Harris, J. (2010). How e-WOM recommendations influence product consideration
and quality of choice: A motivation to process information perspective. Journal of Business
Research, 63(9-10), 1041–1049.
Gursoy, D., & McCleary, K. W. (2004). Travelers’ Prior Knowledge and its Impact on their
Information Search Behavior. Journal of Hospitality & Tourism Research, 28(1), 66-94.
Gretzel, U., Fesenmaier, D. R., Lee, Y.-J., & Tussyadiah, I. (2011). Narrating travel
experiences: The role of new media. In R. Sharpley & P. Stone (Eds.), Tourist
experiences: Contemporary perspectives (pp. 171–182). New York: Routledge.
Gruen, T., Osmonbekov, T., & Czaplewski, A. (2006). EWOM: the impact of customer-to-
customer online know-how exchange on customer value and loyalty. Journal of Business
Research, 59(4), 449–456.
Harmon, R. R., & Coney, K. A. (1982). The persuasive effects of source credibility in buy and
lease situations. Journal of Marketing Research, 19(2), 255–260.
Helm, R., & Mark, A. (2007). Implications from cue utilization theory and signalling theory for
firm reputation and the marketing of new products. International Journal of Product
Development, 4(3), 396–411.
Herr, P. M., Kardes, F. R., & Kim, J. (1991). Effects of word-of-mouth and product-attribute
information on persuasion: An accessibility-diagnosticity perspective. Journal of Consumer
Research, 17, 454-462
Hu, N., Liu, L., & Zhang, J. (2008). Do online reviews affect product sales? The role of reviewer
characteristics and temporal effects. Information Technology and Management, 9(3), 201-
Janis, I., & Hovland, C. (1959). Personality and Persuasibility. New Haven: Yale University
Jarvenpaa, S.L., & Leidner, D.E. (1999). Communication and trust in global virtual teams.
Organisation Science, 10(6), 791–815.
Jeppesen, L., & Frederiksen, L. (2006). Why do users contribute to firm-hosted user communities?
Organization Science, 17(1), 45–63.
Jin, X., Lu, Y., & Shi, C. (2002). Distribution Discovery: Local Analysis of Temporal Rules: the
6th Pacific-Asia Conference on Knowledge Discovery and Data Mining. Taipei, Taiwan, 6-
Johnson, E. J., & Payne, J. W. (1985). Effort and Ac- curacy in Choice. Management Science,
Kennedy, P. (1994). A Guide to Econometrics. 3rd edn. Oxford, England: Blackwell Publishers.
Kincaid, J.P., Fishburne, R.P., Rogers, R.L., & Chissom, B.S. (1975). Derivation of New
Readability Formulas (Automated Readability Index, Fog Count and Flesch Reading Ease
Formula) for Navy Enlisted Personnel.
Korfiatis, N., Garcia-Bariocanal, E., & Sanchez-Alonso, S. (2012). Evaluating content quality and
helpfulness of online product reviews: The interplay of reviews helpfulness vs. review
content. Electronic Commerce Research and Applications, 11(3), 205-217.
Krosnick, J. A., Boninger, D. S., Chuang, Y. C., Berent, M. K., & Camot, C. G. (1993). Attitude
Strength: One Construct or Many Related Constructs. Journal of Personality and Social
Psychology, 65(6), 1132-1151.
Kruglanski, A. W., Chen, X. Y., Pierro, A., Mannetti, L., Erb, H. P., & Spiegel, S. (2006).
Pesuasion according to the unimodel: Implications for cancer communication. Journal of
communication, 56(1), 105-122.
Kumar, N., & Benbasat, I. (2006). The influence of Recommendations on Consumer Reviews on
Evaluations of Websites. Information System Research, 17(4), 425-439.
Lascu, D-N., Bearden, W.O., & Rose, R.L. (1995). Norm extremity and interpersonal influences
on consumer conformity’, Journal of Business Research, 32(3), 201–212.
Lee, H. G. (1998). Do electronic marketplaces lower the price of goods? Communications of the
ACM, 41(1), 73-80.
Long, J.S. (1997). Regression Models for Categorical and Limited Dependent Variables,
Thousand Oaks, CA: Sage Publications.
Lushing, D. (2012). Top 10 Cities With Most Restaurants. Newser. Retrieved from
Mackie, D., Worth, L., & Asuncion, A. (1990). Processing of persuasive in-group messages.
Journal of Personality and Social Psychology, 58(5), 812-822.
Mathwick, C., & Rigdon, E. (2004). Play, flow and the online search experience. Journal of
Consumer Research, 31(2), 324-332.
Mattila, A., & Wirtz, J. (2000). The role of preconception affect in post purchase evaluation of
services. Psychology and Marketing, 17(7), 587–605.
Mudambi, S. M., & Schuff, D. (2010). What makes a helpful online review? A study of customer
reviews on Amazon.com. MIS quarterly, 34(1), 185–200.
Nabeth, T. (2005). Understanding the identity concept of digital social environment. INSEAD
CALT-FIDIS working paper. Retrieved from http://www.calt.insead.fr.
Ogut, H., & Tas, B. K. O. (2012). The influence of internet customer reviews on online sales and
prices in hotel industry. The Service Industries Journal, 32(2), 197-214.
Park, D. H., Lee, J., & Han, I. (2007). The Effect of On-Line Consumer Reviews on Consumer
Purchasing Intention: The Moderating Role of Involvement. International Journal of
Electronic Commerce, 11(4), 125–148.
Pavlou, P., & Dimoka, A. (2006). The Nature and Role of Feedback Text Comments in Online
Marketplaces: Implications for Trust Building, Price Premiums, and Seller Differentiation.
Information Systems Research, 17(4), 392-414.
Payne, J. W., Bettman, J. R., & Johnson, J. R. (1992). Behavioral decision research: a
constructive processing perspective. Annual Review of Psychology. 43, 87-131.
Poston, R., & Speier, C. (2005). Effective Use of Knowledge Management Systems: A Process
Model of Content Ratings and Credibility Indicators. MIS Quarterly, 29(2), 221-244.
Racherla, P., Connolly, D. J., & Christodoulidou, N., (2012). What Determines Consumers'
Ratings of Service Providers? An Exploratory Study of Online Traveller Reviews, Journal
of Hospitality Marketing & Management, 22(2), 135-161.
Racherla, P., & Friske, W. (2012). Perceived usefulness of online consumer reviews: An
exploratory investigation across three services categories. Electronic Commerce Research
and Applications, 11(6), 548–559.
Resnick, P., Zeckhauser, R., Friedman, E., & Kuwabara, K. (2000). Reputation Systems.
Communications of the ACM, 43(12), 45–48.
Russo, J. E., Meloy, M. G., Medvec, V. H. (1998). Predecisional distortion of product information.
Journal of Marketing Research 35(4), 438–452.
SagitterOne (2013). London Restaurant Industry. Sagitter One. Retrieved from
Schein, A. (2011). Yelp! Inc. Hoover’s Company Records. Hoover’s. Retrieved from
Schindler, R. M., & Bickart, B. (2005). Published ‘‘word of mouth’’: Referable, consumer-
generated information on the Internet’, in Haugtvedt, C.P., Machleit, K.A. and Yalch, R.F.
(eds). Online consumer psychology: Understanding and influencing behaviour in the virtual
world. Mahwah, NJ: Lawrence Erlbaum Associates, 35–61.
Schindler, R. M., & Bickart, B. (2012). Perceived helpfulness of online consumer reviews: The
role of message content and style. Journal of Consumer Behaviour, 11, 234-243.
Senter, R. J., & Smith, E.A. (1967). Automated Readability Index. AMRL-TR-66-22.Wright
Patterson AFB, Ohio: Aerospace Medical Division.
Shelat, B., & Egger, F. (2002). What makes people trust online gambling sites? Conference
on Human Factors in Computing Systems, 852–853.
Sparks, B. A., & Browning, V. (2011). The impact of online reviews on hotel booking intentions
and perception of trust. Tourism Management, 32(6), 1310-1323.
Triandis, H. C. (1980). Values, attitudes, and interpersonal behavior. in Howe, H. and Page, M.
(eds.), Nebraska symposium on motivation. Lincoln, NE: University of Nebraska Press, 195–
Sussman, S. W., & Siegal, W. S. (2003). Informational influence in organizations: an integrated
approach to knowledge adoption. Information systems research, 14(1), 47–65.
Santos, S. (2014). 2012-2013 Social media and tourism industry statistics. Stikky Media. Retrieved
Tabachnick, B. G., and L. S. Fidell. (2007). Using multivariate statistics (6th ed.). Boston:
Allyn & Bacon.
Todd, P., & Benbasat, I. (1992). The use of information in decision making: An experimental
investigation of the impact of computer-based decision aids. MIS Quarterly, 16(3), 373-393.
Tidwell, L. C., & Walther, J. B. (2002). Computer‐mediated communication effects on
disclosure, impressions, and interpersonal evaluations: Getting to know one another a bit at
a time. Human Communication Research, 28(3), 317–348.
Van der Heijden, H. (2003). Factors influencing the usage of websites: the case of a generic portal
in The Netherlands. Information & Management, 40(6), 541–549.
Venkatesh, V., Speier, C., & Morris, M. G. (2002). User acceptance enablers in individual decision
making about technology: Toward an integrated model. Decision Sciences, 33(2), 297–316.
Vermeulen, I. E., & Seegers, D. (2009). Tried and tested: The impact of online hotel reviews
on consumer consideration. Tourism Management, 30(1), 123–127.
Vlachos, G. (2012). Online travel statistics. Info Graphics Mania. Retrieved from
Wei, W., Miao, L., & Huang, Z. (2013). Customer engagement behaviors and hotel
responses. International Journal of Hospitality Management, 33, 316-330.
Weiss, A., Lurie, N., & MacInnis, D. (2008). Listening to strangers: whose responses are v, how
valuable are they, and why? Journal of Marketing Research, 45(4), 425–436.
Willemsen, L. M., Neijens, P. C., Bronner, F., & de Ridder, A. J. (2011). Highly Recommended!
The Content Characteristics and Perceived Usefulness of Online Consumer Review. Journal
of Computer-Mediated Communication, 17(1), 19–38.
Xia, L., & Bechwati, N.N. (2008). Word of mouse: The role of cognitive personalization in online
consumer reviews. Journal of interactive Advertising, 9(1), 3–13.
Xiang, Z. & Gretzel, U. (2010). Role of social media in online travel information search.
Tourism Management, 31(2), 179-188.
Ye, Q., Law, R., Gu, B., & Chen, W. (2011). The influence of user-generated content on
traveler behavior: An empirical investigation on the effects of e-word-of-mouth to hotel
online bookings. Computers in Human Behavior, 27, 634-639.
Yelp (2011). An Introduction to Yelp: Metrics as of June 2011. Yelp. Retrieved from http://
Yelp (2013). Ten Things You Should Know About Yelp. Yelp. Retrieved from
Yelp (2014). Yelp data reveals Top 100 places to eat: Bookmark these babies now! Yelp.
Retrieved from http://officialblog.yelp.com/2014/02/yelp-data-reveals-top-100-places-to-
Yoo, K. H., & Gretzel, U. (2008). What motivates consumers to write online travel reviews?
Information Technology & Tourism, 10, 283-295.
Zakaluk, B. L., & Samuels, S.J. (1988). Readability: Its Past, Present, and Future. International
Reading Association, Newark.
Zhang, Z., Ye, Q., Law, R., & Li, Y. (2010). ‘The impact of e-word-of-mouth on the online
popularity of restaurants: a comparison of consumer reviews and editor reviews’,
International Journal of Hospitality Management, 29(4), 694-700.
Fig. 1. The proposed research model
1. Real Photo
2. Real Name
3. Real Address
1. Number of Friends
2. Number of Fans
3. Number of Elite award
Review star rating
Fig. 2. Illustration of the variables in an online consumer review
(Real name, Real address, Real photo)
(Friends, Fans and Elite award)
Review Star ratings
The readability tests
Indicates the educational grade level
required. The lower the grade, the more
readable the text.
Readability Index (ARI)
Indicates the educational grade level
required. The lower the grade, the more
readable the text.
FRE= 206.835- 1.015×
Scores above 40% indicate that the text
is understandable by literally everyone.
As the value of the index decreases, the
comprehensibility of the text becomes
Indicates the educational grade level
required. The lower the grade, the more
readable the text
Source: Gunning, 1969; Flesch, 1951; Kincaid et al., 1975; Coleman & Liau, 1975
Descriptive statistics for the variables
N = 4,908
Note: realname refer to whether the messenger has a real name; realphoto refer to whether the messenger has a real photo; realaddress refer to whether the
messenger has a real address; expertise refer to the numbers of prior reviews that messenger writes; friends refer to the numbers of friends that messenger has in
Yelp.com; fans refer to the numbers of fans that messenger has in Yelp.com; eliteaward refers to how many times that messenger achieves the Elite title in
Yelp.com; wordcount refer to the number of words in each review; star_rating refer to the ratings that reviewer post; Pevotes refer to perceived enjoyment; FOG
refers to Gunning’s FOG Index; ARI refers to Automated Readability Index; CLI refers to the Coleman-Liau Index; FRE refers to Flesch Reading Ease Index;
Results of Model 1 regression analysis
95% Conf. Interval
Note: Log likelihood=-6642.8095; ±Significance: p<0.1. *Significance: p<0.05. **Significance:
Results of Model 2 regression analysis
95% Conf. Interval
Note: Log likelihood=-6541.8392; ±Significance: p<0.1*Significance: p<0.05**Significance:
Results of Model 3 regression analysis
95% Conf. Interval
Note: Log likelihood=-5937.6205; ±Significance: p<0.1*Significance: p<0.05**Significance:
Appendix A. Collinearity Estimation Table 7
The results of VIF and tolerance in the regression model
Note: conditional index places between .001 and 9.346
To assess collinarity between explanatory variables, a series of estimations used for OLS
regression was conducted, including VIF, tolerance and conditional index. Since multicollinearity
is an issue between independent variables, not a dependent variable, it does not depend on the link
function. Accordingly, the steps applied in OLS regression are reasonable approaches in censored
regression (Belsley, Kuhand, & Welsch, 1980). Table 7 presents that VIF values of each fifteen
independent variable are below 10 and tolerance levels are over .10 as the cut-off points (Hair,
Anderson, Tatham, & Black, 1995). Furthermore, the conditional indexes are below 30 (Belsley,
et al., 1980). Thus, based on these diagnostic results, there is limited concern of collinearity
between explanatory factors in the regression model.
Belsley, D. A., Kuhand, E., & Welsch, R. E. (1980). Regression diagnostics. John Wiley & Sons,
Hair, J. F. Jr., Anderson, R. E., Tatham, R. L. & Black, W. C. (1995). Multivariate Data Analysis
(3rd ed). New York: Macmillan.