1

Negative Customer Engagement Behavior: The Interplay of Intensity

and Valence in Online Networks
Jaylan Azer
Department of Management, Adam Smith Business School, University of Glasgow,
Glasgow, UK
&
Matthew Alexander
Department of Marketing, Strathclyde Business School, University of Strathclyde,
Glasgow, UK

Abstract

Recent developments in marketing and service research highlight the potentially detrimental
impact of negative customer engagement behavior (CEB) on the attitudes and behaviors of
other actors in social networks toward service providers. Specifically, in online contexts,
negative CEB is contagious and viral in nature, with ensuing implications that may have short
and long-term financial and reputational repercussions for service providers. Nevertheless,
the extant literature predominantly captures only the negative impact of what customers say
about service providers in their negative reviews and fails to provide any understanding of
different intensity levels of negative engagement. This article marks the first attempt to
provide a more nuanced view of the impact of negative CEB. This paper empirically
investigates the impact of six distinct forms of negatively valenced influencing behavior
(NVIB) using two experimental studies in an online review setting. Our results provide new
insights into different intensity levels of NVIB and how they are moderated by the number of
positive reviews. Practically, this paper addresses one of the challenges for service providers
in managing NVIBs, centered on understanding the heterogeneity of its forms and, hence,
their different intensity levels. The results suggest that service providers use semantic tools to
detect the intensity levels of NVIB and to prioritize handling and/or mitigating the more
intense NVIBs when they occur.

Statement of contribution: Our study contributes by the first empirical more nuanced view
regarding different intensity levels of NVIB and how they are moderated through the number
of positive reviews. Accordingly, we encourage managers to use semantic tools to detect
these different intensity levels of NVIB and to prioritize handling and/or mitigating the more
intense NVIBs when they occur.

Keywords: Customer engagement behavior, digital social networks, E-WoM, experiments,

negative influence, number of reviews, online reviews, valence of reviews
 2

Introduction

In today’s markets, the explosive growth of technology-enabled interfaces has given

customers a dramatic influencing role (Harrigan et al., 2018; Hennig-Thurau et al., 2010;
Lemon and Verhoef, 2016). To capture the enhanced influencing role played by customers,
recent marketing research has identified influencing behavior as a form of customer
engagement behavior (CEB) that captures customer contributions to adjusting other actors’
knowledge, perceptions, and preferences toward a focal service provider (Jaakkola and
Alexander, 2014). Customers rely on each other for authoritative information about offerings
and accept shared influencing behavior to alleviate perceived risks and reduce their reliance
on communications provided by firms (Alexander et al., 2018). As hospitality products and
services are especially difficult to evaluate prior to actual consumption (Azer and Alexander,
2017; Wei et al., 2013), customers utilize the Internet, particularly online reviews, as a prime
source of information about services and products (Azer, 2015; Mathwick and Mosteller,
2017; Wu et al., 2016). Recent market research has revealed that 77% of customers read
online reviews before making a purchase decision while 35% adjust their accommodation
plans based on online reviews (TripAdvisor.co.uk, 2016). Therefore, customers’ influencing
behavior – whether positive or negative – has, via online reviews, the potential to impact the
attitudes and behaviors of other actors (social actors embedded within networks such as
customers, potential customers, or a collection of review readers) toward firms (Bowden et
al., 2017; Brodie et al., 2019), ultimately affecting a firm’s value, its online reputation, and
hence, its revenues (Anderson, 2012; Kumar et al., 2010; Wu et al., 2016).

The pivotal role of influencing behavior is well recognized by marketing and service
research, and firms are increasingly devoting their strategic efforts to fostering influencing
behaviors that are positive for a focal organization (Harmeling et al., 2017; Kumar et al.,
2010). However, the challenge for service providers is to develop appropriate strategies to
manage negatively valenced influencing behavior (NVIB) when it occurs, centered on
understanding the heterogeneity of its forms and, hence, their different intensity levels (Dolan
et al., 2019; Juric et al., 2016). This is because the specific performances that are required
from service providers to foster a firm’s value or avoid value destruction can be different for
each NVIB form based on the relative strength of its negative impact (Dolan et al., 2019).
Nevertheless, the extant e-WOM literature has overlooked addressing the impact of how
customers engage in different forms of NVIB in online reviews (Azer and Alexander, 2018).
 3

Thus, it predominantly captures the impact of what customers say about service providers in
their negative reviews and not the impact of how they say it (Hennig-Thurau et al., 2010),
although, according to recent research, the impact of negative reviews may differ, with some
messages having a stronger negative impact than others (Dolan et al., 2019; King et al.,
2014).

Recent research on online reviews reveals that customers’ negative influences are not
homogenous but that they either advise other actors in their reviews on what not to do (direct)
or just share their negative experience without offering explicit advice (indirect) using six
discrete forms of NVIB (Azer and Alexander, 2018). To illustrate, customers engage in
indirect NVIB by discrediting a service provider, expressing regret for choosing a focal
provider or deriding service providers based on their experiences. In engaging in direct
NVIB, customers explicitly address other actors by dissuading or warning them to avoid
transacting with focal providers based on their unsatisfactory or perilous experiences,
respectively, or by endorsing competitors thereby, encouraging other actors to transact with
over a focal provider.

Although, prior research has shown that customers are influenced by the choices and
opinions of other customers on products and services (Blasco-Arcas et al., 2014; Lee et al.,
2011) this research captures customers’ influential roles only in terms of recommending,
referring, or warning others (Blazevic et al., 2013; Jaakkola and Alexander, 2014; Verhoef et
al., 2010). However, these roles typically involve advising others and, to our knowledge, no
studies have investigated the impact of indirect NVIB. Crucially, the specific heterogeneity of
the forms of NVIB implies difference in their level of intensity. This aspect is new and has
been overlooked in the extant literature, which has studied negative reviews as homogenous
rather than considering different intensity levels of NVIBs that may result in different
negative impacts. This has caused a major limitation in this research stream with inconclusive
findings specifically, regarding the impact of negative valence compared to volume of
juxtaposed positive reviews (Cheung and Thadani, 2012; Hennig-Thurau et al., 2010; King et
al., 2014).

Therefore, to cover the mentioned research gaps, this paper draws on previous research on
CEB and e-WOM to investigate the impact of six forms of NVIB on other actors’ attitudes
and behavioral intentions toward service providers using two experimental studies. It
contributes to both literature streams by firstly providing new insights into the intensity levels
 4

of NVIB by showing the significant differences in the impact of how customers engage in
discrete forms of NVIB on other actors’ attitudes and behaviors toward service providers.
Secondly, it extends the extant knowledge of the moderating role of the number of reviews
with new findings that provide a view that goes beyond the existing understanding - which
presents mixed results - of the aggregate impact of negative reviews when accompanied by
positive reviews.

Theoretical Background

Negatively-valenced Influencing Behavior (NVIB)

Recent research focusing on CEB within online contexts has identified influencing
behavior as a form of CEB that captures customer contributions of their knowledge, skills,
time, and experience to adjust other actors’ knowledge, perceptions, and expectations about
focal service providers (Jaakkola and Alexander, 2014). Influencing behavior is manifested in
a range of communication tools, such as e-WOM (e.g., online reviews), referrals,
recommendations, blogging, and mobile apps, which are utilized to spread influence
(Jaakkola and Alexander, 2014; Kumar et al., 2010). Influencing behavior has the potential to
affect the value and performance of firms in different ways depending on its valence.
Specifically, within online contexts, NVIB is contagious and viral in nature, with ensuing
implications that may involve potential short and long-term financial and reputational
detrimental outcomes for firms, brands, and service providers (Bowden et al., 2017; Wu et
al., 2016). This paper concentrates on NVIB manifested using online reviews, which is
defined as:
‘Customer contributions of resources such as knowledge, skills, experience, and time to
negatively affect other actors’ knowledge, expectations, and perception about a focal
service provider’ (Azer and Alexander 2018, p.469).

Customers engaging in NVIB intend to affect how others feel, think, and behave toward
service providers (Jaakkola and Alexander, 2014; Zimbardo and Leippe, 1991). Hence, they
may negatively influence other actors’ attitudes and behavioral intentions about these
providers (Bowden et al., 2017; Schaefers and Schamari, 2016; Wünderlich et al., 2013). To
illustrate, social psychologists suggest that individuals’ thoughts (cognition) and feelings
(affect) form their attitudes toward a focal object (Fishbein and Ajzen, 1975). Prior marketing
research has found that customers’ attitudes are a global evaluation of a product or service
that involve their perceptions, knowledge, and experiences (Bolton and Drew, 1991;
 5

Wünderlich et al., 2013), and in the absence of experiences, expectations define customers’
attitudes (Oliver, 1980). Moreover, customers might not only affect how others think or feel
but also how they behave toward a focal object. Attitude might affect behavior at a later stage
(Maio et al., 2018; Zimbardo and Leippe, 1991). Therefore, NVIB is expected to negatively
affect other actors’ attitudes and behaviors toward service providers. However, prior research
has provided mixed findings even when studies focus on the same empirical context (Babić
Rosario et al., 2016), which, according to Hennig-Thurau et al. (2010), has led to a major
limitation in e-WOM research.

For instance, prior findings suggest that negative reviews enhance the credibility of online
forums (Doh and Hwang, 2009; Hiura et al., 2010), decrease sales (Chevalier and Mayzlin,
2006; Ho et al., 2006; Sun, 2012), and negatively impact other actors’ attitudes toward
products or brand (Lee et al., 2009). However, other studies suggest that this negative impact
occurs only when customers are exposed to a specific volume of negative reviews (Lee et al.,
2008) for lesser-known brands (Vermeulen and Seegers, 2009) or utilitarian products (Sen
and Lerman, 2007). Although the extant findings suggest a negative impact of negative
reviews on customers’ behavioral intentions (Qiu et al., 2012; Sen and Lerman, 2007; Zhang
et al., 2010), other authors have found a counterintuitive effect of negative reviews in
enhancing behavioral intentions of customers who feel a close connection toward a brand
(Wilson et al., 2017).

In addition to the mixed findings, the extant e-WOM literature captures the negative
impact of what customers say about service providers in their reviews. Building on this, it
would be further interesting to understand the negative impact of how they say it. This has
been overlooked in the extant literature, although this impact may differ based on the way
customers engage in NVIB (King et al., 2014). Customers engage in NVIB by directly
addressing other actors in their reviews, advising them what not to do (direct NVIB) or just
sharing their negative experience without explicitly advising other actors (Indirect NVIB)
(Azer and Alexander, 2018). Prior research has captured only direct roles that encompass an
explicit piece of advice (e.g., recommending, referring, and warning) (Blazevic et al., 2013;
Jaakkola and Alexander, 2014; Verhoef et al., 2010). However, no existing studies have
investigated the impact of indirect NVIB and how this impact may differ from that of direct
NVIB. In addition to the overarching difference between direct and indirect NVIB, the
specific heterogeneity of the six forms of NVIB (see Table 1) within this indirect
 6

(discrediting, regretting, and deriding behaviors) and direct (dissuading, endorsing

competitors, and warning behaviors) classification implies variable impacts, which are
relatively new to the literature and worthy of investigation. In this paper, in addition to
providing new insights about the impact of six forms of NVIB, we will extend the extant
knowledge about the moderating role of the number of reviews.

NVIBs Definitions Exemplars

Indirect NVIB
Literal report of functional details of ‘Unfortunately, the facilities haven’t been updated.
substandard service to discredit a Peeling paint, noisy, food was awful. None of the
Discrediting
service provider without explicitly staff was able to do anything without the manager’s
Behavior
addressing other actors in online approval who conveniently was never available. A
reviews truly horrible place’
Communication of emotions of regret
‘When I spent my night in this hotel it was my worst
Regretting for choosing a service provider
experience. I regretted my decision and
Behavior without explicitly addressing other
I will not stay there again’
actors in online reviews
‘TV seemed to be an Internet stream as it kept
Usage of sarcasm to deride a service buffering and played more like a slide show.
Deriding
provider without explicitly addressing Shaving light cover is lying next to bare bulb. This
Behavior
other actors in online reviews is what I can recall before my brain started to
subliminally bury the horror to protect my sanity’
Direct NVIB
Explicit advice to convince other ‘The waiter was way too busy to listen to us and
Dissuading actors not to transact with a focal brought us vegetarian food we didn’t want or order.
Behavior provider stressing on opposition to The food was greasy and expensive. No one cared
and the refusal of a focal provider that it wasn’t what we ordered. Do not eat here’
Endorsing Explicit recommendation of one or ‘If you are up for all you can eat in Port Madero, go
Competitors more competitors to other actors, over to Gourmet Porteno better than this restaurant by
Behavior service providers far’
‘WARNING!! BEWARE! Absolutely
Warning other actors of a probable HORRIFYING!! We originally planned to stay for
Warning
risk based on a perilous service 2 nights, ended up staying for one only. As we ran
Behavior
experience. away the receptionist then picked up a heavy glass
ashtray to throw at us’
Table 1: Forms of NVIB – definitions and exemplars (Azer and Alexander, 2018, pp. 477-479)
 7

Forms of Negatively valenced Influencing Behavior

Indirect NVIB: Discrediting, Deriding and Regretting Behaviors

Customers engage in indirect NVIB using three different forms, namely, discrediting,
regretting, and deriding behaviors (Azer and Alexander, 2018). Although customers engage
in the three forms of indirect NVIB without explicitly addressing other actors in their
reviews, they do that in different ways, literally (discrediting), emotionally (regretting), and
sarcastically (deriding). We argue that discrediting behavior has a more negative impact on
other actors’ attitudes and behavior toward service providers than regretting behavior while
deriding behavior should have a stronger negative impact than both types of behavior.

Customers engage in discrediting behavior by sharing their negative service experience

without explicitly advising other actors not to transact with this provider. They report in their
reviews details of substandard service in a functional way that incorporates detailed
evaluation of tangible aspects of a firm or service provider, such as service environment,
service quality, and staff (Azer and Alexander, 2018). Similarly, customers engaging in
deriding behavior report informative details of substandard service in their reviews, albeit in a
sarcastic way. One rational explanation for this, according to social psychologists, is that the
reviewer may choose to use sarcasm instead of speaking literally in order to additionally
convey a negative attitude toward service providers (Filik et al., 2016). In their reviews, they
shift the polarity of positive or negative speech to its opposite (Giora et al., 2000; González-
Ibánez et al., 2011) by writing salient, incompatible meanings, specifically known to be more
potent, retainable and memorable than literal comments (Colston, 1997; Giora, 2002).
According to social psychology research, sarcasm enhances the critical effect and, hence, the
negativity of comments compared to literal ones (Colston, 1997; Toplak and Katz, 2000).
This is because sarcasm is perceived as being more condemning than literal comments
(Bowes and Katz, 2011). Accordingly, deriding behavior may have a stronger negative
impact than discrediting behavior.

Regretting behavior differs from discrediting and deriding behavior as customers focus
mainly on communicating emotions of regret for choosing a specific provider.
Communicating emotions of regret is known to elicit an avoidance motivation (Strack and
Deutsch, 2004). Emotional expressions are very common in the context of service and
hospitality (Lee et al., 2017). Although research suggests that embedded emotions in online
reviews play a powerful role in changing customers’ evaluations (Bagozzi et al., 1999), other
 8

authors argue that emotional negative reviews decrease the degree of helpfulness of these
reviews (Stieglitz and Dang-Xuan, 2013). This is because review readers may attribute
emotions internally (to the reviewers’ personal dispositions) or externally (to the reviewed
hotels) (Kim and Gupta, 2012). Thus, they may consider expressed negative emotions in
negative reviews as reviewers’ dispositions or irrationality, and this is also viewed as bad and
unpleasant behavior (Diefendorff and Richard, 2003).

Importantly, other actors tend to consider negative reviews as less informative when
negative emotional expressions are embedded (Kim and Gupta, 2012; Lee et al., 2017). This
is because emotional negative reviews provide less diagnostic information and, thus, do not
assist the readers in understanding and evaluating the quality of the hotel services. This
implies that by engaging in regretting behavior, customers may decrease the informativity of
their reviews and, hence, decrease their impact compared to the more informative deriding
and discrediting behaviors. Moreover, on top of being informative compared to the emotional
regretting behavior, deriding behavior is expected to be more negative in its impact on other
actors compared to literal discrediting behavior since sarcastic comments are more intense,
potent, retainable, and memorable than literal ones (Bowes and Katz, 2011; Colston, 1997;
Giora, 2002). Following this theorizing, we hypothesize that:

H1: Discrediting behavior will have a stronger negative impact than regretting
behavior while deriding behavior will have a stronger negative impact than both on
other actors’ (i) attitudes and (ii) behavioral intentions toward service providers.

Direct NVIB: Dissuading, Endorsing Competitors, and Warning Behaviors

Customers engage in direct NVIB by using three forms, dissuading, endorsing
competitors, and warning behaviors. Although customers engage in the three forms of direct
NVIB by explicitly advising other actors not to transact with service providers, nuances of
stressing opposition are different within each form, encompassing literal opposition
(dissuading), opposition with alternative (endorsing competitor), and severe alarming
(warning). We argue that endorsing competitors will have a stronger negative impact than
dissuading behavior on other actors’ attitudes and behaviors toward service providers, while
warning will have a stronger negative impact compared to both types of behaviors.

Customers engage in NVIB by dissuading other actors from transacting with a focal firm
or service provider, stressing opposition to providers based on a service experience (Azer and
Alexander, 2018). By engaging in dissuading behavior, customers detail their unfavorable
 9

service experience, explicitly advising other actors not to transact with focal service
providers. The extant empirical research shows that injunctive messages are persuasive as
they include advice to other actors on what to or not to do and, hence, have a strong impact
on changing an individual’s behavior (Lee et al., 2008; Schultz et al., 2008). However,
customers may not limit their review to stressing opposition to providers; instead, they
explicitly endorse competitors (Azer and Alexander, 2018). They recommend alternative
competitors over the focal provider, attempting to persuade others to choose competitors by
making them attractive to other actors over focal providers. Prior marketing research shows
that providing customers with alternatives (competitors) can decrease their post-purchase
evaluations, such as satisfaction toward the chosen brand and the decision to continue a
service relationship (Jones et al., 2000; Lemon et al., 2002). Specifically, the attractiveness of
alternatives influences the commitment to a service relationship (Lemon et al., 2002; Yim et
al., 2007). Customers engaging in endorsing competitorsbehavior provide others with
alternatives and compare them with focal providers, making those alternatives attractive,
which is not what they do when they engage in dissuading behavior. Accordingly, we expect
a stronger negative impact from endorsing competitors than dissuading behaviors on other
actors’ attitudes and behaviors toward service providers.

Unlike endorsing competitors and dissuading behaviors, customers engage in warning

behavior to warn others from a perilous service experience that involves a severe threat or
danger assessment (Azer and Alexander, 2018). Prior research shows that customers’ feelings
of insecurity based on perilous service experiences involving threat assessments exclusively
triggers them to engage in warning behavior (Meloy et al., 2012). Customers engage in
warning behavior differently; they use capital letters and stress on warning and alarming
words (Azer and Alexander, 2018). Capital letters and alarming words play a central role in
how others perceive their alerting reviews, as suggested by labeling and advertising research
(Godfrey et al., 1983). According to health and safety studies, warning messages most likely
induce fear of detrimental consequences and, accordingly, succeed in shifting the behavior of
the receivers of these messages (Hammond et al., 2004; Stacy et al., 1993). Crucially, prior
research shows that depriving customers of their fundamental human needs (e.g., security) is
considered to be a more potent driver of the negative valence of behavior compared to
dissatisfying levels of service (Patterson et al., 2006). Accordingly, we argue that customers
warning other actors against a service provider to avoid a perilous service experience will
have a stronger negative impact on their attitudes and behavioral intentions toward these
 10

providers compared to dissuading and endorsing competitors, which are focused on

dissatisfying service experiences.

H2: Endorsing competitors behavior will have a stronger negative impact than
dissuading behavior while warning behavior will have a stronger negative impact
than both on other actors’ (i) attitudes and (ii) behavioral intentions toward service
providers.

The Moderating Role of the Number of Reviews

Customers are usually exposed to negative as well as positive influencing information

about the same offering, which, for customers, represent part of the information collection
process (Sparks and Browning, 2011). According to previous research, customers are likely
to look for negative reviews as they are perceived to be diagnostic and informative (Berezina
et al., 2015). Consequently, customers are more influenced by negative reviews, especially
about service experiences than positive ones (Lee et al., 2017; Park and Lee, 2009; Racherla
and Friske, 2012). Such assumptions are justified by psychology literature based on the
negativity bias’s concept, which claims that negative events are more potent and dominant
(Rozin and Royzman, 2001). Additionally, it is argued that customers give more weight and
are attracted more to negative than positive information (Fiske et al., 1999); therefore,
negative reviews can have more impact than positive reviews (Lee et al., 2008; Papathanassis
and Knolle, 2011). Nevertheless, prior research presents mixed results regarding the power of
positive over negative messages and their impact, with authors contrasting the relative power
of negative messages (Chevalier and Mayzlin, 2006; Heitmann et al., 2007; Hollebeek and
Chen, 2014) or positive messages (Adjei et al., 2010; Cheung et al., 2009; East et al., 2008;
Kim and Gupta, 2012) for having the stronger impact. Although negative messages shared
online are known to have a detrimental effect on a focal offering, service, or brand
(Hollebeek and Chen, 2014; Qiu et al., 2012), authors argue that a few negative messages
might serve to promote the creditability of a website or a review site (Doh and Hwang, 2009)
since it is the number of positive reviews that matters (Babić Rosario et al., 2016; Berger et
al., 2010; Wirtz et al., 2013). Importantly, these studies have studied negative reviews as
homogenous rather than considering the interplay of different levels of intensity of NVIBs
and valence. Accordingly, to approach this gap, we look at the interaction between the six
forms of NVIB and higher and an equal number of positive reviews that go along with these
forms, and we hypothesize that:
 11

H3: The number of positive reviews will moderate the impact of NVIB, but not
equally; the more intense the type of NVIB the weaker the moderation role of the
number of positive reviews on its impact on other actors’ (i) attitudes and (ii)
behavioral intentions toward service providers.

To test these hypotheses, two experiments were conducted. The following section
addresses the procedures, scenarios, manipulation checks, measurements, and data analysis of
each experiment.

Study 1: Indirect forms of NVIB and number of positive reviews

Design and Procedures

This experiment tests hypotheses H1 and H3 using a 3 (Indirect NVIB: discrediting,

regretting, and deriding behaviors) × 2 (Number of positive reviews: equal & greater than)
factorial design resulting in six scenarios (see Appendix A). The stimulant material was
developed similarly to a TripAdvisor page, showing positive and negative TripAdvisor hotel
reviews to ensure realism and believability. Also, to control for the content of reviews, we
kept all scenarios about service staff; the difference is only in the way customers engage in
indirect NVIB. The number of positive reviews for the ‘greater than’ condition was
considered to be three reviews alongside one negative review as customers usually check four
to twelve reviews before booking a hotel (TripAdvisor.co.uk, 2016). Thus, considering
respondent fatigue, this study provided respondents with a maximum of four reviews (three
positive reviews and one negative). Following the recommendations of Hair et al. (2010) for
sample size requirements (0.05 alpha, 0.8 statistical power, and large effect size), a sample of
330 individuals (cell size=55, females 62.5%, average age = 37.3 years, SD = 1.30) was
recruited through ‘M-TURK’, a specialized purchased panel provider. Therefore, the sample
was drawn randomly from a subset population of real customers. We asked MTURK to
ensure that the participants checked online reviews, especially on TripAdvisor. This was in
addition to the screening question of TripAdvisor’s usage. Using the randomization facility
provided by Qualtrics, blocks of scenarios were designed to randomly capture 55 participants
per scenario, assigned randomly (between-subjects) to conditions. The scenario realism was
tested using an item adapted from (Gelbrich et al., 2015): ‘I think the description of the
situation is realistic’ and results show that the participants found the scenarios realistic
(M=6.15, SD=1.08).
 12

Manipulation Check and Measurements

The experimental manipulations were developed using the extant definitions of the three
indirect NVIBs (Appendix A). The results of the manipulation checks resulted in dropping
five participants from each group (they failed to answer manipulation questions correctly),
leaving 300 participants (N=50/group). test for the indirect NVIB indicates the different

answer patterns between manipulations, (3) =45.3, p < .001. Similarly, the test for

number of positive reviews, (2) =68.4, p < .001.

Based on previous research on e-WOM, credibility of forum, recipient utilitarian value,

and homophily were considered to affect the acceptance and persuasiveness of e-WOM
(Ayeh et al., 2013; Babić Rosario et al., 2016; Reichelt et al., 2014). Moreover, the reaction
to negative reviews might depend not only on the mere knowledge that a service failure
occurred but also on the customer’s attribution of this failure (Albrecht et al., 2016).
Accordingly, four confounding variables were selected: attitude toward checking online
reviews (utilitarian value), perceived forum credibility, blame attributions, and attributions of
reviewers’ motives to avoid the perceived bias of the reviewers. Some of the NVIB forms
were more strongly worded than others such as warning behavior; this might have made
participants believe there was malicious intent behind these reviews. Additionally, we
deliberately hid any factor that reflected the reviewers’ age, personal picture, gender,
nationality, or social status (Babić Rosario et al., 2016; Steffes and Burgee, 2009).

After reading the scenarios, participants completed a questionnaire that comprised items to
measure dependent variables, manipulation checks, and demographic items (age and gender)
in this order. Confirmatory factor analysis (CFA) was performed on all the key scales used in
this research using AMOS23 with maximum likelihood estimation (Bagozzi and Yi, 1988).
The CFA revealed a good fit with the data (χ²/df= 1.95; CFI = .96; RMSEA = 0.02). Factors
loading and reliability of scales were all above the recommended threshold of .7 (Hair et al.,
2010) – (see Appendix B). Tests were undertaken to confirm convergent (AVE ˃ .5) and
discriminant (the maximum shared variance and the average shared variance were both less
than the value of the AVE) validity (Bagozzi and Yi, 1988). As shown in Table 2, the square
root of AVE for each construct was greater than the correlations between them and all other
constructs. The results suggest an adequate discriminant validity of all measurements (Fornell
and Larcker, 1981). Additionally, correlations among the study constructs show no threats of
 13

multicollinearity (R<.80) (Hair et al., 2010). We examined common method variance bias
with Harman’s single factor test. The results from this test show that the most variance
explained by one factor was 35%, indicating that common method bias is not a likely
contaminant of our results (Harman, 1976; Podsakoff et al., 2003).

Table 2: Correlation Matrix – Study 1

AT BE Conf Conf Conf Conf
1 2 3 4
Attitudes (AT) .88
Behavioral Intentions (BE) .457 .82
Attitudes toward checking online reviews (Conf1) .304 .255 .87
Perceived Credibility of TripAdvisor (Conf2) .329 .217 .202 .89
Attributed Motives of Reviewer (Conf3) .310 .345 .250 .215 .90
Blame Attributions toward service providers (Conf4) .290 .281 .120 .170 .201 92
Note: Italicized diagonal elements are the square root of AVE for each construct. Off-diagonal elements are the correlations
between constructs

Results

After satisfying preliminary checks of the assumption of homoscedasticity (Levene’s Test

p ˃.05) for the two dependent variables (Attitude: p = .815, Behavior: p=.708), the equality of
the entire variance-covariance matrices (Box’s Test p = .251), MANCOVA was conducted
and the results reveal a significant interaction effect between NVIB and the number of
positive reviews (Wilk’s lambda = .98, F (4, 544) =18.23, p ˂.001). Any effects for the
confounding variables were non-significant under both conditions of the positive reviews.
The interaction was significant for both attitudes and behaviors (p ˂ .001) plotted for each of
the dependent variables (see Figure 1), showing different intensity levels of indirect NVIBs.
The mean scores show a stronger negative impact of the discrediting than regretting
behaviors and a stronger negative impact of deriding behavior compared to both behaviors,
which confirms H1. As seen in Table 3, regretting, discrediting and deriding behaviors have
negative effects on other actors’ attitudes and behavioral intentions. However, in cases where
there were more positive than negative reviews, regretting behavior had less negative impact
compared to discrediting behavior, while deriding behavior showed a stronger negative
impact than both behaviors on other actors’ attitudes (Mdis=3.00, Mreg=4.00, Mder=1.80;
p<.001) and behavioral intentions (Mdis=2.89, Mreg =3.90, Mder=1.90; p<.001). Similarly, in
cases where there were an equal number of positive and negative reviews, regretting behavior
(Attitude: Mreg=3.25, Behavior: Mreg=3.00) had less negative impact compared to discrediting
behavior (Attitude: Mdis=2.15, Behavior: Mdis=2.00), while deriding behavior (Attitude:
Mder=1.57, Behavior: Mder=1.51) showed a stronger negative impact than both behaviors.
 14

Thus, H3 is confirmed as the number of positive reviews moderates the impact of indirect
NVIBs, however differently; the more intense the type of NVIB, as seen in the scores of
deriding behavior compared to the other two types of behaviors, the weaker the moderation
role of the number of positive reviews on its negative impact.

Figure 1: Interaction Effect for Dependent Variables (Study 1)

Table 3: Means of significant interaction effect on dependent variables, p<.001

Number of Positive
Dependent Variables Forms of NVIB Means
Reviews
Greater 4.01
Regretting
Equal 3.25
Greater 3.00
Attitude Discrediting
Equal 2.15
Greater 1.80
Deriding
Equal 1.57
Greater 3.90
Regretting
Equal 3.00
Behavioral Greater 2.89
Discrediting
Intentions Equal 2.00
Greater 1.90
Deriding
Equal 1.51

Study 2: Direct forms of NVIB & Number of positive reviews

Design and Procedures

This experiment tests hypotheses H2 and H3 using a 3 (direct NVIB: dissuading,

endorsing competitors and warning behaviors) × 2 (number of positive reviews: equal &
 15

greater than) factorial design, resulting in six scenarios (see Appendix A). A sample of 330
individuals (cell size=55, females 54.9%, average age = 42.5 years, SD = 1.21) was recruited
through MTurk. As with Study 1, the scenarios were designed using the same procedures and
were tested for their realism. The results show that participants found the scenarios realistic
(M=6.28, SD=1.15).

Manipulation Check and Measurements

The results of the manipulation checks left 312 participants (N=52/group). An test for

direct NVIB indicates different answer patterns between manipulations, (4) =105, p <

.001; similarly, the test for the number of positive reviews (2) =75.01, p < .001. The
same confounding and dependent variables from Study 1 were used in this experiment too.
The CFA revealed a good fit with the data (χ²/df = 1.90; CFI = .92; RMSEA= 0.03) (Hu and
Bentler, 1995). Factors loading and reliability of scales were all above the recommended
threshold of .7 (Hair et al., 2010) – (see Appendix B). Additionally, as with Study 1,
convergent (AVE ˃ .5) and discriminant validity were satisfied (Bagozzi and Yi, 1988). In
addition, the correlations among the study constructs showed no threats of multicollinearity
(R<.80) (Hair et al., 2010) (see Table 4).

Table 4: Correlation Matrix – Study 2

AT BE Conf Conf Conf Conf
1 2 3 4
Attitudes (AT) .85
Behavioral Intentions (BE) .480 .83
Attitudes toward checking online reviews (Conf1) .310 .315 .88
Perceived Credibility of TripAdvisor (Conf2) .212 .238 .222 .89
Attributed Motives of Reviewer (Conf3) .310 .360 .250 .215 .80
Blame Attributions toward service providers (Conf4) .322 .354 .110 .180 .251 81
Note: Italicized diagonal elements are the square root of AVE for each construct. Off-diagonal elements are the correlations
between constructs

Results

The results of a MANCOVA reveal a significant interaction effect between NVIB and the
number of positive reviews (Wilk’s lambda = .92, F (4, 200) =38.47, p ˂.001), and any
effects for the confounding variables were non-significant under both conditions of positive
reviews. The interaction was significant for both attitudes and behaviors (p ˂ .001) plotted for
each of the dependent variables (see Figure 2), showing different levels of intensity of direct
NVIBs. The mean scores show a stronger negative impact of endorsing competitors than
dissuading behavior, while there was a stronger impact of warning behavior compared to both
 16

behaviors, thus, confirming H2. As seen in Table 5, dissuading, endorsing competitors, and
warning behaviors have negative effects on other actors’ attitudes and behavioral intentions.
However, in cases where there were more positive than negative reviews, dissuading
behavior had less negative impact compared to endorsing competitors, while warning
behavior showed a stronger negative impact than both behaviors on other actors’ attitudes
(Mdiss=2.9, Mend=1.98, Mwar=1.12; p<.001) and behavioral intentions (Mdiss=3.00, Mend =1.90,
Mwar=1.21; p<.001). Similarly, in cases where there were an equal number of positive and
negative reviews, dissuading behavior (Attitude: Mdiss=2.15, Behavioral Intentions:
Mdiss=2.50) had less negative impact compared to endorsing competitors (Attitude: Mend=1.43,
Behavior: Mend=1.45) while warning behavior (Attitude: Mwar=1.01, Behavior: Mwar=1.01)
showed a stronger negative impact than both behaviors. The differences in means could be
attributed to the relatively extreme tone of wording customers use when engaging in warning
behavior. Thus, H3 is confirmed as the number of positive reviews moderates the impact of
the direct NVIBs, however differently. Therefore, the more intense the type of NVIB, as seen
in the scores of warning behavior compared to the other two types of behaviors, the weaker
the moderation role of the number of positive reviews on its negative impact.

Figure 2: Interaction effect for dependent variables (Study 2)

 17

Table 5: Means of significant interaction effect on dependent variables, p<.001

Number of Positive
Dependent Variables Forms of NVIB Means
Reviews
Greater 2.90
Dissuading
Equal 2.15
Greater 1.98
Attitude Endorsing Competitors
Equal 1.43
Greater 1.12
Warning
Equal 1.01
Greater 3.00
Dissuading
Equal 2.50
Behavioral Greater 1.90
Endorsing Competitors
Intentions Equal 1.45
Greater 1.21
Warning
Equal 1.01

General Discussion
Theoretical Implications

This paper advances the empirical research on negatively valenced CEB and represents
one of the first studies to provide a more nuanced view regarding different intensity levels of
NVIB and how they are moderated by the number of positive reviews, thereby contributing to
both the CEB and e-WOM literature. As opposed to the extant literature, which limits the
influence of customers on other actors to the direct recommendation of service providers or
warning other actors from transacting with focal providers (Blazevic et al., 2013; Kumar et
al., 2010), this paper provides new knowledge about negative influence that includes six
indirect (discrediting, regretting, and deriding behavior) and direct (dissuading, endorsing
competitors, and warning behaviors) NVIB. Importantly, this study extends the impact of
what customers say in their reviews to provide empirical evidence of the impact of how they
say it, thereby responding to recent research calls made in several e-WOM studies (e.g. Balaji
et al., 2016; Hennig-Thurau et al., 2010; King et al., 2014).

This paper investigates the impact of three different ways in which customers engage in
indirect NVIB that involve different intensity levels: literally (discrediting behavior),
emotionally (regretting behavior), and sarcastically (deriding behavior). The empirical results
of this paper show a stronger negative impact of deriding behavior on other actors’ attitudes
and behavioral intentions toward service providers compared to both discrediting and
regretting behaviors. Customers engage in deriding behavior, sharing informative details of
substandard service in their reviews, although in a sarcastic way. Our results extend on the
 18

results of prior research (e.g. Colston, 1997; Toplak and Katz, 2000), which show the
intensity of sarcasm compared to literal comments, by showing the intensity of derision
compared not only to literal ones (discrediting) but also to emotional comments (regretting).
The results also show a stronger negative impact of discrediting than regretting behavior.
Prior research suggests the tendency of other actors to consider negative reviews as less
informative when negative emotional expressions are embedded (Kim and Gupta, 2012; Lee
et al., 2017). However, these results addressed the impact of emotional negative reviews
compared to positive reviews. Our results, therefore, offer new insights into the impact of
emotional negative reviews, not only in comparison to positive reviews but also to other
types of indirect NVIB.

By explicitly addressing other actors in their reviews, customers advise them not to
transact with a service provider; yet, the nuances of stressing opposition to service providers
differ to encompass different intensity levels: literal opposition (dissuading), opposition with
alternative (endorsing competitor), and severe alarming of a perilous experience (warning).
Although prior research shows that injunctive messages are persuasive and have the potential
to shift a receiver’s behavior as they provide explicit advice (Lee et al., 2008; Schultz et al.,
2008), this research did not consider different intensity levels of direct NVIBs. For example,
when customers just stress opposition to service providers compared to recommending
competitors or more intensely alarming of a perilous experience. Our results show the
intensity of warning behavior compared to endorsing competitors and dissuading behaviors.
Moreover, we extend the current understanding of the negative valence of behavior by
showing the negative impact of warning behavior, which is based on a perilous service
experience that involves a danger or threat assessment compared to the impact of endorsing
competitors and dissuading behaviors that are based on dissatisfying levels of services.
Additionally, we show a stronger negative impact of endorsing competitors than dissuading
behavior, which is consistent with prior findings that suggest providing customers with
alternatives negatively impacts their decisions to continue a service relationship (Lemon et
al., 2002; Yim et al., 2007).

Additionally, this study extends the extant e-WOM research with new findings regarding
the interaction effect of the number of positive reviews on the cause-effect relationship
between different forms of negative reviews and other actors’ attitudes and behavioral
intentions toward service providers. These findings provide a view that goes beyond the
 19

existing understanding of the aggregate impact of negative reviews when accompanied by

positive reviews, which present mixed results, with authors contrasting the relative power of
negative messages (Chevalier and Mayzlin, 2006; Heitmann et al., 2007; Hollebeek and
Chen, 2014) or positive messages (Adjei et al., 2010; Cheung et al., 2009; Kim and Gupta,
2012) for having the stronger impact. However, none of these studies considered the interplay
of different levels of intensity of NVIBs and positive valence. This study extends previous
research findings by revealing that not all negative reviews have the same impact when
accompanied by positive reviews. Less intense NVIBs (direct and indirect) can be moderated
by positive reviews, especially, if these positive reviews are greater than the negative ones.
Prior research shows that a few negative messages might serve to promote the creditability of
a website or a review site (Doh and Hwang, 2009) since it is the number of positive reviews
that matters (Babić Rosario et al., 2016; Berger et al., 2010; Wirtz et al., 2013). Instead, our
results reveal that it is the intensity level of NVIB that matters. Warning, deriding, and
endorsing competitors behaviors have a stronger negative impact on other actors’ attitudes
and behavioral intentions toward service providers than other types of NVIB, even when
juxtaposed with a greater number of positive reviews, while a more favorable outcome of
greater number of positive reviews is shown when accompanied by regretting, discrediting,
and dissuading behaviors in this order (most to least favorable). In other words, there are
forms of negative influence that customers enact in their negative reviews; hence, an
aggregate view of the negative impact of negative online reviews is no longer useful.

Managerial Implications

This paper enhances the managerial understanding of negative engagement by providing

a nuanced view of the different intensity levels of NVIB and how they are moderated by
number of positive reviews. Based on the results, less intense NVIBs (regretting, discrediting,
and dissuading behaviors) can be moderated by positive reviews, especially, if these positive
reviews are greater than the negative ones. Hence, we recommend that review sites may
follow/precede them with more positive reviews to soften their negative impact. Importantly,
it is highly recommended that managers should not neglect the impact of a single negative
review, especially if it shows one of the more intense NVIBs (warning, deriding, and
endorsing competitors behaviors) being of strong negative impact, even with the presence of
a greater number of positive reviews. Hence, managers are recommended to detect these
different intensity levels on review sites using semantic tools (e.g. google alerts, SentiOne,
social mention, trackur…etc.) early on. Semantic tools can help managers to pay attention to
 20

what customers say about their products, services or brands. For example, alarming words
and capital letters to detect warning behavior, nonliteral language and sarcasm to detect
deriding behavior, and for endorsing competitors, the mention of the names of competitors
could be detected by text-link or social mention tools.

Importantly, to prevent other actors from drawing their own, negative inferences about
this service provider (Wan, 2013; Xie et al., 2016) a response to warning, deriding, and
endorsing competitors behaviors is likely to be critical, given their consistent negative impact
on other actors’ attitudes and behavioral intentions toward service providers. If time is a
challenge, it is recommended that managers respond to reviews that demonstrate warning,
deriding, endorsing behaviors, preferably in this order, which is based on their level of
intensity. The managerial response to warning behavior which is exclusively based on
perilous experiences is recommended to show concrete evidence that any insecurity issue has
been readdressed. In addition, it is highly recommended that this response should include a
managerial promise of future security to reduce potential anxiety (Meloy et al., 2012). In
response to deriding behavior, it is recommended that the managerial response include an
acknowledgement of the issue reported in the customer’s review, an apology, and a highlight
of the service provider or firm’s values, maybe in a friendly good-humoured way, as friendly
and humorous responses are known to counter the negative effects of derision (Collinson,
2002). Endorsing competitors behavior incorporates direct recommendations of competitors;
therefore, it is recommended that managers use a text-link analysis to know what customers
are saying about specific competitors and their services (Balaji et al., 2016). Importantly, it is
recommended that managers in their response to the endorsing competitors form should not
refute what customers are saying about the competitors; this is likely to communicate distrust
on the part of the service provider, and may discourage other actors from using their services
(Sparks et al., 2016). Instead, in their response to this form, managers are recommended to
highlight the aspects that encourage customers to choose them over their competitors.

Limitations and Further Research

The limitations of this study offer the potential for future research in this area. TripAdvisor
reviews were selected for the reason of appropriateness rather than representativeness
(Kozinets, 2010); future research could replicate this study using other online forums. It was
advantageous that the reviews in all the experiment’s conditions did not show the nationality,
age, name, social status, gender, or personal pictures of the reviewers as this eliminated any
 21

confounding of homophily. However, in a real-life setting, these factors are neither

anonymous nor the same in all reviews. Future research might consider manipulating these
factors and test their impact as moderators or mediators. To ensure scenario realism we
designed the scenarios similar to what customers write in real-life reviews using the six forms
of NVIB. However, customers use extreme tone of wording when engaging in some forms
such as warning behavior, which could attribute to the differences in the mean scores, despite
confounding for attributions of reviewers’ motives. Customers are exposed to different forms
of negative reviews. Managerial responses to these reviews are expected to yield favorable
customers’ inferences (Sparks et al., 2016). Future research might also measure the impact of
each form of NVIB paired with the tailored and standard managerial responses.

Service providers’ situations may change (e.g., if a hotel is refurbished) and customer
reviews may change over time as well. Future research may replicate this study by including
the time factor (e.g., when the reviews were written) in the research design. Additionally,
future research could investigate the impact of the six forms moderated by the dynamic prices
of hotel rooms or their aggregate ratings. Importantly, this paper provides results about the
impact of six types of direct and indirect NVIB and the moderating role of the volume of
accompanied positive reviews given the absence of a typology of positively valenced
influencing behavior (PVIB). It is recommended that future research provide a typology of
PVIB and consider its moderating role on the impact of the six forms of NVIB. Finally, this
paper shows the difference in the impact of distinct forms of NVIB within online reviews. It
is recommended that future e-WOM research move beyond the aggregate view of
negative/positive e-WOM to offer more nuanced views of the typologies of positive and
negative e-WOM, which is expected to enrich this research stream.
 22

Appendix A: Scenarios and Manipulation

(The stimulant material shown to participants was developed so as to appear similar in appearance to a TripAdvisor page).
You are planning a vacation and while checking Hotels at your planned
Manipulation Check
destination on TripAdvisor, you came across these reviews:

“Disgusting” In this review, the reviewer is trying to

Discrediting

a. Inform other reviewers of a bad experience

We were disgusted at the service of the staff in this hotel, very rude and b. Advise other reviewrs explicitly not to transact with a
completely unhelpful. They don’t seem to be bothered about any form of service provider
customer service.
Regretting Behavior

“Never again” In this review, the reviewer is trying to

a. Express their regret for choosing a service provider
A regretful choice, the staff was so rude and completely unhelpful. They b. Advise other reviewers explicitly not to transact with a
didn’t seem to be bothered about any form of customer service. service provider
Definitely, won’t be going back there again.

In this review, the reviewer is trying to

Deriding Behavior

“Award of the worst service!”

a. Deride a service provider based on a bad service
experience
The staff were running around like headless chickens, no attention
b. Advise other reviewers explicitly not to transact with a
whatsoever was given to us. So rude and not bothered to serve!
service provider
Congratulations! The hotel deserves the award of the worst service.
 23

Dissuading Behavior
“Don’t go there!” In this review, the reviewer is trying to
a. Inform other reviewers of a bad experience
Terrible Customer Service!! The receiptionist was very rude. The attitude b. Advise other reviewers explicitly not to transact with a
of the room service staff was unhelpful too they didn’t seem to be service provider
bothered about any form of customer service. Don’t go there!
Endorsing Competitors

“Terrible experience”
In this review, the reviewer is trying to
Behavior

a. Inform other reviewers of a bad service experience

The staff is very rude and not bothered to serve you. Go to Sofitel
b. Recommend competitors over a service provider based
instead, I’d recommend it over this place. Staff there is very helpful, so
on a bad experience
responsible and would go the extra mile to provide a better service.
Warning Behavior

“WARNING!!!!!!”
In this review, the reviewer is trying to
a. Inform other reviewers of a bad service experience
BECAREFUL! The staff is so rude and unhelpful, not bothered about
b. Warn other reviewers based on a perilous experience.
service, nor the SAFETY of your LIFE or your possessions.
BEWARNED AND AVOID LIKE A PLAGUE!!!!
 24

Equal number of positive

“Good Hotel” The number of positive reviews is ……… the negative

reviews

reviews
This is a good hotel. We think it is perfect for a few days sightseeing. a. Greater than
Staff is good and the service is too. Would book again b. Equal to

“Good Hotel”

This is a good hotel. We think it is perfect for a few days sightseeing.

Staff is good and the service is too. Would book again
Greater number of positive reviews

The number of positive reviews is ……… the negative

“Nice experience”
reviews
The hotel is perfect for a short stay. Good service and kind receptionists. a. Greater than
Close to everything, very near to the bus station too. b. Equal to

“Good stay”

Had a night stay and we had a great time. Service is good, room service
is so helpful and the receptionist too. All in all it was a good stay.
 25

Appendix B: Items and Reliability

Factor Loadings &

α/CR
Constructs and Items
Exp.1 Exp.2
.92/ .91/
Attitude (Bansal and Taylor, 1999; Cronin and Taylor, 1992; Putrevu and Lord, 1994) .95 .90
My overall feeling about this hotel can be best described as (Very Unfavorable /Very favorable) .91 .90
I think the quality of service of this hotel is (Very Poor / Excellent) .92 .91
The decision to book this hotel is considered a good one (Strongly Disagree / Strongly Agree) .90 .89
ͣ
Behavioral Intentions (Gelbrich, 2010; Park et al., 2004) .95/ .93/
.95 .90 .92 .96
Would you consider booking this hotel?
Would you recommend it to other people? .91 .95
.90 .90
I would tell other people good things about the hotel
Attitude toward Checking Online Reviews (Donthu and Gilliland, 1996; Qiu et al., .89/ .87/
.95 .92
2012)
Online reviews are helpful for my purchase decision making .81 .88
.85 .91
If I do not read online reviews prior to purchase, I will feel worried about my decision
.88 .89
I want to be sure about my purchase decisions
Perceived TripAdvisor Credibility (Qiu et al., 2012) .95/ .98/
.90 .92 .95 .90
In general, I think TripAdvisor is trustworthy
In general, I think TripAdvisor is reliable .95 .92
.90 .97
In general, I think TripAdvisor is credible
Perceived reviewers’ motives (Sen and Lerman, 2007) .85/ .87/
.90 .95 .90 .90
The motive of the reviewers is to inform others about this hotel
I feel the reviewers’ comments are based on their true experience .80 .89
.81 .83
Other motives might influence the reviewers to post these review
Blame Attribution toward service providers (Gelbrich, 2010) .91/ .93/.
.90 .95 .89 90
The service provider is responsible for the problem
The reason for the problem is something the service provider had control over .88 .91
.93 .93
The service provider could have taken action to prevent the problem but has not
Note. ͣ7-point definitely will not/definitely will Likert scale, with exception of Attitude, which was a 7-point Likert
scale labeled as shown. ᵇAll confounding items were anchored on 7-point strongly disagree/strongly agree Likert-type
scale. ¹ Standardized Loadings: all loadings were significant at p ˂ .001.
 26

References

Adjei, M., Noble, S., & Noble, C. (2010). The influence of C2C communications in online brand
communities on customer purchase behavior. Journal of the Academy of Marketing Science,
38(5), pp. 634-653. doi:10.1007/s11747-009-0178-5
Albrecht, A., Walsh, G., & Beatty, S. (2016). Perceptions of Group Versus Individual Service Failures
and Their Effects on Customer Outcomes. Journal of Service Research, 20(2), pp. 188-203.
doi:10.1177/1094670516675416
Alexander, M., Jaakkola, E., & Hollebeek, L. (2018). Zooming out: actor engagement beyond the
dyadic. Journal of Service Management, 29(3), pp. 333-351. doi:10.1108/josm-08-2016-0237
Anderson, C. (2012). The impact of social media on lodging performance. Cornell Hospitality Report,
12(15), pp. 4-11.
Ayeh, J. K., Au, N., & Law, R. (2013). Do We Believe in TripAdvisor?" Examining Credibility
Perceptions and Online Travelers' Attitude toward Using User-Generated Content. Journal of
Travel Research, 52(4), pp. 437-452. doi:10.1177/0047287512475217
Azer, J. (2015). Facebook from socializing to advertising: an empirical study on the effect of
Facebook as an advertising tool. African Journal of Business Management, 9(24), pp. 796-
813.
Azer, J., & Alexander, M. (2017). Negatively-valenced Customer Engagement Behavior: Forms and
Drivers. 5th Naples Forum on Service: 6-9 June 2017., Naples, Italy.
Azer, J., & Alexander, M. (2018). Conceptualizing negatively valenced influencing behavior: forms
and triggers. Journal of Service Management, 29(3), pp. 468-490. doi:10.1108/JOSM-12-
2016-0326
Babić Rosario, A., Sotgiu, F., De Valck, K., & Bijmolt, T. (2016). The effect of electronic word of
mouth on sales: A meta-analytic review of platform, product, and metric factors. Journal of
Marketing Research, 53(3), pp. 297-318.
Bagozzi, R., Gopinath, M., & Nyer, P. (1999). The role of emotions in marketing. Journal of the
Academy of Marketing Science, 27(2), pp. 184-206.
Bagozzi, R., & Yi, Y. (1988). On the evaluation of structural equation models. Journal of the
Academy of Marketing Science, 16(1), pp. 74-94.
Balaji, M. S., Khong, K. W., & Chong, A. Y. L. (2016). Determinants of negative word-of-mouth
communication using social networking sites. Information & Management, 53(4), pp. 528-
540. doi:10.1016/j.im.2015.12.002
Bansal, H., & Taylor, S. (1999). The Service Provider Switching Model (SPSM): A Model of
Consumer Switching Behavior in the Services Industry. Journal of Service Research, 2(2),
pp. 200-218.
Berezina, K., Bilgihan, A., Cobanoglu, C., & Okumus, F. (2015). Understanding Satisfied and
Dissatisfied Hotel Customers: Text Mining of Online Hotel Reviews. Journal of Hospitality
Marketing & Management, 25(1), pp. 1-24. doi:10.1080/19368623.2015.983631
Berger, J., Sorensen, A., & Rasmussen, S. (2010). Positive effects of negative publicity: When
negative reviews increase sales. Marketing Science, 29(5), pp. 815-827.
Blasco-Arcas, L., Hernandez-Ortega, B., & Jimenez-Martinez, J. (2014). The online purchase as a
context for co-creating experiences. Drivers of and consequences for customer behavior.
Internet Research, 24(3), pp. 393-412. doi:10.1108/IntR-02-2013-0023
Blazevic, V., Aksoy, L., van Riel, A., Kandampully, J., Hammedi, W., Garnefeld, I., . . . Donthu, N.
(2013). Beyond traditional word-of-mouth: an expanded model of customer-driven influence.
Journal of Service Management, 24(3), pp. 294-313.
Bolton, R., & Drew, J. (1991). A longitudinal analysis of the impact of service changes on customer
attitudes. Journal of Marketing, 55(1), pp. 1-9.
Bowden, J., Conduit, J., Hollebeek, L., Luoma-aho, V., & Solem, B. A. (2017). Engagement valence
duality and spillover effects in online brand communities. Journal of Service Theory and
Practice, 27(4), pp. 877-897. doi:10.1108/jstp-04-2016-0072
Bowes, A., & Katz, A. (2011). When Sarcasm Stings. Discourse Processes, 48(4), pp. 215-236.
doi:10.1080/0163853x.2010.532757
 27

Brodie, R. J., Fehrer, J. A., Jaakkola, E., & Conduit, J. (2019). Actor Engagement in Networks:
Defining the Conceptual Domain. Journal of Service Research, 22(2), pp. 173-188.
doi:10.1177/1094670519827385
Cheung, C., & Thadani, D. (2012). The impact of electronic word-of-mouth communication: A
literature analysis and integrative model. Decision Support Systems, 54(1), pp. 461-470.
doi:10.1016/j.dss.2012.06.008
Cheung, M., Chuan Luo, Choon Ling Sia, & Chen, H. (2009). Credibility of Electronic Word-of-
Mouth: Informational and Normative Determinants of Online Consumer Recommendations.
International Journal of Electronic Commerce, 13(4), pp. 9-38.
Chevalier, J., & Mayzlin, D. (2006). The effect of word of mouth on sales: Online book reviews.
Journal of Marketing Research, 43(3), pp. 345-354.
Collinson, D. (2002). Managing humour. Journal of Management Studies, 39(3), pp. 269-288.
Colston, H. (1997). Salting a wound or sugaring a pill: The pragmatic functions of ironic criticism.
Discourse Processes, 23(1), pp. 25-46.
Cronin, J., & Taylor, S. (1992). Measuring Service Quality: A Reexamination and Extension. Journal
of Marketing, 56(3), p 55. doi:10.2307/1252296
Diefendorff, J., & Richard, E. (2003). Antecedents and consequences of emotional display rule
perceptions. Journal of Applied Psychology, 88(2), pp. 284-294.
Doh, S., & Hwang, J. (2009). How consumers evaluate e-WOM (electronic word-of-mouth)
messages. Cyber Psycholology & Behaviour, 12(2), pp. 193-197. doi:10.1089/cpb.2008.0109
Dolan, R., Seo, Y., & Kemper, J. (2019). Complaining practices on social media in tourism: A value
co-creation and co-destruction perspective. Tourism Management, 73, pp. 35-45.
doi:10.1016/j.tourman.2019.01.017
Donthu, N., & Gilliland, D. (1996). The Infomercial Shopper. Journal of Advertising Research, 36(2),
pp. 69-76.
East, R., Hammond, K., & Lomax, W. (2008). Measuring the impact of positive and negative word of
mouth on brand purchase probability. International Journal of Research in Marketing, 25(3),
pp. 215-224. doi:10.1016/j.ijresmar.2008.04.001
Filik, R., Turcan, A., Thompson, D., Harvey, N., Davies, H., & Turner, A. (2016). Sarcasm and
emoticons: Comprehension and emotional impact. Q J Exp Psychol (Hove), 69(11), pp. 2130-
2146. doi:10.1080/17470218.2015.1106566
Fishbein, M., & Ajzen, I. (1975). Belief, Attitude, Intention, and Behaviour: An Introduction to
Theory and Research Reading, MA: Addison-Wesley.
Fiske, S., Xu, J., & Cuddy, A. (1999). (Dis)respecting versus (Dis)liking: Status and Interdependence
Predict Ambivalent Stereotypes of Competence and Warmth. Journal of Social Issues, 55(3),
pp. 473-489.
Fornell, C., & Larcker, D. (1981). Evaluating Structural Equation Models with Unobservable
Variables and Measurement Error. Journal of Marketing Research, 18(1), pp. 39-50.
Gelbrich, K. (2010). Anger, Frustration, and Helplessness after Service Failure: Coping Strategies and
Effective Informational Support. Journal of the Academy of Marketing Science, 38(5), pp.
567-585.
Gelbrich, K., Gäthke, J., & Grégoire, Y. (2015). How Much Compensation Should a Firm Offer for a
Flawed Service? An Examination of the Nonlinear Effects of Compensation on Satisfaction.
Journal of Service Research, 18(1), pp. 107-123. doi:10.1177/1094670514543149
Giora, R. (2002). Literal vs. figurative language: Different or equal? Journal of Pragmatics, 34(4), pp.
487-506.
Giora, R., Zaidel, E., Soroker, N., Batori, G., & Kasher, A. (2000). Differential effects of right-and
left-hemisphere damage on understanding sarcasm and metaphor. Metaphor and Symbol,
15(1-2), pp. 63-83.
Godfrey, S., Allender, L., Laughery, K., & Smith, V. (1983). Warning messages: Will the consumer
bother to look? In Proceedings of the Human Factors Society Annual Meeting, CA: Los
Angeles.
González-Ibánez, R., Muresan, S., & Wacholder, N. (2011). Identifying sarcasm in Twitter: a closer
look. Proceedings of the 49th Annual Meeting of the Association for Computational
Linguistics: Human Language Technologies: Short Papers-Volume 2.
 28

Hair, J., Black, W., Babin, B., & Anderson, R. (2010). Multivariate Data Analysis: A Global
Perspective Upper Saddle, NJ: Pearson.
Hammond, D., Fong, G., McDonald, P., Brown, K., & Cameron, R. (2004). Graphic Canadian
cigarette warning labels and adverse outcomes: evidence from Canadian smokers. American
Journal of Public Health, 94(8), pp. 1442-1445.
Harman, H. (1976). Modern factor analysis Chicago: University of Chicago press.
Harmeling, C., Moffett, J., Arnold, M., & Carlson, B. (2017). Toward a theory of customer
engagement marketing. Journal of the Academy of Marketing Science, 45(3), pp. 312-335.
doi:10.1007/s11747-016-0509-2
Harrigan, P., Evers, U., Miles, M. P., & Daly, T. (2018). Customer engagement and the relationship
between involvement, engagement, self-brand connection and brand usage intent. Journal of
Business Research, 88, pp. 388-396. doi:10.1016/j.jbusres.2017.11.046
Heitmann, M., Lehmann, D., & Herrmann, A. (2007). Choice goal attainment and decision and
consumption satisfaction. Journal of Marketing Research, 44(4), pp. 234-250.
Hennig-Thurau, T., Malthouse, E. C., Friege, C., Gensler, S., Lobschat, L., Rangaswamy, A., &
Skiera, B. (2010). The Impact of New Media on Customer Relationships. Journal of Service
Research, 13(3), pp. 311-330. doi:10.1177/1094670510375460
Hiura, K., Kikumori, M., Kishimoto, K., Matsumoto, N., Nakagawa, M., & Ujita, M. (2010). Does
Negative e-WOM Affect Consumer Attitude Negatively or Positively? Journal of eBusiness,
10(1), pp. 38-42.
Ho, T., Lim, N., & Camerer, C. (2006). Modeling the psychology of consumer and firm behavior with
behavioral economics. Journal of Marketing Research, 43(3), pp. 307-331.
Hollebeek, L., & Chen, T. (2014). Exploring positively- versus negatively-valenced brand
engagement: a conceptual model. Journal of Product & Brand Management, 23(1), pp. 62-74.
doi:10.1108/jpbm-06-2013-0332
Hu, L., & Bentler, P. (1995). Evaluation model fit. In R. Hoyle (Ed.), Structural equation modeling:
Concepts, issues, and applications (pp. 76-99). Thousand Oaks: Sage.
Jaakkola, E., & Alexander, M. (2014). The role of customer engagement behavior in value co-
creation. Journal of Service Research, 17(3), pp. 247-261. doi:10.1177/1094670514529187
Jones, M., Mothersbaugh, D., & Beatty, S. (2000). Switching Barriers and Repurchase Intentions in
Services,. Journal of Retailing, 76(2), pp. 259-274.
Juric, B., Smith, S., & Wilks, G. (2016). Negative customer brand engagement: an overview of
conceptual and blog-based findings. In R. brodie, L. Hollebeek & J. Conduit (Eds.), Customer
engagement: Contemporary issues and challenges (pp. 272-286). New Zealand: Routledge.
Kim, J., & Gupta, P. (2012). Emotional expressions in online user reviews: how they influence
consumers’ product evaluations. Journal of Business Research, 65(7), pp. 985-992.
King, R. A., Racherla, P., & Bush, V. D. (2014). What We Know and Don't Know About Online
Word-of-Mouth: A Review and Synthesis of the Literature. Journal of Interactive Marketing,
28(3), pp. 167-183. doi:10.1016/j.intmar.2014.02.001
Kozinets, R. (2010). Netnography: Doing ethnographic research online London: Sage publications.
Kumar, V., Aksoy, L., Donkers, B., Venkatesan, R., Wiesel, T., & Tillmanns, S. (2010). Undervalued
or overvalued customers: Capturing total customer engagement value. Journal of Service
Research, 13(3), pp. 297-310. doi:10.1177/1094670510375602
Lee, J., Park, D., & Han, I. (2008). The effect of negative online consumer reviews on product
attitude: An information processing view. Electronic Commerce Research and Applications,
7(3), pp. 341-352. doi:10.1016/j.elerap.2007.05.004
Lee, M., Jeong, M., & Lee, J. (2017). Roles of negative emotions in customers’ perceived helpfulness
of hotel reviews on a user-generated review website. International Journal of Contemporary
Hospitality Management, 29(2), pp. 762-783. doi:10.1108/ijchm-10-2015-0626
Lee, M., Rodgers, S., & Kim, M. (2009). Effects of Valence and Extremity of eWOM on Attitude
toward the Brand and Website. Journal of Current Issues & Research in Advertising, 31(2),
pp. 1-11. doi:10.1080/10641734.2009.10505262
Lee, M., Shi, N., Cheung, C., Lim, K., & Sia, C. (2011). Consumer's decision to shop online: The
moderating role of positive informational social influence. Information & Management,
48(6), pp. 185-191. doi:10.1016/j.im.2010.08.005
 29

Lemon, K., & Verhoef, P. (2016). Understanding Customer Experience Throughout the Customer
Journey. Journal of Marketing, 80(6), pp. 69-96.
Lemon, K., White, T., & Winer, R. (2002). Dynamic Customer Relationship Management:
Incorporating Future Considerations into the Service Retention Decision. Journal of
Marketing, 66(1), pp. 1-14.
Maio, G., Haddock, G., & Verplanken, B. (2018). The psychology of attitudes and attitude change
London: Sage Publications Limited.
Mathwick, C., & Mosteller, J. (2017). Online Reviewer Engagement. Journal of Service Research,
20(2), pp. 204-218. doi:10.1177/1094670516682088
Meloy, R., Hoffmann, J., Guldimann, A., & James, D. (2012). The role of warning behaviors in threat
assessment: an exploration and suggested typology. Behavioural Science Law, 30(3), pp. 256-
279. doi:10.1002/bsl.999
Oliver, R. (1980). A cognitive model of the antecedents and consequences of satisfaction decisions.
Journal of Marketing Research, 17(1), pp. 460-469.
Papathanassis, A., & Knolle, F. (2011). Exploring the adoption and processing of online holiday
reviews: A grounded theory approach. Tourism Management, 32(2), pp. 215-224.
Park, C., & Lee, T. (2009). Information direction, website reputation and eWOM effect: a moderating
role of product type. Journal of Business Research, 62(1), pp. 61-67.
Park, J., Robertson, R., & Wu, C.-L. (2004). The effect of airline service quality on passengers’
behavioural intentions: a Korean case study. Journal of Air Transport Management, 10(6),
pp. 435-439. doi:10.1016/j.jairtraman.2004.06.001
Patterson, P., Yu, T., & De Ruyter, K. (2006). Understanding customer engagement in services:
Advancing theory, maintaining relevance. Proceedings of ANZMAC Brisbane.
Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in
behavioral research: a critical review of the literature and recommended remedies. J Appl
Psychol, 88(5), pp. 879-903. doi:10.1037/0021-9010.88.5.879
Putrevu, S., & Lord, K. (1994). Comparative and Non-comparative Advertising : Attitudinal Effects
under Cognitive and Affective Involvement Conditions. Journal of Advertising, 23(2), pp. 77-
90.
Qiu, L., Pang, J., & Lim, K. (2012). Effects of conflicting aggregated rating on eWOM review
credibility and diagnosticity: The moderating role of review valence. Decision Support
Systems, 54(1), pp. 631-643. doi:10.1016/j.dss.2012.08.020
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), pp. 548-559.
Reichelt, J., Sievert, J., & Jacob, F. (2014). How Credibility Affects eWOM Reading: The Influences
of Expertise, Trustworthiness, and Similarity on Utilitarian and Social Functions. Journal of
Marketing Communications, 20(1), pp. 65-81.
Rozin, P., & Royzman, E. (2001). Negativity bias, negativity dominance, and contagion. . Personal
Social Psychology Review, 5(4), pp. 269-320.
Schaefers, T., & Schamari, J. (2016). Service Recovery via Social Media. Journal of Service
Research, 19(2), pp. 192-208. doi:10.1177/1094670515606064
Schultz, W., Khazian, A., & Zaleski, A. (2008). Using normative social influence to promote
conservation among hotel guests. Social Influence, 3(1), pp. 4-23.
Sen, S., & Lerman, D. (2007). Why are you telling me this? An examination into negative consumer
reviews on the Web. Journal of Interactive Marketing, 21(4), pp. 76-94.
doi:10.1002/dir.20090
Sparks, B., & Browning, V. (2011). The impact of online reviews on hotel booking intentions and
perception of trust. Tourism Management, 32(6), pp. 1310-1323.
doi:10.1016/j.tourman.2010.12.011
Sparks, B., So, K., & Bradley, G. (2016). Responding to negative online reviews: The effects of hotel
responses on customer inferences of trust and concern. Tourism Management, 53, pp. 74-85.
doi:10.1016/j.tourman.2015.09.011
 30

Stacy, A., MacKinnon, D., & Pentz, M. (1993). Generality and specificity in health behavior:
Application to warning-label and social influence expectancies. Journal of Applied
Psychology, 78(4), pp. 611-627.
Steffes, E., & Burgee, E. (2009). Social ties and online word of mouth. Internet Research, 19(1), pp.
42-59.
Stieglitz, S., & Dang-Xuan, L. (2013). Emotions and Information Diffusion in Social Media—
Sentiment of Microblogs and Sharing Behavior. Journal of Management Information
Systems, 29(4), pp. 217-248. doi:10.2753/mis0742-1222290408
Strack, F., & Deutsch, R. (2004). Reflective and impulsive determinants of social behavior.
Personality and social psychology review, 8(3), pp. 220-247.
Sun, M. (2012). How Does the Variance of Product Ratings Matter? Management science, 58(4), pp.
696-707.
Toplak, M., & Katz, N. (2000). On the uses of sarcastic irony. Journal of Pragmatics, 32(1), pp.
1467-1488.
TripAdvisor.co.uk. (2016). TripAdvisor Fact Sheet. Available at:
http:www.tripadvisor.co.uk/presscentre-c4-Fact_Sheet.html. Retrieved Date.
Verhoef, P., Reinartz, W., & Krafft, M. (2010). Customer engagement as a new perspective in
customer management. Journal of Service Research, 13(3), pp. 247-252.
doi:10.1177/1094670510375461
Vermeulen, I. E., & Seegers, D. (2009). Tried and tested: The impact of online hotel reviews on
consumer consideration. Tourism Management, 30(1), pp. 123-127.
doi:10.1016/j.tourman.2008.04.008
Wan, L. (2013). Culture's impact on consumer complaining responses to embarrassing service failure.
Journal of Business Research, 66(3), pp. 298-305. doi:10.1016/j.jbusres.2011.08.009
Wei, W., Miao, L., & Huang, Z. (2013). Customer engagement behaviors and hotel responses.
International Journal of Hospitality Management, 33(1), pp. 316-330.
doi:10.1016/j.ijhm.2012.10.002
Wilson, A. E., Giebelhausen, M., & Brady, M. (2017). Negative word of mouth can be a positive for
consumers connected to the brand. Journal of the Academy of Marketing Science, 45(4), pp.
534-547. doi:10.1007/s11747-017-0515-z
Wirtz, J., Aksoy, L., den Ambtman, A., Bloemer, J., Horváth, C., Ramaseshan, B., . . . Kandampully,
J. (2013). Managing brands and customer engagement in online brand communities. Journal
of Service Management, 24(3), pp. 223-244. doi:10.1108/09564231311326978
Wu, L., Mattila, A., Wang, C.-Y., & Hanks, L. (2016). The impact of power on service customers’
willingness to post online reviews. Journal of Service Research, 19(2), pp. 224-238.
doi:10.1177/1094670516630623
Wünderlich, N., Wangenheim, F., & Bitner, M. (2013). High Tech and High Touch. Journal of
Service Research, 16(1), pp. 3-20. doi:10.1177/1094670512448413
Xie, K. L., Zhang, Z., Zhang, Z., Singh, A., & Lee, S. K. (2016). Effects of managerial response on
consumer eWOM and hotel performance. International Journal of Contemporary Hospitality
Management, 28(9), pp. 2013-2034. doi:10.1108/ijchm-06-2015-0290
Yim, C., Chan, K., & Hung, K. (2007). Multiple reference effects in service evaluations: Roles of
alternative attractiveness and self-image congruity. Journal of Retailing, 83(1), pp. 147-157.
doi:10.1016/j.jretai.2006.10.011
Zhang, J. Q., Craciun, G., & Shin, D. (2010). When does electronic word-of-mouth matter? A study
of consumer product reviews. Journal of Business Research, 63(12), pp. 1336-1341.
doi:10.1016/j.jbusres.2009.12.011
Zimbardo, P., & Leippe, M. (1991). The psychology of attitude change and social influence England:
Mcgraw-Hill Book Company.