Innocent Until Proven Guilty: Suspicion of Deception in Online

Publications

4-15-2022

Innocent Until Proven Guilty: Suspicion of Deception in Online Innocent Until Proven Guilty: Suspicion of Deception in Online

Reviews Reviews

Maria Petrescu

Embry-Riddle Aeronautical University

, [email protected]

Philip Kitchen

ICN Business School ARTEM

Costinel Dobre

West University of Timisoara

Selima Ben Mrad

Nova Southeastern University

Anca Milovan-Ciuta

West University of Timisoara

See next page for additional authors

Follow this and additional works at: https://commons.erau.edu/publication

Part of the Marketing Commons

Scholarly Commons Citation Scholarly Commons Citation

Petrescu, M., Kitchen, P., Dobre, C., Mrad, S. B., Milovan-Ciuta, A., Goldring, D., & Fiedler, A. (2022). Innocent

Until Proven Guilty: Suspicion of Deception in Online Reviews.

European Journal of Marketing, 56

(4).

https://doi.org/10.1108/EJM-10-2019-0776

This Article is brought to you for free and open access by Scholarly Commons. It has been accepted for inclusion in

Publications by an authorized administrator of Scholarly Commons. For more information, please contact

[email protected].

Authors Authors

Maria Petrescu, Philip Kitchen, Costinel Dobre, Selima Ben Mrad, Anca Milovan-Ciuta, Deborah Goldring,

and Anne Fiedler

This article is available at Scholarly Commons: https://commons.erau.edu/publication/1837

Innocent until proven guilty: suspicion of deception in online reviews

Structured Abstract

- Purpose: This study formulates a new framework for identifying deception in consumer

reviews through the lens of Interpersonal Deception Theory and the Persuasion

Knowledge Model. It evaluates variables contributing to consumer intentions to purchase

after reading deceptive reviews and proposes deception identification cues to be

incorporated into the interpersonal communication theoretical framework.

- Methodology: The first study is qualitative and quantitative, based on sentiment and

lexical analysis of 1000 consumer reviews. The second study employs a USA national

consumer survey with a PLS-SEM and a Process-based mediation-moderation analysis.

- Findings: The study shows deceptive characteristics that cannot be dissimulated by

reviewing consumers that represent review legitimacy based on review valence,

authenticity, formalism, and analytical writing. The results also support the central role of

consumer suspicion of an ulterior motive, with a direct and mediation effect regarding

consumer emotions and intentions, including brand trust and purchase intentions.

- Research implications: This paper presents a new framework for identifying deception

in consumer reviews based on IDT and PKM, adding new theoretical elements that help

adapt these theories to written digital communication specificities. The study clarifies the

role of suspicion in a deceptive communication context and shows the variables

contributing to consumers’ purchase intention after reading deceptive reviews. The

results also emphasize the benefits of lexical analysis in identifying deceptive

characteristics of reviews.

- Practical implications: Companies can consider the vulnerability of certain generations

based on lower levels of suspicions and different linguistic cues to detect deception in

reviews. Long-term, marketers can also implement deception identification practices as

potential new business models and opportunities.

- Social implications: Policymakers and regulators need to consider critical deception cues

and the differences in suspicion levels among segments of consumers in the formulation

of preventative and deception management measures.

- Originality/value: This study contributes to the literature by formulating a new

framework for identifying deception in consumer reviews, adapted to the characteristics

of written digital communication. The study emphasizes deception cues in eWOM and

provides additional opportunities for theorizing deception in electronic communication.

Keywords: consumer deception; online reviews; incentivized reviews; Persuasion Knowledge

Model; Interpersonal Deception Theory; lexical analysis.

Article classification: research paper

1. Introduction

Modern consumers tend to evaluate products based on their peers’ opinions and reviews

in an overwhelming amount (Bambauer-Sachse and Mangold, 2013; Chakraborty and Bhat,

2018; Decker and Trusov, 2010; De Langhe et al., 2016). Studies found that 97% of consumers

read online reviews for local businesses, and 93% of individuals are influenced by online

reviews in their consumption decisions (Schoenmueller, Netzer, and Stahl, 2020). Since

consumers rely on online reviews when deciding which products and services to purchase, some

marketers have injudiciously started employing fake reviews to influence potential customer

decisions (Hu et al., 2012; Malbon, 2013; Steward et al., 2020).

Consequently, on modern etailing platforms, we encounter incentivized, sponsored, and

even fake reviews, which leads to misleading situations for online shoppers. This situation is

even more problematic, considering that the influence of consumer review volume and valence is

still debated in the literature (Kordrostami, Liu-Thompkins, and Rahmani, 2021). Even the FTC

updated its guidelines for endorsements in 2009, in response to pressures from consumer groups,

requiring the identification of any material connection between the seller and the reviewer, and

took legal measures in multiple fake review cases (FTC, 2009; Plotkina, Munzel, and Pallud,

2020; Steward et al., 2020). Studies have also discussed regulatory intervention to prohibit the

use of deceptive consumer communications (Malbon, 2013; Mayzlin et al., 2014; Plotkina,

Munzel, and Pallud, 2020; Steward et al., 2020).

The debate on regulatory and industry intervention is essential, as more than two-thirds of

consumers trust online reviews when evaluating products and making purchase decisions

(Dellarocas, 2006; Singh et al., 2017). As deception in persuasive marketing communication

through online consumer reviews can represent a threat for consumers, businesses, policymakers,

and society, a better understanding of how consumers perceive, detect, and interpret deception is

needed (Plotkina, Munzel, and Pallud, 2020; Steward et al., 2020).

Our study aims to expand on the previous literature and analyze the use and identification

of deception in online consumer reviews through the lens of Interpersonal Deception Theory

(IDT) (Buller et al., 1996; Burgoon et al., 1996), explaining the deceptive communication

process, and the Persuasion Knowledge Model (PKM) (Campbell and Kirmani, 2000; Friestad

and Wright, 1994, 1995; Kirmani and Campbell, 2004), complementing our theoretical

framework with the antecedents of suspicion and information acquisition. Studies on these two

theoretical frameworks have called for more research on persuasive communication and

deception in modern contexts, especially on digital platforms and computer-mediated

communication (Burgoon et al., 2010; Evans and Park, 2015; George and Robb, 2018; Isaac and

Grayson, 2017).

We propose an improved framework of deception identification in consumer reviews as

an update of IDT and PKM in the written language spectrum. Applying these theories in the

context of online consumer reviews will help further research on deceptive communication,

heuristics used to interpret deception, and related to the effects of suspicion in digital word-of-

mouth communication. We assess linguistic cues that consumers can use to identify fake and

incentivized reviews in the under-explored written digital communication context and evaluate

how consumer suspicion of deceptive communication influences purchase intentions. The study

also explores incentivized reviews and differences in a deceptive communication relationship

among age generations. From a managerial standpoint, this research will help business managers

understand the short and long-term impact of deceptive reviews on consumers and the

implication brought by consumer suspicion of an ulterior motive. Further, it will inspire

practitioners to find solutions to actual or perceived online deception via reviews to increase

consumer brand trust.

After a literature review focused on applying IDT and PKM in the context of online

consumer reviews, we performed two multi-method empirical studies. The first identifies

common elements of digital written deception cues via qualitative and quantitative content

analysis and lexical analysis. The second study extends the analysis and assesses consumer

reactions to a possible deceiving environment by evaluating the role played by suspicion of

reviewers’ ulterior motive.

2. Conceptual framework: deception in consumer reviews

Deception in communication is “a communicator’s deliberate attempt to foster in others a

belief or understanding which the communicator considers to be untrue” (DePaulo and DePaulo,

1989, p. 1553). It is an intentional or deliberate act, accomplished by manipulating information

in some way, with an instrumental end goal, to generate or preserve a belief or conclusion that

the communicator knows to be false (Buller and Burgoon, 1996; DePaulo et al., 2003; Munzel,

2015; Peng et al., 2016; Xiao and Benbasat, 2011).

Deception in marketing includes actions or messages that impact consumer decisions,

make them believe something that is not verifiably true about consumption elements, or create

distrust in the consumption process (Aditya, 2001, p.743). Consumers have a specific set of

expectations regarding the number of details that should be provided in the communication, its

truthfulness, its relevance to the conversation, and message clarity; through deception,

expectations can be exploited or damaged with the widespread use of technology (Buller et al.,

1996; Burgoon et al., 1996; McCornack, 1992; McCornack et al., 1992). Some common

deceptive practices include automatically filtering out negative consumer reviews (concealment),

posing as consumers to write positive reviews about products and services received from the

company (through falsification), and offering incentives to encourage positive reviews

(concealment or equivocation) (Dellarocas, 2006; Hu et al., 2011, 2012; McCornack, 1992;

Munzel, 2016; Xiao and Benbasat, 2011).

2.1 Interpersonal Deception Theory

According to IDT, message receivers are active agents whose cognitions and behavior are

essential in explaining deceptive messages (Burgoon et al., 1996). IDT focuses on the dyadic

relationship between a sender and a receiver in which the sender might be falsifying information

(Buller et al.,1996). IDT is a theory of reaction to perceived deception, a combination of

interpersonal communication and deception principles related to credibility and honest

communication (Buller et al., 1996). Suspicion is defined as a belief, without certainty and

enough evidence or proof, that an individual’s speech or behavior may be duplicitous (Burgoon

et al., 1996).

This model of deceptive interpersonal communication considers the level of suspicion of

the receiver, knowledge, expectations, and type of deception (Burgoon et al., 1996; Buller et al.,

1996). Message receivers (or readers) are active agents whose own cognitions and behaviors are

indispensable in explaining the consequences of deceptive messages (Burgoon et al., 1996).

Impersonation is also a deceptive practice, as businesses and their representatives deliberately

pretend to be other persons to post deceptive reviews deliberately written to sound authentic to

deceive consumers (Munzel, 2016; Ott et al., 2012). While research in psychology has shown

that individuals use verbal and nonverbal cues to detect deception in face-to-face

communications, in digital settings, consumers do not have access to the same range of cues,

making identification of false reviews much more difficult (Anderson and Simester, 2014; Ott et

al., 2012). Also, consumers might not be generally aware of the degree of deception in online

reviews (Peng et al., 2016), in which case PKM can help assess marketing knowledge and

suspicion levels.

2.2 The Persuasion Knowledge Model

To complement IDT, the Persuasion Knowledge Model talks about the way consumers

become knowledgeable about persuasion attempts through social interactions, conversations,

observation, and discussions about marketers, advertisers, and salespeople (Friestad and Wright,

1994; Lawlor, Dunne, and Rowley, 2016; Lunardo and Roux, 2015). This process leads to

personal knowledge about influence attempts used in marketing, shaping how consumers

respond as persuasion targets (Friestad and Wright, 1994, 1995).

Consumers can eventually use their knowledge to identify marketers trying to influence

them and can manage the interaction and the relationship for their own goals (Campbell and

Kirmani, 2000; Friestad and Wright, 1994, 1995; Kirmani and Campbell, 2004; Lawlor, Dunne,

and Rowley, 2016). This information depends on consumer accessibility to ulterior motives,

cognitive capacity, and consumer experience, aspects that come to complement the variables

included in IDT (Campbell and Kirmani, 2000; Lawlor, Dunne, and Rowley, 2016). Like IDT

(Burgoon et al., 1996), PKM incorporates the concept of suspicion generated by the acquisition

of persuasion and market knowledge (Nelson et al., 2017) while providing additional

information on the factors that affect the development of suspicion and its antecedents.

As presented in Table 1, there are thousands of citations in the research databases for

these theories and many discussions on deception, suspicion, and persuasion knowledge in

different areas of research (Burgoon et al., 2010; Evans and Park, 2015; George and Robb, 2018;

Isaac and Grayson, 2017). However, as some of the studies in Table 1 show, the research and

practice trends underline an evolution in the literature and knowledge gaps regarding deception

and persuasion in digital and computer moderated communication (Burgoon et al., 2010;

Burgoon and Nunamaker Jr. a,b 2004; Evans and Park, 2015; Fuller et al., 2013; George and

Robb, 2018; Isaac and Grayson, 2017; Kim, Kim, and Marshall, 2016). Based on previous

findings and the critical research gaps identified, we focus on formulating an improved model of

deception and persuasion in digital reviews communication, centered on written language cues

and consumer behavior.

(Please place Table 1 about here)

Considering both IDT and PKM, we base our analysis on the conceptual framework

included in Figure 1. While IDT formulates the critical explanatory variables related to the

deceptive communication process, PKM adds to this framework by providing information related

to the consumer suspicion process and consumption information acquisition.

(please insert Figure 1 here)

2.3 Review characteristics

As companies started offering online reputation management services, some digital

marketers have increased the use of manipulated online reviews to promote products and

services in the online environment (Malbon, 2013; Munzel, 2015). Anderson and Simester

(2014) underlined that lower ratings in a review were associated with reduced demand for that

product over the next 12 months and that reviews from 15 customers influenced the behavior of

other 985 customers.

Each reviewer discusses a different set of product features, based on personal

consumption experiences, despite expressing opinions about the same product feature (Moon and

Kamakura, 2016). Ott et al. (2012) explain that review communities’ role is to reduce the

inherent information asymmetry between buyers and sellers in online marketplaces by providing

buyers with a priori knowledge of the underlying quality of the products sold.

Review quality includes the extent to which consumers perceive it as logical and reliable,

including the perceived justification for reviewers’ recommendations. Consumers are more likely

to consider the message legitimate if the reviewer provides detailed and valid arguments

(Chakraborty and Bhat, 2018; Hong et al., 2017). Considering previous findings related to online

consumer reviews, as well as the interactions of consumer suspicion with the level of marketing

knowledge, we expect that reviews legitimacy will reduce the impact of consumer suspicion on

the consumption process (Burgoon et al., 1996; Buller et al., 1996; Campbell and Kirmani, 2000;

Friestad and Wright, 1994, 1995; Kirmani and Campbell, 2004; Lawlor, Dunne, and Rowley,

2016).

H1a: The level of review legitimacy reduces the effect of consumer suspicion on brand

trust.

H1b: The level of review legitimacy reduces the effect of consumer suspicion on

purchase intentions.

Consumer perceptions of reviewers’ ability and willingness to tell the truth moderate the

effectiveness of a comment that includes both positive and negative evaluations (De Langhe et

al., 2016; Lin and Xu, 2017). High variance in review valence represents uncertainty and

positively affects attitudes and intentions towards poorly rated products but negatively impacts

highly rated products (Kordrostami, Liu-Thompkins, and Rahmani, 2021; Kostyra et al., 2016).

A study analyzing online consumer reviews using a sentiment mining approach found that their

length and longevity positively influenced their readership and helpfulness (Salehan and Kim,

2016). Some reviewers have never purchased or tried the product and are providing fake reviews

for self-gain, while others are incentivized to submit online posts (Steward et al., 2020). In the

case of incentivized word-of-mouth, the incentivization process can induce biased self-interest

on the side of the recommender (Pongjit and Beise-Zee, 2015).

Large online retailers, such as Amazon, have issues with deceptive reviews, including

incentivized reviews where the vendor or a reputation-management company offer free or

discounted products to reviewers in exchange for recipients’ “honest opinion” on the item in a

review on Amazon (Perez, 2016; Soper, 2015). These reviewers are more likely to post positive

reviews overall, with approximately 4.74 stars out of five, compared with an average rating of

4.36 for non-incentivized reviews (Perez, 2016).

H2a: Incentivized reviews have a more positive valence compared to other categories.

In the context of automated linguistic features analysis and classification models,

researchers have explored different frameworks based on IDT and have called for a unification

of the framework linguistic cues used to identify deception (Fuller et al., 2013). Burgoon and

Qin (2006) have developed a framework based on eight constructs, while Fuller et al. (2013)

have retained seven constructs – quantity, specificity, affect, diversity, uncertainty,

nonimmediacy, and activation. Other studies have focused on computer-mediated

communication and cues specific to this type of modern linguistics (Carlson et al., 2004; Zhou et

al., 2004).

The perceived level of manipulation and the degree of review authenticity are also

reflected in consumer attitudes and the perceived helpfulness of online product reviews (Mayzlin

et al., 2014; Peng et al., 2016; Steward et al., 2020). Moreover, incentivized review experiences

influence even review writers and their writing styles, motivating them to seek other rewards,

such as impulsive buying (Motyka et al., 2018).

Deceivers use less analytical information and less clear and complete messages to

manipulate content, not adding much detail and relevant information to their responses (Burgoon

et al., 1996, Carlson et al., 2004). There are significant differences in authentic communication,

analytical writing style, and text formalism as a function of the type of review (Chakraborty and

Bhat, 2018; Hong et al., 2017).

H2b: The level of authenticity is lower for incentivized reviews.

Research on automated linguistic analysis has highlighted that deceiving individuals lack

the support of real experiences and memory, so they tend to communicate in a language that

lacks complexity, detail and omits specific, analytical language (Zhou et al., 2004). The same

study on deceptive language emphasized that deceptive senders employ more informality in their

messages than their respective receivers, including more typographical errors in written

messages (Zhou et al., 2004).

Buller and Burgoon (1996) pointed out that a deceiving message is more likely vague and

short, including words of withdrawal rather than involvement, and indicates a disassociation of

the sender. Therefore, we hypothesize that, for digital consumer reviews, we should expect lower

levels of formalism and analytical writing for text resulting from incentives.

H2c: The level of analytical writing is lower for incentivized reviews.

H2d: The level of text formalism is lower for incentivized reviews.

2.4 Attitude toward reviews

Even though consumers are aware of the possibility of manipulation, they might only

partially correct it as a function of their expected level of manipulation (Hu et al., 2011). A

generalized effect of consumer distrust created by misleading activities can activate a defensive

stereotyping mechanism in the online environment, adversely affecting marketing

communication strategies (Friestad and Wright, 1995; Riquelme and Roman, 2014).

As consumers become more aware that marketers can manipulate reviews, they are less

likely to believe and trust them, especially when reviews are dissonant to consumers’ initial

evaluations of a particular product or service (Dellarocas, 2006; Hu et al., 2011). Therefore, we

expect that consumers who have a more positive attitude towards reviews will be more likely to

have a positive attitude towards the reviewed brand.

H3a: Consumers’ attitude toward reviews is positively related to consumers’ brand trust

in a reviewed brand.

According to IDT, deceivers use various control attempts when they negotiate the

outcomes with their partner, and the message receiver responds with strategic moves based on

the information received, various cues, and already formed attitudes. Considering these potential

issues, as well as the emphasis that IDT places on consumer expectations and experience in a

deception context (Burgoon et al., 1996; Buller et al., 1996), this study includes consumer

attitudes toward reviews in the model to represent already formed views on this type of digital

content (Khare, Labrecque, and Asare, 2011).

Moreover, PKM highlights the effect of previous experience, persuasive knowledge, and

previously formed attitudes toward advertising and persuasive intent on consumer-level

suspicion and skepticism (Evans and Park, 2015; Friestad and Wright, 1994). A prerequisite for

persuasion knowledge development is individuals’ direct and indirect experiences with the

marketing messages that helped them form attitudes toward this way of communication (Evans

and Park, 2015).

Based on the two theories in our framework, we hypothesize that consumers’ attitude

toward reviews is likely to have a negative effect on consumers’ level of suspicion of an ulterior

motive from the reviewer. This is influenced by the information and experiences consumers

previously acquired in the marketplace.

H3b: Consumers’ attitude toward reviews is negatively related to their suspicion of an

ulterior motive from the reviewer.

2.5 Suspicion of an ulterior motive

According to IDT and PKM, the correspondent of deception on the side of the message

receiver is perceived suspicion, the belief, without enough specific evidence, that an individual’s

message may be deceiving, a knowledge that consumers learn through experience (Buller and

Burgoon, 1996; Friestad and Wright, 1994; Kirmani and Campbell, 2004). When consumers start

to doubt a reviewer’s honesty based on acquired consumption knowledge, external influences, or

intrinsic behavior, their suspicion becomes an essential catalyst in the transaction, as it may alter

both their behaviors and those of the message communicators (Bambauer-Sachse and Mangold,

2013; Burgoon et al., 1996; Friestad and Wright, 1994; Kirmani and Campbell, 2004).

Consumer suspicion of ulterior motives refers to questioning the reasons that inspire

another person’s behavior or doubting the authenticity of that conduct (DeCarlo, 2005; DeCarlo

et al., 2013; Friestad and Wright, 1994; Kirmani and Campbell, 2004). When consumers become

suspicious, their message acceptance levels are lower, and they pay more attention to the

communicator’s agenda (DeCarlo et al., 2013). When consumers believe that a review was

written with ulterior motives, they perceive a higher level of untrustworthiness in the review

(Bambauer-Sachse and Mangold, 2013; Lin and Xu, 2017; Reimer and Benkenstein, 2016).

Suspicion makes individuals look for additional information, and it could negatively

affect the attitude formation process and purchase intentions (DeCarlo, 2005; DeCarlo et al.,

2013; Friestad and Wright, 1994; Kirmani and Campbell, 2004). Studies on word-of-mouth

communication have found that when consumers are suspicious of ulterior motives, the

effectiveness of the message will decrease (Godes and Mayzlin, 2009; Mayzlin, 2006). The use

of rewards for recommendations hurts the receiver’s attitude toward the brand because the

impression that a business has motivated friends to profit from a personal relationship (Pongjit

and Beise-Zee, 2015).

Providing biased incentivized electronic word-of-mouth (eWOM) changes the

communicator’s brand evaluation. That individual is likely to remember the prejudiced

recommendation and use it as a learning opportunity to acquire knowledge and update his

attitude (Friestad and Wright, 1994, 1995; Kim et al., 2016; Kirmani and Campbell, 2004).

When consumers know that a business is offering rewards for engaging in WOM, they consider

these reviewers as having lower source trustworthiness levels (Martin, 2014). Seeing the place of

perceived suspicion in the IDT and PKM, we hypothesize that this variable will act as a mediator

between consumers’ attitudes toward reviews and their brand trust, as well as their purchase

intentions.

H4a: Consumers’ level of suspicion of an ulterior motive mediates the relationship

between attitude toward reviews and brand trust.

H4b: Consumers’ level of suspicion of an ulterior motive mediates the relationship

between attitude toward reviews and purchase intentions.

2.6 Brand trust

Perceived trustworthiness of the source mediates the impact of the cues that consumers

use, including reviewer identity and persuasion knowledge, on these behavioral intentions (Ma

and Lee, 2014; Munzel, 2016). Honest reviews affect purchase intentions in the same direction

as review valence, while untrustworthy reviews lead to a “boomerang effect” that causes positive

reviews to decrease and negative reviews to increase purchase intention (Reimer and

Benkenstein, 2016).

Consumer reviews decrease brand trust’s influence on purchase decisions and indicate a

brand’s online reputation while reducing the impact of a brand’s general reputation once they are

displayed together (Kostyra et al., 2016). The impact of user-generated content, such as online

reviews, is starting to significantly affect consumers (Decker and Trusov, 2010; De Langhe et

al., 2016). Considering these effects shown by the previous literature, we hypothesize that brand

trust formed when reading an online review will mediate the relationship between suspicion of

an ulterior motive and purchase intentions. We also test the relationship between brand trust and

consumers’ intentions to purchase the reviewed product as a replication.

H5a: Brand trust mediates the relationship between suspicion of an ulterior motive and

purchase intentions.

H5b: Brand trust mediates the relationship between attitude toward reviews and purchase

intentions.

2.7 Generational cohort

Under the PKM lens, the results are mixed regarding the influence of age on persuasion

knowledge and consumer use of cognitive and information processing ability (Campbell and

Kirmani, 2000; Carlson, Bearden, and Hardesty, 2007; Kirmani and Campbell, 2004; Lunardo

and Roux, 2015). Some studies have emphasized that a diminution of the cognitive and

information processing ability in adults can interfere with their activation of persuasion

knowledge (Carlson et al., 2007; Kirmani and Campbell, 2004), while others have shown its

application in the context of advertising for children (Lawlor, Dunne, and Rowley, 2016).

Persuasion knowledge is based on consumers’ direct and indirect experiences with the marketing

content and platform of communication, which becomes even more critical in a digital context

(Evans and Park, 2015).

Regarding online reviews, our attention focuses on differences in the deception model as

a function of the generational cohort. Various age groups have different experience levels with

consumer reviews, online shopping, and exposure to deceptive circumstances. The Millennial

generation is a distinct age group, including confident and better-educated members, with

frequent and extended social contact with peer groups and digital interactions (Doster, 2013;

Hübner Barcelos and Vargas Rossi, 2014). Less than half of Baby Boomers (48%) and retirees

(45%) read online reviews before making a purchase, while 76% of Millennials and 63% of Gen

X members do (Vantiv, 2018). Moreover, the younger generations also emphasize online

reviews in their decision-making process (Vantiv, 2018).

As generational cohorts have different attitudes and perceptions regarding online reviews,

we hypothesize a moderation position for this variable in the deception model. As the younger

generational cohorts have greater market experience on the digital platform in the environment

of online reviews and more persuasive knowledge in this regard, we expect their suspicion levels

to have a more significant negative impact on their brand trust levels and purchase intentions.

H6a: Younger generational cohorts enhance the effect of consumer suspicion of an

ulterior motive on brand trust.

H6b: Younger generational cohorts enhance the effect of consumer suspicion of an

ulterior motive on purchase intentions.

3. Methodology

We employ a multi-method approach, using the insight advantages of qualitative studies

and the benefits that mining social commerce sites like Amazon represent for word-of-mouth

communication (Humphreys and Wang, 2018; Kirmani and Campbell, 2004). We first use an

automated text analysis method, sometimes used by researchers to make discoveries, find

patterns in positive vs. negative consumer reviews, and evaluate differences between expert and

consumer discourse in product-related comments (Lee and Bradlow, 2011; Netzer et al., 2012;

Situmeang, de Boer, and Zhang, 2020). This method also allows us to measure differences in

language among groups and types of reviews (Humphreys and Wang, 2018).

3.1 Study 1: Identifying deception in online reviews

Researchers have tried to find algorithms and markers that help consumers assess the

degree of deception in reviews by leveraging their textual characteristics (Banerjee and Chua,

2017; Ott et al., 2012). Previous analyses have focused on review comprehensibility

(readability), specificity (informativeness and relevance), exaggeration (sentiment), and

negligence (Banerjee and Chua, 2017). Researchers have analyzed deceptive messages’ writing

style and linguistic characteristics, including deceptive travel reviews, deceptive emails, and

online dating profiles (Anderson and Simester, 2014; Hu et al., 2012; Markowitz and Hancock,

2016; Ott et al., 2011). Our analysis in Study 1 has exploratory and empirical purposes of testing

the hypotheses related to the critical linguistic cues that can be used to identify deception.

3.1.1 Analysis

Approximately 1000 reviews for a scented spray product were downloaded from Amazon

from January 2014 to January 2018 to perform an exploratory study. During this period, the

branding company also ran an incentivized consumer review campaign through a third-party

online reputation company by offering the product for free, making it possible for us to identify

incentivized reviews that mentioned: “I received the product for free.” Our analysis included 105

incentivized reviews, 620 verified reviews (Amazon verifies as product purchasers), and 150

unverified reviews.

This study used a content analysis approach based on sentiment and lexical analysis of a

text. We first performed sentiment analysis of the three types of reviews using NVivo (Tang and

Guo, 2015). This type of analysis focuses on sentiment strength detection to classify text for the

overall strength of positive and negative sentiment and its polarity (Thelwall, 2016). The

helpfulness of sentiment classification is shown by analyzing consumer reviews and assigning

them to appropriate sentiment categories (Bai, 2011; Salehan and Kim, 2016). NVivo searches

for expressions of sentiment in the text content, based on a sentiment dictionary and an

algorithmic method, where each word containing sentiment has a predefined score, with a range

on a scale from very negative to very positive; neutral words are not coded (Tang and Guo,

2015).

Then we performed a content and semantic analysis using LIWC, based on existing

psychometrically tested scales and algorithms that include analytical thinking (Pennebaker et al.,

2014) and authenticity (Newman et al., 2003). Studies have called for more unstructured textual

content and semantic analysis of online reviews (Büschken and Allenby, 2016; Tirunillai and

Tellis, 2014). LIWC2015 is based on three internal dictionary systems, with a master dictionary

composed of almost 6,400 words, word stems, selected emoticons, and a corresponding

dictionary entry that defines word categories (Pennebaker et al., 2015). In the context of

deception, LIWC was used successfully in analyzing five independent samples, where it

correctly classified liars and truth-tellers at a rate of 67% when the topic was constant and a rate

of 61% overall (Newman et al., 2003). Some of the indices included in our analysis are

analytical thinking (Pennebaker et al., 2014) and authenticity (Newman et al., 2003), based on

the findings of previous literature (Chakraborty and Bhat, 2018; Hong et al., 2017). These were

derived from previously published findings and converted to percentiles based on standardized

scores from large comparison samples. The analytical thinking index is measured by identifying

formal, logical, and hierarchical thinking patterns in text based on function words and grammar

words (Pennebaker et al., 2014; Plotkina et al., 2020). Authenticity analyzes whether individuals

communicate honestly, based on research showing that consumers are more personal, disclosed,

and vulnerable when authentic (Pennebaker et al., 2015; Plotkina et al., 2020). Informality is

measured with a language dictionary that contains 380 words, including such categories as swear

words, netspeak, nonfluencies, and fillers. For the LIWC 2015 version, the corrected alphas for

the indices used range from 0.55 to 0.84, computed on a sample of about 181,000 text files from

several language corpora (blogs, natural language, media), based on the Spearman Brown

method (Pennebaker et al., 2015).

3.1.2 Results

We first explored in automated content analysis in NVivo, as shown in Figure 2, the most critical

themes in the consumer comments downloaded. The incentivized reviews emphasize a positive

focus on the product, on different scent options, “good,” “nice,” “strong,” and “great,” without

concrete and practical references regarding the purchase or the consumption experience, and

with very few negative mentions. As discussed in our conceptual framework, the main aspects

emphasized in the incentivized reviews category confirm a focus on general, non-specific, non-

descriptive elements, with positive inclinations.

(please insert Figure 2 here)

The themes in the category of unverified reviews (which can also be incentivized or fake) reveal

more positive elements, the benefits of low price, and a good smell. As expected through our

theoretical framework, they show an inclination towards positive word-of-mouth. Consumers in

the verified category focus more on their overall experience with the product, favorite scent,

duration of the scent, and its freshness, in a more concrete and specific way of informing their

peers. All three categories of reviews showed elements from the themes “scent” and “smell,” but

their focus was different: for incentivized comments, the discussions were centered on product

lines and scents, while for the verified reviews, the text exhibited a more emotional presentation,

with sub-themes including “favorite scent,” “calming scent,” and “perfumy.”

All these findings have made it necessary to assess whether deception can be identified

from the tone and sentiment of reviews; therefore, we performed a sentiment analysis in NVivo.

The results show that incentivized reviews have a significantly more optimistic tone than

unverified and verified reviews, with lower negative content and a higher level of very positive

content. Their text shows a significantly more extreme positive sentiment (44%) than unverified

(32%) and verified (27.95%) reviews, including stealth incentivized or fake reviews. `

Some examples of verified reviews include “Weak to nonexistent smell. Had 3 going on

in one room and still virtually no smell. All were set on the highest setting”, “It works! I keep

one by the laundry and another by the shoes. Generally, I keep them dialed a bit low, the scent

can get intense if you turn it up.”. At the same time, the incentivized reviews state the reward

part, as in the following: “Great scent, long lasting. I received to sample from (…) for free for an

honest review. I am very pleased with this product. I tried the clean linen scent in my son’s room

which has a tendency of having a strong musty smell. This has kept room smelling clean and

fresh for at least two weeks. I totally recommend”.

Our findings underline that consumers who write reviews intending to manipulate are

more likely to use sentiments to influence a potential reader’s purchase behavior (Hu et al.,

2012). The lexical analysis performed in LIWC looks at differences in writing style and cues that

can signal deceptive practices to test hypotheses 2 a-d and provide exploratory information about

other variables. The results obtained in LIWC show the differences among verified, unverified,

and incentivized reviews, emphasizing the formal character of incentivized reviews and the more

personal tone of verified reviews. Incentivized reviews received a higher score for the

authenticity index compared to the other two, which is fascinating and surprising, especially

considering the purpose of this index; nevertheless, this can be justified by the number of details

and further explanations provided by consumers who received rewards, as shown by the

word/sentence count.

The index measuring the use of negations and the employment of numbers for each of the

three categories of reviews confirms the NVivo results related to review valence. It underscores a

focus on information in verified reviews, and much more positive sentiment and lack of

numerical information in incentivized reviews. LIWC also provides an index of words related to

affective processes and feelings (such as happiness, satisfaction…) and perceptions related to

seeing, hearing, and feeling. The linguistic analysis results reflect a more objective and

descriptive approach than an experiential style for receiving rewards. To test these findings, we

performed a t-test analysis in Table 2 using the indices obtained in LIWC for each group,

confirming the differences hypothesized in H2a-d.

(please insert Table 2 here)

Our findings in the qualitative analysis emphasize that incentivized consumers are less

likely to exhibit affective processes, concrete perceptions, and feelings related to their experience

with the product reviewed than actual product buyers. An analysis of the amount of informal

language used shows that incentivized reviews are more likely to incorporate this type of

content. The results highlight significant differences among the three types of reviews and show

deceptive characteristics that consumers cannot dissimulate, enriching the potential repertoire of

cues that marketers and consumers can use in deception identification.

3.2 Study 2: Consumers as review evaluators

The second study considers variables included in the conceptual framework based on IDT

and PKM, such as the level of suspicion of the reader, consumer expectations, and the type of

deception in the context of online consumer reviews. To test the model discussed, an online

survey was used to collect data from a national sample of U.S. consumers through Qualtrics.

3.2.1 Analysis

The final sample includes 505 consumers with heterogeneous demographic

characteristics, 47% male, similar distributions in each age group and income level, and the

majority have a college degree. Consumers were randomly distributed into three groups and

shown an actual Amazon consumer review about a fictional room deodorizer brand from one of

the three categories: incentivized, verified, and unverified, as shown in the example in Figure A1

in the Appendix. A fictional brand was used to eliminate attitudes toward the brand potentially

formed before seeing the review; however, the advertisement used was very similar to the image

of an existing product, while genuine reviews from Amazon for a similar brand were used in the

three experimental groups to ensure they are realistic.

We employed product images and text already analyzed in Study 1 to assess their level of

legitimacy and text characteristics. To perform a post-hoc manipulation check and to assess

confounding effects, we performed an ANOVA test. The results of the analysis show the success

of the manipulation, with significant differences in consumer level of suspicion based on the

three experimental groups, while the confounding effects, based on the social desirability bias

scale (Crowne and Marlowe, 1960), show no significance (Hauser and Gonzalez, 2018; Perdue

and Summers, 1986).

The constructs in the model were measured based on established scales of measurement,

as shown in TableA1 in the Appendix. Unless the scale developers noted specific

recommendations, a 7-item Likert rating option was used. Information on the rigor of the model

is further provided in Table 3, which presents data on reliability, validity, and common method

bias. To increase reliability and minimize bias, we included some reverse-coded items and a few

attention filters that removed inattentive respondents from the survey (Podsakoff et al., 2003).

(please insert Table 3 here)

All factor loadings for the measures in the PLS-SEM analysis were above the

recommended value of 0.60, as shown in Table A1 in the Appendix (Hair, Ringle, and Sarstedt,

2013). An analysis of the reliability of the measurement scales shows that Cronbach’s alpha

coefficients and the composite reliability of each construct are higher than the minimal accepted

value of 0.70 (Bagozzi and Yi, 1988; Nunnally, 1978). The Fornell-Larcker criterion in Table 3

shows that the AVE values for each construct are all above the recommended 0.50 level

(Bagozzi and Yi, 1988; Fornell and Larcker, 1981), concluding discriminant validity between all

constructs. The heterotrait-monotrait (HTMT) ratio shows values lower than 0.85 for

conceptually distinct constructs.

3.2.2 Results

To test the hypotheses presented in the conceptual model, we employed a PLS-SEM

procedure using SmartPLS 3. The results of the overall model are presented in Figure 3 and

Table 4. The output of the PLS-SEM analysis produced an SRMR of 0.074 for the model, which

is at the recommended cutoff of equal or less than 0.08 (Hu and Bentler, 1999).

(please insert Figure 3 here)

The rms Theta is 0.207, slightly higher than the accepted limits presented by the current

literature for SmartPLS (Henseler et al., 2014). The Q² value also shows levels above 0, with

good predictive relevance for the selected endogenous construct, especially purchase intentions.

Figure 3 presents the significance for each hypothesized relationship and the R-square for the

endogenous variables included in the model, emphasizing good predictive power. The effect-

power presented by f-square shows a value of 0.13 for the influence of attitude toward reviews

on brand trust, and 0.061 for the influence of suspicion on purchase intentions.

(please insert Table 4 here)

All the hypothesized relationships were significant at the p<0.001 level. Also, the R-

square level showed an excellent explanation for brand trust and an excellent percentage of

explaining consumer purchase intentions at R-square 0.509. The specific and total indirect

effects support the hypothesized role of consumer-level suspicion of an ulterior motive as a

mediator in the relationship between attitude toward reviews and brand trust and between

attitude toward reviews and purchase intentions. Brand trust is a mediator in the relationship

between suspicion of an ulterior motive and purchase intentions.

We also performed a PLS-MGA group analysis procedure in SmartPLS using the model

presented in Figure 1 for the three experimental groups: consumers who viewed verified,

unverified, and incentivized reviews before answering the survey questions. The effect of

consumer suspicion of an ulterior motive on brand trust is different for consumers who saw an

incentivized review vs. a verified review at p<0.05 level and for incentivized and unverified

reviews at p<0.01 level. While there are no differences regarding purchase intentions, the group

analysis does confirm the mediation effect of the review type and the significant differences

between the reviews with the highest vs. lowest level of legitimacy, verified vs. incentivized.

To test hypothesis 6, related to the moderation effects of the generational cohort, we

performed a PLS-MGA group analysis procedure in SmartPLS for four age groups: Generations

Z (65 consumers), Y (90), and X (167), as well as Baby Boomers (165). The multigroup analysis

emphasizes differences in the effects of suspicion of an ulterior motive on brand trust and

purchase intentions. The effect of brand trust on purchase intention is also different for

consumers from different generations. Regarding the effects of suspicion, the most significant

differences appear at the extremes of the age spectrum between Baby Boomers and Generation

Z. However, there are also significant differences between Generations X and Z. In the case of

brand trust and its impact on purchase intentions, differences are also appearing between

Millennials and Gen. Z, providing support for our expectations regarding a higher level of

market knowledge and suspicion from younger consumers, which also impacts the way they

interpret online reviews.

We also employed a conditional moderation-mediation regression analysis based on the

Process method (Hayes, 2017, 2018). We tested the model presented in Figure 1 and focused on

the moderation and mediation results and the unconditional and conditional effects. The results

show that the effects of suspicion of ulterior motive on brand trust and purchase intentions are

moderated by the type of review consumers read and the generational cohort. The findings

summarized in Table 5 show that consumers have various levels of consumption and deception

knowledge as a function of their experience and exposure.

(please insert Table 5 here)

Overall, the results of our quantitative analysis provide support for our proposed

improved framework of deception identification in consumer reviews and emphasize the

complementary role of IDT and PKM in the online written language context. The findings

emphasize the role of consumer knowledge and experience in the framework of online deception

and show the effects of review characteristics and consumer generational cohort on the impact of

online peer-to-peer communications.

4. Discussion and conclusions

The essential steppingstone that can be used by future research is represented by our new

framework of deception identification in the persuasive context of digital written reviews. The

legitimacy of online reviews can be identified based on our study’s four significant dimensions:

valence, authenticity level, analytical writing, and formal expressions. An integrative element of

interpersonal and computer-mediated communication in a persuasive context is represented by

suspicion, at the center of our framework, exhibited even in the asynchronous, written

environment of online reviews. This new framework represents a significant step forward for the

literature on identifying the main characteristics of deceptive messages among online consumer

reviews based on linguistic cues and lexical analysis. Also, the newly emphasized elements of

deception identification have a role in the previously formulated theories on interpersonal

deception and help adapt the theoretical agenda to the current digital circumstances. Theories

like IDT and PKM need to incorporate these elements of deception in digital language to account

for the specifics of electronic communication and to theorize information concealment,

falsification, and equivocation aspects in online consumer reviews based on different types of

deceptive cues.

4.1 Theoretical contributions

We evaluated consumer reaction to a possibly deceptive environment by assessing the

role played by suspicion of an ulterior motive of the reviewer in the deception model in an

integrative theoretical framework based on IDT and PKM. The results show the applicability of

the PKM and IDT model in deception in digital consumer reviews, especially in asynchronous

digital communication among consumers. The findings reiterate the importance of consumer

expectations and experience in a deception context (Burgoon et al., 1996; Buller et al., 1996) by

showing the importance of consumer attitudes toward reviews in the model in direct and

mediated relationships (Khare, Labrecque, and Asare, 2011). Therefore, applying this framework

in the context of online consumer reviews will help further research on deceptive

communication, heuristics used to interpret deception, and related to the effects of suspicion in

digital word-of-mouth communication.

We also assessed linguistic cues that consumers can use to identify fake and incentivized

reviews written digital communication context and evaluate how consumer suspicion of

deceptive communication influences their purchase intentions. We found that incentivized

reviews have a significantly more positive valence than both unverified and verified reviews,

with a significantly more extreme positive sentiment than even unverified reviews. This confirms

previous findings and underlines the effect of incentives in generating extreme positive valence

in reviews (Perez, 2016). Further, the qualitative and quantitative results highlight the formal

character of incentivized reviews and the more personal tone of verified reviews. The analysis

performed in LIWC also exhibits a higher level of details and extra-explanations provided by

consumers who received rewards, as shown by the authenticity index and the word/sentence

count. These cues are in tune with previous discussions on deception, noting that deceptive

communicators use fewer self-references and individual formulations (Buller et al., 1996; Zhou

et al., 2004). The indices also confirm the objective and detached styles on affective processes

and feelings (such as satisfaction) and perceptions of seeing, hearing, and feeling.

Communicators in incentivized reviews use a more objective and descriptive approach rather

than an experiential style in their comments. They do not describe their affective processes,

perceptions, and feelings related to their experience with the product reviewed. These results

represent potential cues that marketers and consumers can use in assessing deception in online

consumer comments.

Finally, the conclusions support the central role of consumer suspicion of an ulterior

motive from the reviewer and show the impact of incentivized consumer reviews. Our findings

show that consumer suspicion can negatively affect the attitude formation process and purchase

intentions in the context of online reviews (DeCarlo, 2005; DeCarlo, Laczniak, and Leigh, 2013).

Consumer suspicion regarding the reviewer is also a mediator for the effects of attitude toward

online reviews on consumer emotions and intentions, including brand trust and intentions to

purchase the product. The moderation role of review type as review legitimacy expresses the

adverse effects of suspicion on attitudinal and behavioral variables, as well as the potential that

the characteristics of one review for a fictitious brand have on the formation of consumer

intentions (DeCarlo, 2005; DeCarlo, Laczniak, and Leigh, 2013). The moderator effects shown

by the generational cohort also exhibit the role played by experience and skills in the deception

detection process (Buller and Burgoon, 1996).

4.2 Practical contributions

From a practical standpoint, companies need to consider the vulnerability of specific

generations based on lower levels of suspicion and distrust and formulate their short and long-

term marketing communication strategies accordingly. This analysis also provides businesses

with different cues to detect deception in online comments. Our results show the main

difficulties in dissimilating deception and emphasize deceptive characteristics of incentivized

reviews, which marketers can use to identify attempts of deception through concealment,

falsification, and equivocation from their competitors.

It is now easier for marketing practitioners to identify potentially deceptive online

reviews for their brand and their competitors based on the framework we propose, and the four

main characteristics analyzed: valence, authenticity, formalism, and analytical writing.

Moreover, by evaluating the essential cues in persuasive communication and the role played by

suspicion when interpreting consumer reviews, marketers can now more easily formulate a

digital reputation management campaign, manage their digital content, provide, and request

feedback from their consumers. In the short run, marketers can also improve their content

marketing strategies by promoting consumer reviews that attenuate consumer suspicion based on

the essential characteristics emphasized in this study.

In the long and medium-term, entrepreneurs also have business opportunities to create

services for organizations interested in increasing their level of legitimate reviews and their

feedback relation with consumers. These deception identification cues can create new long-term

opportunities for marketing communicators to adapt and change their business model and

integrate more proactive deception management measures. Also, long-term, marketers can

reduce the overall level of consumer suspicion and skepticism and increase brand trust by

promoting reviews that have lower levels of deceptive characteristics and by providing

consumers access to constructive market knowledge. Considering the exponential growth of

digital communication, policymakers and regulators also need to reflect on the role of suspicion,

consumer skepticism, the potential for market knowledge to help consumers deal with deception,

and the role of policies and education campaigns in reducing deceptive communication and

decreasing overall consumer suspicion levels.

4.3 Future research

This study has some limitations, mainly related to its sample focused on the U.S. market

and reviews posted on Amazon. Therefore, it would be interesting to see the differences in

results when examining reviews for other types of products, such as high-value products and

services, and in a cross-cultural context, especially in a high/low context cultural framework.

Nevertheless, there is also potential interest in studying this topic on different review platforms,

including etailers, retailers, as well as review aggregators.

Moreover, thanks to big data mining and natural language processing (NLP), we can now

perform larger scale, cross-cultural, integrative, and comprehensive machine-learning-based

analyses on a text to identify the critical markers of deception. In practice, we have some review

checker software options, and in research, we have attempts of studies on deception. However,

we need more thorough studies that can provide comprehensive frameworks based on theory and

data, allowing businesses and consumers to identify deception in various eWOM circumstances,

especially in computer-mediated communication and based on automated lexical analysis.

Finally, numerous topics stem from the widespread use of incentivized and deceptive

reviews that need additional attention. As mentioned, there are numerous options now of using a

review deception checker, usually integrated into the browser, which calculates the authenticity

score of a particular online seller. It would be interesting to analyze consumer attitudes towards

deceptive reviews and their intentions of using these aids and behavioral outcomes. Moreover, as

online reputation management companies and review influencers are becoming more common

and accepted on the market, research will also need to focus on differences within incentivized

reviews and consumer attitudes and behaviors towards this new type of marketing content.

References

Aditya, R. N. (2001), “The psychology of deception in marketing: A conceptual framework for

research and practice”, Psychology and Marketing, Vol. 18 No. 7, pp. 735–761.

Anderson, E. T. and Simester, D. I. (2014), “Reviews without a purchase: low ratings, loyal

customers, and deception”, Journal of Marketing Research, Vol. 51 No. 3, pp. 249–269. s

Artz, N. and Tybout, A.M. (1999), “The moderating impact of quantitative information on the

relationship between source credibility and persuasion: A Persuasion Knowledge Model

interpretation”, Marketing Letters, Vol. 10 No. 1, pp. 51–63.

Bagozzi, R. P. and Youjae Y. (1988), “On the evaluation of structural equation models”, Journal

of the Academy of Marketing Science, Vol. 16 No. 1, pp. 74-94.

Bai, X. (2011), “Predicting consumer sentiments from online text”, Decision Support Systems,

Vol. 50 No. 4, pp. 732-742.

Bambauer-Sachse, S. and Mangold, S. (2013), “Do consumers still believe what is said in online

product reviews? A persuasion knowledge approach”, Journal of Retailing and

Consumer Services, Vol. 20 No. 4, pp. 373–381.

Banerjee, S. and Chua, A. Y. K. (2017), “Authentic versus fictitious online reviews: A textual

analysis across luxury, budget, and mid-range hotels”, Journal of Information Science,

Vol. 43 No. 1, pp. 122–134.

Buller, D. B., Burgoon, J. K., Buslig, A. and Roiger, J. (1996), “Testing Interpersonal Deception

Theory: The language of interpersonal deception”, Communication Theory, Vol. 6 No. 3,

pp. 268–289.

Burgoon, J. K., Buller, D. B., Ebesu, A. S., White, C. H. and Rockwell, P. A. (1996), “Testing

interpersonal deception theory: Effects of suspicion on communication behaviors and

perceptions”, Communication Theory, Vol. 6 No. 3, pp. 243–267.

Burgoon, J.K., Chen, F. and Twitchell, D.P. (2010), “Deception and its detection under

synchronous and asynchronous computer-mediated communication”, Group Decision

and Negotiation, Vol. 19 No. 4, pp. 345–366.

Burgoon, J.K. and Nunamaker, J.F. (2004), “Toward computer-aided support for the detection of

deception”, Group Decision and Negotiation, Vol. 13 No. 1, pp. 1–4.

Büschken, J. and Allenby, G. M. (2016), “Sentence-based text analysis for customer reviews”,

Marketing Science, Vol. 35 No. 6, pp. 953-975.

Campbell, M. C. and Kirmani, A. (2000), “Consumers’ use of persuasion knowledge: The effects

of accessibility and cognitive capacity on perceptions of an influence agent”, Journal of

Consumer Research, Vol. 27 No. 1, pp. 69–83.

Chakraborty, U. and Bhat, S. (2018), “The effects of credible online reviews on brand equity

dimensions and its consequence on consumer behavior”, Journal of Promotion

Management, Vol. 24 No. 1, pp. 57–82.

Crowne, D. P. and Marlowe, D. A. (1960), “A new scale of social desirability independent of

psychopathology”, Journal of Consulting Psychology, Vol. 24 No. 4, pp. 349-354.

de Langhe, B., Fernbach, P. M. and Lichtenstein, D. R. (2016), “Navigating by the stars:

Investigating the actual and perceived validity of online user ratings”, Journal of

Consumer Research, Vol. 42 No. 6, pp. 817–833.

DeCarlo, T. E. (2005), “The effects of sales message and suspicion of ulterior motives on

salesperson evaluation”, Journal of Consumer Psychology, Vol. 15 No. 3, pp. 238–249.

DeCarlo, T. E., Laczniak, R. N. and Leigh, T. W. (2013), “Selling financial services: The effect

of consumer product knowledge and salesperson commission on consumer suspicion and

intentions” Journal of the Academy of Marketing Science, 41(4), 418–435.

https://doi.org/10.1007/s11747-012-0319-0

Decker, R. and Trusov, M. (2010), “Estimating aggregate consumer preferences from online

product reviews”, International Journal of Research in Marketing, Vol. 27 No. 4, pp.

293–307.

Dellarocas, C. (2006), “Strategic manipulation of internet opinion forums: implications for

consumers and firms”, Management Science, Vol. 52 No. 10, pp. 1577-1593.

DePaulo, P. J. and DePaulo, B. M. (1989), “Can attempted deception by salespersons and

customers be detected through nonverbal behavioral cues?” Journal of Applied Social

Psychology, Vol. 19 No. 18, pp. 1552–1577.

DePaulo, B. M., Malone, B. E., Lindsay, J. J., Muhlenbruck, L., Charlton, K. and Cooper, H.

(2003), “Cues to deception”, Psychological Bulletin, Vol. 129 No. 1, pp. 74–118.

Doster, L. (2013), “Millennial teens design and redesign themselves in online social networks”,

Journal of Consumer Behavior, Vol. 12 No. 4, pp. 267-279

Evans, N.J. and Park, D. (2015), “Rethinking the Persuasion Knowledge Model: schematic

antecedents and associative outcomes of persuasion knowledge activation for covert

advertising”, Journal of Current Issues and Research in Advertising, Vol. 36 No. 2, pp.

157–176.

Fornell, C. and Larcker, D. F. (1981), “Evaluating structural equation models with unobservable

variables and measurement error”, Journal of Marketing Research, Vol. 18 No. 1, pp. 39-

50.

Friestad, M. and Wright, P. (1995), “Persuasion knowledge: lay people’s and researchers’ beliefs

about the psychology of advertising”, Journal of Consumer Research, Vol. 22 No. 1, pp.

62-74.

Friestad, M. and Wright, P. (1994), “The Persuasion Knowledge Model: How people cope with

persuasion attempts”, Journal of Consumer Research, Vol. 21 No. 1, pp. 1-31.

FTC (2009), “Federal Trade Commission 16 CFR Part 255. Guides concerning the use of

endorsements and testimonials in advertising”, available at:

https://www.ftc.gov/sites/default/files/attachments/press-releases/ftc-publishes-final-

guides-governing-endorsements-testimonials/091005revisedendorsementguides.pdf

Fuller, C.M., Biros, D.P., Burgoon, J. and Nunamaker, J. (2013), “An examination and validation

of linguistic constructs for studying high-stakes deception”, Group Decision and

Negotiation, Vol. 22 No. 1, pp. 117–134.

George, J.F. and Robb, A. (2008), “Deception and computer-mediated communication in daily

life”, Communication Reports, Vol. 21 No. 2, pp. 92–103.

Godes, D. and Mayzlin, D. (2004), “Firm-Created word-of-mouth communication: a field-based

quasi-experiment”, Marketing Science, Vol. 28 No. 4, pp. 721–739.

Hair, J. F., Ringle, C. M., and Sarstedt, M. (2013), “Partial least squares structural equation

modeling: Rigorous applications, better results, and higher acceptance”, Long Range

Planning, Vol. 46 Nos. 1–2, pp. 1-12.

Hauser, D.J., Ellsworth, P.C. and Gonzalez, R. (2018), “Are manipulation checks necessary?”,

Frontiers in Psychology, Vol. 9 No. 21, pp. 998-1008.

Hayes, A. F. (2018), “Partial, conditional, and moderated mediation: Quantification, inference,

and interpretation”, Communication Monographs, Vol. 85 No. 1, pp. 4-40.

Hayes, A. F. (2017), “Introduction to mediation, moderation, and conditional process analysis”,

2nd Ed., The Guilford Press, New York.

Henseler, J., Dijkstra, T. K., Sarstedt, M., Ringle, C. M., Diamantopoulos, A., Straub, D. W.,

Ketchen Jr, D. J., Hair, J. F., Hult, G. T. M. and Calantone, R. J. (2014), “Common

beliefs and reality about PLS: Comments on Ronkko and Evermann (2013)”,

Organizational Research Methods, Vol. 17 No. 2, pp. 182-209.

Hong, H., Xu, D., Wang, G. A. and Fan, W. (2017), “Understanding the determinants of online

review helpfulness: A meta-analytic investigation”, Decision Support Systems, Vol. 102,

pp. 1–11.

Hu, L. and Bentler, P. M. (1999), “Cutoff criteria for fit indexes in covariance structure analysis:

conventional criteria versus new alternatives,” Structural Equation Modeling, Vol. 6 No.

1, pp. 1-55.

Hu, N., Bose, I., Koh, N. S. and Liu, L. (2012), “Manipulation of online reviews: An analysis of

ratings, readability, and sentiments”, Decision Support Systems, Vol. 52 No. 3, pp. 674–

684.

Hu, N., Liu, L. and Sambamurthy, V. (2011), “Fraud detection in online consumer reviews”,

Decision Support Systems, Vol. 50 No. 3, pp. 614–626.

Hübner Barcelos, R. and Vargas Rossi, C. A. (2014) “Paradoxes and strategies of social media

consumption among adolescents”, Young Consumers, Vol. 15 No. 4, pp. 275-295.

Humphreys, A. and Wang, R. J. H. (2018), “Automated text analysis for consumer research”,

Journal of Consumer Research, Vol. 44 No. 6, pp. 1274–1306.

Isaac, M.S. and Grayson, K. (2016), “Beyond skepticism: can accessing persuasion knowledge

bolster credibility?”, Journal of Consumer Research, Vol. 43 No. 1994, pp. 895-912.

Khare, A., Labrecque, L. I. and Asare, A. K. (2011), “The assimilative and contrastive effects of

word-of-mouth volume: an experimental examination of online consumer ratings”,

Journal of Retailing, Vol. 87 No. 1, pp. 111-126.

Kim, J., Kim, J.E. and Marshall, R. (2016), “Are two arguments always better than one?:

Persuasion knowledge moderating the effect of integrated marketing communications”,

European Journal of Marketing, Vol. 50 Nos. 7–8, pp. 1399–1425.

Kirmani, A. and Campbell, M. C. (2004), “Goal seeker and persuasion sentry: How consumer

targets respond to interpersonal marketing persuasion”, Journal of Consumer Research,

Vol. 31 No. 3, pp. 573–582.

Kordrostami, E., Liu-Thompkins, Y. and Rahmani, V. (2021), “Investigating the influence of

regulatory focus on the efficacy of online review volume versus valence”, European

Journal of Marketing, Vol. 55 No. 1, pp. 297–314.

Kostyra, D. S., Reiner, J., Natter, M. and Klapper, D. (2016), “Decomposing the effects of online

customer reviews on brand, price, and product attributes”, International Journal of

Research in Marketing, Vol. 33 No. 1, pp. 11–26.

Lawlor, M.A., Dunne, Á. and Rowley, J. (2016), “Young consumers’ brand communications

literacy in a social networking site context”, European Journal of Marketing, Vol. 50 No.

11, pp. 2018–2040.

Lee, T. Y. and Bradlow, E. T. (2011), “Automated marketing research using online customer

reviews”, Journal of Marketing Research, Vol. 48 No. 5, pp. 881–894.

Lepkowska-White, E., Brashear, T. G. and Weinberger, M. G. (2003), “A test of ad appeal

effectiveness in Poland and the United States - the interplay of appeal, product, and

culture”, Journal of Advertising, Vol. 32 No. 3, pp. 57–66.

Lin, C. A. and Xu, X. (2017), “Effectiveness of online consumer reviews: The influence of

valence, reviewer ethnicity, social distance and source trustworthiness”, Internet

Research, Vol. 27 No. 2, pp. 362–380.

Lunardo, R. and Roux, D. (2015), “In-store arousal and consumers’ inferences of manipulative

intent in the store environment”, European Journal of Marketing, Vol. 49 No. 5–6, pp.

646–667.

Ma, Y. J. and Lee, H. H. (2014), “Consumer responses toward online review manipulation”,

Journal of Research in Interactive Marketing, Vol. 8 No. 3, pp. 224–244.

Malbon, J. (2013), “Taking fake online consumer reviews seriously”, Journal of Consumer

Policy, Vol. 36 No. 2, pp. 139–157.

Markowitz, D. M. and Hancock, J. T. (2016), “Linguistic obfuscation in fraudulent science”,

Journal of Language and Social Psychology, Vol. 35 No. 4, pp. 435–445.

Martin, W. C. (2014), “Independent versus incentivized word-of- mouth: effects on listeners”,

Academy of Marketing Studies Journal, Vol. 18 No. 1, pp. 261-271.

Mayzlin, D. (2006), “Promotional chat on the Internet”, Marketing Science, Vol. 25 No. 2, pp.

155–163.

Mayzlin, D., Dover, Y. and Chevalier, J. (2014), “Promotional reviews: An empirical

investigation of online review manipulation”, American Economic Review, Vol. 104 No.

8, pp. 2421–2455. s

McCornack, S.A. (1992), “Information manipulation theory”, Communication Monographs,

Vol. 59, pp. 1-16.

McCornack, S. A., Morrison, K., Paik, J. E., Wisner, A. M. and Zhu, X. (2014), “Information

Manipulation Theory 2: A propositional theory of deceptive discourse production”,

Journal of Language and Social Psychology, Vol. 33 No. 4, pp. 348–377.

Moon, S. and Kamakura, W. A. (2017), “A picture is worth a thousand words: Translating

product reviews into a product positioning map”, International Journal of Research in

Marketing, Vol. 34 No. 1, pp. 265–285.

Motyka, S., Grewal, D., Aguirre, E., Mahr, D., de Ruyter, K. and Wetzels, M. (2018), “The

emotional review–reward effect: how do reviews increase impulsivity?” Journal of the

Academy of Marketing Science, Vol. 46 No. 6, pp. 1032–1051.

Munzel, A. (2015), “Malicious practice of fake reviews: Experimental insight into the potential

of contextual indicators in assisting consumers to detect deceptive opinion spam”,

Recherche et Applications en Marketing (English Edition), Vol. 30 No. 4, pp. 24–50.

Munzel, A. (2016), “Assisting consumers in detecting fake reviews: The role of identity

information disclosure and consensus”, Journal of Retailing and Consumer Services, Vol.

32, pp. 96–108.

Nelson, M. R., Atkinson, L., Rademacher, M. A. and Ahn, R. (2017), “How media and family

build children’s persuasion knowledge”, Journal of Current Issues & Research in

Advertising, Vol. 38 No. 2, pp. 165-183.

Netzer, O., Feldman, R., Goldenberg, J., and Fresko, M. (2012), “Mine your own business:

market-structure surveillance through text mining”, Marketing Science, Vol. 31 No. 3,

pp. 521–43.

Newman, M. L., Pennebaker, J. W., Berry, D. S. and Richards, J. M. (2003), “Lying words:

Predicting deception from linguistic style”, Personality and Social Psychology Bulletin,

Vol. 29, pp. 665- 675.

Nunnally, J. (1978), Psychometric Theory, McGraw-Hill, New York, NY.

Ott, M., Cardie, C. and Hancock, J. (2012), “Estimating the prevalence of deception in online

review communities”, Proceedings of the 21st international conference on World Wide

Web, pp. 201–210.

Peng, L., Cui, G., Zhuang, M. and Li, C. (2016), “Consumer perceptions of online review

deceptions: an empirical study in China”, Journal of Consumer Marketing, Vol. 33 No. 4,

pp. 269–280.

Pennebaker, J. W., Chung, C. K., Frazee, J., Lavergne, G. M. and Beaver, D. I. (2014), “When

small words foretell academic success: The case of college admissions essays”, PloS

One, Vol. 9, pp. 110.

Pennebaker, J.W., Booth, R.J., Boyd, R.L. and Francis, M.E. (2015), “Linguistic Inquiry and

Word Count: LIWC2015”, Pennebaker Conglomerates (www.LIWC.net), Austin, TX.

Perdue, B.C. and Summers, J.O. (1986), “Checking the success of manipulations in marketing

experiments”, Journal of Marketing Research, Vol. 23 No. 4, pp. 317–326.

Perez, S. (2016), “Amazon makes good on its promise to delete “incentivized” reviews”, Tech

Crunch, available at: https://techcrunch.com/2016/11/23/amazon-made-good-on-its-

promise-to-delete-incentivized-reviews/

Plotkina, D., Munzel, A. and Pallud, J. (2020), “Illusions of truth—Experimental insights into

human and algorithmic detections of fake online reviews”, Journal of Business Research,

Vol. 109, pp. 511–523.

Podsakoff, P. M., S. B. MacKenzie, J. Y. Lee, and N. P. Podsakoff (2003), “Common method

biases in behavioral research: a critical review of the literature and recommended

remedies”, Journal of Applied Psychology, Vol. 88 No. 5, pp. 879-903.

Pongjit, C. and Beise-Zee, R. (2015), “The effects of word-of-mouth incentivization on

consumer brand attitude”, Journal of Product and Brand Management, Vol. 24 No. 7, pp.

720–735.

Reimer, T. and Benkenstein, M. (2016), “When good WOM hurts and bad WOM gains: The

effect of untrustworthy online reviews”, Journal of Business Research, Vol. 69 No. 12,

pp. 5993–6001.

Riquelme, I. P. and Román, S. (2014), “The influence of consumers’ cognitive and

psychographic traits on perceived deception: A comparison between online and offline

retailing contexts”, Journal of Business Ethics, Vol. 119 No. 3, pp. 405–422.

Salehan, M. and Kim, D. J. (2016), “Predicting the performance of online consumer reviews: A

sentiment mining approach to big data analytics”, Decision Support Systems, Vol. 81, pp.

30–40.

Schoenmueller, V., Netzer, O. and Stahl, F. (2020), “The polarity of online reviews: prevalence,

drivers and implications”, Journal of Marketing Research, Vol. 57 No. 5, pp. 853–877.

Singh, J. P., Irani, S., Rana, N. P., Dwivedi, Y. K., Saumya, S. and Kumar Roy, P. (2017),

“Predicting the “helpfulness” of online consumer reviews”, Journal of Business

Research, Vol. 70, pp. 346–355.

Situmeang, F., de Boer N. and Zhang, A. (2020), “Looking beyond the stars: A description of

text mining technique to extract latent dimensions from online product reviews”,

International Journal of Market Research, Vol. 62 No. 2, pp. 195-215.

Soper, S. (2015), “Amazon targets web freelancers peddling fake customer reviews”, Bloomberg,

available at: http://www.bloomberg.com/news/articles/2015-10-16/amazon-targets-web-

freelancers-peddling-fake-customer-reviews

Steward, M. D., Burns, A. C., Morgan, F. N. and Roehm, M. L. (2020), “Credible effects: the

impact of disclosure of material connections within online product reviews”, Journal of

Public Policy and Marketing, Vol. 39 No. 3, pp. 353–368.

Tang, C. and Guo, L. (2015), “Digging for gold with a simple tool: validating text mining in

studying electronic word-of-mouth (eWOM) communication”, Marketing Letters, Vol.

26 No. 1, pp. 67-80.

Thelwall, M. (2016), “Sentiment analysis for small and big data”, in N. G. Fielding, & R. M. Lee

(Eds.), The SAGE Handbook of Online Research Methods, Sage, pp. 344-360.

Tirunillai, S. and Tellis, G. J. (2014), “Mining marketing meaning from online chatter: Strategic

Brand analysis of big data using Latent Dirichlet Allocation”, Journal of Marketing

Research, Vol. 51 No. 4, pp. 463–479.

Toma, C. L. and Hancock, J. T. (2012), “What lies beneath: The linguistic traces of deception in

online dating profiles”, Journal of Communication, Vol. 62 No. 1, pp. 78–97.

Vantiv (2018), “Do online reviews affect your institution’s reputation?” available at:

https://www.vantiv.com/vantage-point-fi/smarter-payments/online-reviews-reputation

Xiao, B. and Benbasat, I. (2011), “Product related deception in e-commerce: a theoretical

perspective”, MIS Quarterly, Vol. 35 No. 1, pp. 169–195.

Zhou, U., Burgoon, J. K., Twitchell, D. P., Qin, T. and Nunamaker Jr., J. F. (2004), “A

comparison of classification methods for predicting deception in computer-mediated

communication”, Journal of Management Information Systems, Vol. 20 No. 4, pp. 139-

165.

Figure 1: Conceptual framework

Figure 2: Themes in consumer reviews: incentivized, unverified, verified

3 3

16 16

Verified

Unverified

Incentivized

Verified

Unverified

Incentivized

Unverified

Verified

Air Clean Good

deals

Kitty Price Sample Scent Scented Shipping Smell

References by Theme

Figure 3: Consumer reviews PLS model results

Table 1: The IDT and PKM framework

Theory

Core Theory

Domains

Takeaway

Gaps

IDT

Deception is “a message

knowingly transmitted by

a sender to foster a false

belief or conclusion by the

receiver,” including the

interactive process of

monitoring and adjusting

communication based on

verbal and non-verbal cues

and mutual responses

(Buller and Burgoon,

1996)

decision sciences

(Burgoon et al., 2010)

• participants in synchronous communication are

more involved, perceive more mutuality and more credibility

- Scarcity of empirical

and theoretical work at the

intersection of computer-

mediated communication and

deception;

- The need for more

research on automated

linguistics-based cues

decision sciences (Carlson

et al., 2004)

• success in detecting deception depends on the

medium used

decision sciences (Zhou et

al., 2004)

• linguistic constructs and LBC components are

helpful in detecting deception

decision sciences (Fuller

et al., 2013)

• linguistic-based cues effectiveness: quantity,

specificity, affect, diversity, uncertainty, nonimmediacy,

activation

communication (George

and Robb, 2018)

• deception is common in communication, for 22% to

25% of social interactions

PKM

When consumers are

exposed to persuasion

attempts, they activate

strategies to defend against

these messages based on

experience and knowledge

of the market, leading to

increased skepticism

(Friestad and Wright,

1994, 1995)

advertising (Evans and

Park, 2015)

• alternative outcomes to attitude formation even

when P.K. is activated

- The process of

persuasion acceptance;

managing the explicit

recognition of advertising or

persuasive intent;

- Conditions under

which consumers use P.K. in

digital communication;

- Assess circumstances

with traditional, skepticism-

focused manipulations of P.K.

access

consumer research

(Campbell and Kirmani,

2000; Kirmani and

Campbell, 2004)

• the accessibility of persuasion motives and the

cognitive capacity of the consumer affect the use of P.K.

consumer research (Isaac

and Grayson, 2017)

• P.K. access can lead to a higher level of credibility,

trust, and belief in persuasive messages

IMC (Kim et al., 2016)

• reduced synergistic effects of advertising and

publicity when consumers activate P.K.

marketing (Artz and

Tybout, 1999)

• source-message incongruity and source bias lead to

negative inferences about the manipulative intent of the

source

Table 2: T-test results

Sig. (2-

tailed)

Mean

Dif.

95% Conf. Int.

Lower

Upper

H2a:Negations

10.674

0.009

1.783

1.065

2.502

H2b:Authentic

7.430

0.018

60.267

25.365

95.169

H2c:Analytical

Thinking

27.314

0.001

57.483

48.428

66.538

H2d:Informal

language

7.874

0.016

0.593

0.269

0.918

WPS

12.418

0.006

13.130

8.581

17.679

Numbers

3.079

0.091

2.307

-0.917

5.530

Affective

processes

11.754

0.007

8.367

5.304

11.429

Perceptual

processes

13.360

0.006

7.103

4.816

9.391

Table 3: Construct reliability and validity

Att.

Reviews

Brand

Trust

Purchase

Intentions

Suspicion

Cronbach’s

Alpha

Comp.

Reliab.

AVE

Att.

Reviews

0.840 0.863 0.905 0.705

Brand

Trust

-0.380 0.917

0.905 0.940 0.840

Purchase

Intentions

-0.431 0.692 0.906

0.890 0.932 0.820

Suspicion -0.254 0.240 0.333 0.877 0.861 0.909 0.769

Table 4: Consumer reviews model results

Path Coef.

t value

p-value

Direct effects

Att. Reviews -> Brand trust

0.341 7.488 0.001

Att. Reviews -> Suspicion

-0.254

6.713

0.001

Brand trust -> Purchase

0.650

19.279

0.001

Suspicion -> Brand trust

-0.153 3.580 0.001

Suspicion -> Purchase

-0.178

4.939

0.001

Specific and total indirect effects

Att. Reviews -> Brand trust

0.039

2.955

0.003

Att. Reviews -> Purchase

0.292

8.078

0.001

Att. Reviews -> Suspicion -> Brand trust

0.039

2.955

0.003

Att. Reviews -> Brand trust -> Purchase

0.221

6.375

0.001

Att. Reviews -> Suspicion -> Brand trust ->

Purchase

0.025 2.968 0.003

Att. Reviews -> Suspicion -> Purchase

0.045

3.576

0.001

Suspicion -> Purchase

-0.100

3.576

0.001

Table 5: Overall results

Hypothesis

Study/

analysis

value

Result

H1a

The level of review legitimacy reduces the effect of

consumer suspicion on brand trust.

Study 2,

PLS-

MGA

2.307

0.360

partially

supported*

H1b

The level of review legitimacy reduces the effect of

consumer suspicion on purchase intentions.

not

supported

H2a

Incentivized reviews have a more positive valence

compared to other categories.

Study 1,

LIWC, t-

test

10.674

0.009

supported

H2b

The level of authenticity is lower for incentivized

reviews.

7.430

0.018

supported

H2c

The level of analytical writing is lower for incentivized

reviews.

27.314

0.001

supported

H2d

The level of text formalism is lower for incentivized

reviews.

7.874

0.016

supported

H3a

Consumers’ attitude toward reviews is positively

related to consumers’ brand trust in a reviewed brand.

Study 2,

PLS-SEM,

Process

7.488

0.001

supported

H3b

Consumers’ attitude toward reviews is negatively

related to their level of suspicion of an ulterior motive

from the reviewer.

6.713

0.001

supported

H4a

Consumers’ level of suspicion of an ulterior motive

mediates the relationship between attitude toward

reviews and brand trust

2.955

0.003

supported

H4b

Consumers’ level of suspicion of an ulterior motive

mediates the relationship between attitude toward

reviews and purchase intentions

3.576

0.001

supported

H5a

Brand trust mediates the relationship between

suspicion of an ulterior motive and purchase intentions

2.968

0.004

supported

H5b

Brand trust mediates the relationship between attitude

toward reviews and purchase intentions

6.375

0.001

supported

H6a

Younger generational cohorts enhance the effect of

consumer suspicion of an ulterior motive on brand

trust.

Study 2,

PLS-

MGA

2.306

0.036

partially

supported**

H6b

Younger generational cohorts enhance the effect of

consumer suspicion of an ulterior motive on purchase

intentions.

1.816

0.961

partially

supported**

*Incentivized vs. verified

**Millennials vs. Baby Boomers

Appendix

Figure A1: Survey set-up example

Table A1: Measurement model

Att. Reviews

(Khare, Labrecque,

and Asare, 2011)

I am comfortable with reading online reviews.

0.848

I have used online reviews to help me make a decision about a

product or service.

0.853

In the past, my decisions have been influenced by reviews that I

read online.

0.804

I like to learn about others’ product and service experiences.

0.854

Brand Trust

(Goldsmith, Lafferty

and Newell 2001;

MacKenzie and Lutz

1989)

I trust this brand:

0.931

This brand is safe

0.897

This is an honest brand

0.922

Purchase

Intentions

(Lepkowska-White,

Brashear, and

Weinberger, 2003;

Lepkowska-White,

2005)

If I were looking for this type of product my likelihood of

purchasing this product would be high.

0.863

If I were to buy this type of product, the probability that I would

consider buying this product...

0.927

If I had to buy this type of product, my willingness to buy this

product would be high.

0.925

Suspicion (DeCarlo,

Laczniak, and Leigh,

2013)

The reviewer of the product has an ulterior motive.

0.804

The reviewer’s comments are suspicious.

0.932

The reviewer is motivated to exaggerate the benefits of the

product.

0.891