Journal of Business Research 177 (2024) 114582

Contents lists available at ScienceDirect

Journal of Business Research

journal homepage: www.elsevier.com/locate/jbusres

“Look me in the eye, customer”: How do face-to-face interactions in

peer-to-peer sharing economy services affect customers’ misbehavior
concealment intentions?
Edna Ozuna a, b, *, Lena Steinhoff a
a
Faculty of Business Administration and Economics, Paderborn University, Warburger Strasse 100, 33098 Paderborn, Germany
b
Faculty of Economics and Social Sciences, University of Rostock, Ulmenstrasse 69, 18057 Rostock, Germany

A R T I C L E I N F O A B S T R A C T

Keywords: Customer misbehavior poses a major risk in the sharing economy. For example, property damage to shared
Sharing economy accommodations imposes burdens on both sharing platforms and hosts, especially if misbehaving guests pur­
Customer misbehavior posefully, not coincidentally conceal, or fail to report damages. Such misbehavior might be facilitated by remote
Peer-to-peer services
listing management and the lack of face-to-face interactions between hosts and guests. Therefore, this study
Face-to-face interactions
Experimental research
investigated the effects of host–guest interaction modes (face-to-face, online-only) and frequency on guests’
misbehavior concealment intentions. Social identification and irritation emerged as bright- and dark side me­
diators, respectively. Guests who interacted face-to-face (vs. online-only) with hosts exhibited weaker intentions
to conceal their misbehavior due to increased social identification. Platforms can elicit social identification by
engaging guests in virtual communities. However, when face-to-face interactions become excessive, guests
experience irritation and are more likely to conceal their misbehavior. These insights offer practical implications
for both peer-to-peer sharing platforms and hosts.

1. Introduction intentional or unintentional, overt, or covert actions that violate

generally accepted norms of conduct in consumption settings (Fullerton
Don’t do what some of my guests do: say nothing and hope I don’t notice. & Punj, 2004; Harris & Reynolds, 2003). In a 2016 survey of 484
One guest used the hoover to clear the hot ash from the wood burner, with German Airbnb hosts, 37 % reported at least one case of property
inevitable results. Instead of telling me, an attempt was made to hide the damage, with 5 % reporting more than three such cases (Busch et al.,
fact that the hoover was seriously fire damaged. 2018). The risk of asset damage and theft, along with personal safety
—Forum entry dated March 26, 2019, by Airbnb host JohnClohesy1 concerns, ranks among the top five fears of U.K., U.S., and Chinese
(username) on quora.com providers that offer services through the sharing economy (Lloyd’s of
London, 2018). Such violations have an immediate impact on service
[The fact that the guests] left … pink wine or makeup [stains] on the white providers’ financial, physical, and psychological outcomes and may also
towels, a black splotch on the carpet, and oil drips on the fabric shower harm other customers (Harris & Reynolds, 2003) and operating
curtain concerned me. I was able to remove the stains but noticed they had platforms.
moved the coffee table trunk to cover the carpet stain! Why not just say Services in the sharing economy have a particularly high risk of
something? I’d be fine with that. I don’t like that they tried to hide it. misuse owing to their intrinsic nature. Users leverage the provided assets
—Forum entry dated August 7, 2019, by Airbnb host Willene1 without supervision or transfer of ownership, evoking a lower sense of
(username) on the Airbnb Community Center responsibility toward the accessed goods (Bardhi & Eckhardt, 2012;
Schaefers et al., 2016b). Building on a plethora of anecdotal evidence
These reviews are examples of thousands of blog entries on online and initial academic research on this potential dark side of the sharing
discussion forums for hosts of peer-to-peer (P2P) accommodation ser­ economy, we examine a specific manifestation of customer misbehavior,
vices that report instances of customer misbehavior, which include namely, the intention to conceal it in a purposeful (not coincidental)

* Corresponding author.
E-mail addresses: edna.ozuna@uni-paderborn.de, edna.ozuna@uni-rostock.de (E. Ozuna), lena.steinhoff@uni-paderborn.de (L. Steinhoff).

https://doi.org/10.1016/j.jbusres.2024.114582
Received 11 November 2022; Received in revised form 2 February 2024; Accepted 13 February 2024
Available online 3 April 2024
0148-2963/© 2024 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

attempt to avoid accountability. When guests in P2P accommodation Regarding the interaction mode, a face-to-face interaction increases
services refuse to assume financial responsibility for the intentional or guests’ social identification with the host, which subsequently reduces
unintentional damages they cause, whether by denying, failing to their intentions to conceal misbehavior. The timing of the face-to-face
report, or concealing them, hosts and platforms are on the hook. Airbnb interaction also has differential effects on social identification, such
reportedly spends US $50 million annually on payouts to hosts, that a personal interaction at check-in elicits greater social identification
including litigation settlements and property damage compensations and weaker intentions to conceal misbehavior than at check-out.
(Carville, 2021). Theoretically, we extend the battery of aspects that may enhance
Hosts’ concerns about property damage may also have intensified in group cohesion put forth by social identity theory (Tajfel & Turner,
response to professionalization and convenience trends, driven by the 1986). Specifically, we introduce the mode of interaction as a factor that
increasing use of technology to facilitate service interactions. That is, sparks differential group allegiance levels. However, based on attribu­
many hosts manage their listings remotely (e.g., exclusively online, tion processes (Heider, 1958; Weiner, 1985), more frequent host–guest
property management companies) and embrace self-check-in options (e. face-to-face interaction can lead to irritation without a significant in­
g., lockbox, keypad). Consequently, guests can access a property without crease in social identification, thereby enhancing guests’ misbehavior
any face-to-face or in-person interaction with the host. While concealment intentions. Extending previous service research that
technology-enabled interactions simplify commercial exchanges by investigated the factors within service providers’ control that may
overcoming geographical and synchronous communication barriers, impact customer misbehavior (Danatzis & Möller-Herm, 2023; Schae­
technologies that deprive communication partners of non-verbal fers et al., 2016b; Srivastava et al., 2022), we propose that service
communication cues can significantly modify their interaction dy­ providers in the sharing economy can deliberately influence customers’
namics and outcomes (Ahearne et al., 2022). Face-to-face interactions behavioral intentions by sparking differential psychological mecha­
reduce the risk of personal misrepresentation (McAlexander et al., 2002) nisms. We offer three key learnings for service managers. First, face-to-
and enhance group identity (Bouas & Arrow, 1996). Prior service face interactions, compared with online-only interactions, elicit social
research has extensively investigated the nature of technology-enabled identification and reduce guests’ misbehavior concealment intentions.
interactions (e.g., digital platforms, chatbots, service frontline robots, Second, to achieve this effect, the face-to-face interaction should occur
and artificial intelligence) and their outcomes for customers’ service during check-in. Third, too many service provider-initiated personal
evaluation and subsequent behavioral intentions (Belanche et al., 2020; interactions provoke irritation and subsequently increase guests’
Makarem et al., 2009; McLeay et al., 2021). Some initial studies have misbehavior and concealment intentions.
also noted the influence of interpersonal interactions and non-verbal Finally, we empirically test whether a virtual community created by
communication on sharing economy services (Luo & Zhang, 2016; the sharing platform functions as a contingency factor. Theoretically,
Moon et al., 2019). However, to date, we know of no research that tests membership in a virtual community can act as a natural mechanism that
their impact on customers’ intentions to conceal misbehavior. increases social identification and reduces misbehavior concealment
Considering this research gap, we explore the impact of face-to-face intentions. Therefore, face-to-face interactions are even more relevant
interactions among hosts and guests in the P2P sharing economy on guests’ for guests who are not members of a virtual community. Managerially,
misbehavior concealment intentions, test the psychological mechanisms un­ our study sheds light on practical instruments within the control of
derlying the effect, and identify contingencies. Our research responds to platforms and hosts to prevent customers’ misbehavior concealment.
calls to investigate the dark side of the sharing economy (Buhalis et al.,
2020; Eckhardt et al., 2019), particularly the legal and ethical tensions 2. Literature review
among different platform actors (Mosaad et al., 2023; Perren & Kozinets,
2018), and promising social strategies to prevent customer misbehavior Driven by technological and societal transformations (Wirtz et al.,
(Fombelle et al., 2020). By exploring the differential outcomes of face- 2019), the sharing economy has transformed the way customers access
to-face vs. online-only interactions, we respond to Moon et al.’s products and services, gaining popularity across various sectors,
(2019) call to assess the processes of offline and online social in­ including transportation and hospitality. According to Eckhardt et al.,
teractions among peer service providers and peer consumers. In so (2019, p. 3), the sharing economy refers to “a scalable socioeconomic
doing, we take into consideration “temporal factors as well as the in­ system that employs technology-enabled platforms to provide users with
tensity of the interaction, both of which may influence peers’ percep­ temporary access to tangible and intangible resources that may be
tions of P2P social interactions and outcome variables” (Moon et al., crowdsourced.” In the sharing economy, we can distinguish two relevant
2019, p. 413). business models based on the ownership of shared assets: in business-to-
With insights from our two sequential studies, we contribute to peer models, shared assets are owned by the platform (e.g., Zipcar and
extant knowledge in three main ways. First, we extend the literature on Lime), whereas in P2P models, they are owned by other platform users
customer misbehavior by analyzing misbehavior concealment intentions (e.g., Airbnb and Uber) (Benoit et al., 2017; Wirtz et al., 2019). This
among guests who interact with hosts, either face-to-face or only online. study focuses on P2P sharing economy services provided according to a
The results reveal that guests who experience a face-to-face interaction triadic structure (service provider–mediating platform–customer). Wirtz
with their host exhibit weaker misbehavior concealment intentions et al., (2019, p. 458) define P2P sharing economy services as “two- or
compared with those who interact with them only online. Theoretically, more-sided P2P online platforms through which people collaboratively
our research offers a novel perspective on customer misbehavior by provide and use capacity-constrained assets and resources.” In the P2P
examining concealment intentions, a specific misbehavior manifestation sharing economy, both service interactions and customer misbehavior
that is common in P2P sharing economy services due to the absence of represent relevant issues; however, they have not been linked in previ­
service provider supervision. We also extend previous work on service ous research. Table 1 lists selected contributions related to service en­
encounters and interactions (Luo & Zhang, 2016; Moon et al., 2019) by counters, specifically service interactions and customer misbehavior.
emphasizing the relevance of personal interactions and non-verbal This delineates this study’s contributions to various research areas.
communication for customers’ evaluations of service and behavioral
intentions, even in technology-enabled service settings. For managers, 2.1. Service interactions in the P2P sharing economy
these findings highlight the relevance of face-to-face interactions for
services in the sharing economy as an effective means of encouraging Research on service encounters can inform our understanding of the
customer compliance with social norms. role of service interactions in a P2P sharing economy setting. Because
Second, we explore the mechanisms underlying the effects of service interactions in the P2P sharing economy occur between
host–guest interactions on misbehavior concealment intentions. strangers, both customers and service providers tend to rely on

2
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

Table 1
Review of selected literature
Reference Focus of Theoretical Industry Context Research Key Findings Research Fields
Investigation Background Design
Service Customer Sharing
Interactions Misbehavior Economy

Face-to-face vs. Technology-enabled service interactions

Giebelhausen Employee rapport Role theory, script Hospitality Survey, In high-rapport service ✓ – –
et al. (2014) and service theory industry data encounters, the use of
encounter / Experiment technology results in
evaluation customers’ psychological
discomfort and a lower
evaluation of the
interaction than if
technology were not
present. The opposite
pattern occurs in service
encounters with low
rapport.
Jörling et al. Attribution of Attribution theory Interviews / - Customers are more ✓ – –
(2019) responsibility for Experiment likely to externally
outcomes obtained attribute positive
by service robots, outcomes of service robots,
and customer while negative outcomes
satisfaction are more likely to be
attributed internally,
though only if the
customer perceives
ownership of the robot.
- Lower responsibility for
positive outcomes could
decrease customers’ long-
term satisfaction with the
service robot.
Makarem et al. Customer Literature Call-center Survey / - Service convenience, ✓ – –
(2009) satisfaction in Experiment touch service process, tech
technology-enabled service process, and
service encounters service outcomes are
significant predictors of
customer satisfaction in
technology-enabled
service encounters.
- Customer satisfaction is
linked to positive word-of-
mouth and future business.
- Human touch is an
important factor in
customer satisfaction and
behavioral intentions,
even for relatively young,
tech-savvy customers.
Meuter et al. Sources of customer Self-service Survey / - Sources of satisfaction: ✓ – –
(2000) satisfaction/ technologies Critical ability to solve an
dissatisfaction with incident intensified need, relative
technology-based technique perceived advantage,
service encounters ability to do its intended
job.
- Sources of dissatisfaction:
technology failure, process
failure, poor design,
customer-driven failure.

Antecedents of Customer Misbehavior in General

Daunt and Future misbehavior General theory of Hospitality Survey - Past experience of – ✓ –
Harris intention crime, moral successful misbehavior is
(2011) determined by development associated with future
personality values, theory misbehavior intentions.
demographics and - Consumer alienation,
past misbehavior Machiavellianism,
sensation-seeking,
aggressiveness, self-
esteem, gender, age, and
education are associated
with past misbehaviors.
Daunt and Dysfunctional Broken windows Hospitality Survey - Customer dissatisfaction (✓) ✓ –
Harris customer behavior theory, routine and inequity may predict
(2012) determined by activity theory particular forms of
(continued on next page)

3
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

Table 1 (continued )
Reference Focus of Theoretical Industry Context Research Key Findings Research Fields
Investigation Background Design
Service Customer Sharing
Interactions Misbehavior Economy

physical and social customer misbehavior.

servicescape - Customer misbehavior is
variables, influenced by contextual
disaffection with factors. Through the
service, and control of servicescape
inequity variables (e.g., layout,
design, crowd density,
cleanliness), incidents of
deviant behavior could be
reduced.
Daunt and Spatio-temporal and Routine activity Retail Experiment - Socio-temporal factors (✓) ✓ –
Greer (2015) social drivers of theory, social and personality traits
customer theft impact theory affect theft likelihood.
- Environmental
opportunities for theft can
be mitigated by a high
sense of moral
development and self-
monitoring.
- Individuals are more
likely to steal in a crowd of
shoppers, especially if the
crowd is unknown to
them.
Wirtz and Kum Opportunistic Subjective Hospitality Experiment Both situational and – ✓ –
(2004) consumer behavior expected utility personality factors
theory, game influence cheating
theory, dissonance behavior.
theory - Situational: Potential
material gain has no effect
on consumer cheating,
whereas repeat purchase
intentions reduce
opportunistic behavior.
- Personality: Customer
satisfaction, morality, and
self-monitoring reduce
cheating, while
Machiavellianism
increases it.
Wirtz and Opportunistic Justice theory, Interview / - Customers are more – ✓ –
McColl- customer claiming theory of Experiment likely to be opportunistic
Kennedy neutralization, in their claiming during
(2010) moral intensity service recovery when
theory, dual they experience low
concern theory distributive, procedural,
and interactional justice.
- Customers are more
likely to be opportunistic
when dealing with large
companies, especially in
the case of one-time
transactions.

Customer Misbehavior in the Sharing Economy

Danatzis and Misbehavior Social information Co-working spaces Experiments / - Misbehavior contagion ✓ ✓ (✓)
Möller-Herm contagion and processing theory, (Sharing Field study occurs because customers
(2023) blame attributions attribution theory Economy), blame service personnel
airlines, and trains for the severe misbehavior
of other customers.
- Service personnel can
reduce misbehavior
contagion and increase
service provider attitude
through active
interventions.
Kim et al. Two-way customer Peer-to-peer Experiment When consumers receive – ✓ ✓
(2021) rating systems mobility services low ratings from the
company or service
providers on peer-to-peer
platforms, they likely
exhibit worse behavior
(continued on next page)

4
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

Table 1 (continued )
Reference Focus of Theoretical Industry Context Research Key Findings Research Fields
Investigation Background Design
Service Customer Sharing
Interactions Misbehavior Economy

than they had shown

before.
Jin et al. Consumer indirect Deterrence theory, Bike sharing Experiment - Customer misbehavior – ✓ ✓
(2022) misbehavior rational and intentions result from the
ethical decision- comprehensive
making theories consideration of benefits,
costs, potential sanctions,
and moral beliefs.
- Platforms should not use
isolated sanction-based
measures to inhibit
misbehavior but instead
should consider other
influencing factors.
- Platforms should
implement mechanisms to
reduce customers’
perception of the benefits
of misbehavior.
Schaefers et al. Misbehavior Broken windows Car sharing Interview / Misbehavior contagion is – ✓ ✓
(2016b) contagion theory Experiment attenuated through greater
brand strength and less
anonymity of the product’s
owner. Increased
communal identification
among access-based
service consumers reverses
the contagious effect.
Srivastava Misbehavior Focus theory of Bike sharing Experiment Misbehavior contagion is – ✓ ✓
et al. (2022) contagion normative attenuated through
conduct, social customer-company
identity theory, identification and reduced
broken windows interpersonal anonymity.
theory
Present study Misbehavior Social identity Peer-to-peer Experiment - Face-to-face (vs. online- ✓ ✓ ✓
concealment theory, attribution accommodation only) interactions elicit
intention theory services guests’ social
identification, which
reduces their intentions to
conceal their misbehavior.
- Three (vs. two) face-to-
face interactions cause
irritation, which enhances
guests’ intentions to
conceal their misbehavior.
- Platforms can elicit
guests’ social
identification by engaging
them in virtual
communities, such that
face-to-face interactions
are particularly effective in
reducing intentions to
cover misbehavior among
community nonmembers.

Notes: ✓ = covered by the study / (✓) = partially covered by the study / – = not covered by the study.

information provided by the platform, the other party, and other users cannot fully rely on self-service technologies because the human
(e.g., reviews and ratings) (Wirtz et al., 2019) to reduce information component is a significant determinant of customer satisfaction and
asymmetries and minimize risks (Ert et al., 2016). However, P2P com­ behavioral intentions, even for tech-savvy customers (Makarem et al.,
mercial exchanges also require relational mechanisms, such as trust (Ert 2009). Nevertheless, online-only interactions between customers and
et al., 2016; Ye et al., 2019), attachment (Yang et al., 2019), empathy service providers have become more prevalent in the digital age
(Pera et al., 2019), and solidarity (Suess et al., 2021), which, in turn, (Steinhoff et al., 2019). In accommodation service settings, for example,
influence behavioral intentions. Extant research highlights the impor­ increasingly professionalized platforms (e.g., Airbnb) rely largely on
tance of personal interactions for customers’ positive evaluation of a remote listing management. Hosts can grant guests access to their
service and their subsequent behavioral intentions. Non-verbal cues and property without any face-to-face contact, such that only approximately
emotions entailed in face-to-face service interactions play a vital role in 19 % of Airbnb U.S. revenues come from listings in which the owner is
customers’ assessment of the overall service experience and its quality actually present at the site (American Hotel and Lodging Association,
(Bitner et al., 1990; Farrell et al., 2001; Moon et al., 2019; Varca, 2009). 2017). Among the extensive academic research on service interactions,
In services involving both human and technological elements, firms we found a pertinent research gap regarding investigations of face-to-

5
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

face interactions between peers in sharing economy services and their (Schaefers et al., 2016b). In addition, Srivastava et al. (2022) revealed
outcomes regarding customers’ behavioral intentions. that low interpersonal anonymity (i.e., two consecutive customers) and
customer–company identification attenuated the effects of previous
2.2. Customer misbehavior in the P2P sharing economy misbehavior on misbehavior contagion in dockless bike-sharing systems.
However, when Danatzis and Möller-Herm (2023) investigated misbe­
Customer misbehavior represents a salient risk (Bardhi & Eckhardt, havior contagion in shared workspaces with service personnel, their
2012; Schaefers et al., 2016b). Transactions in the sharing economy findings indicated that it occurs because customers blame service staff
entail access to a certain resource instead of a transfer of ownership, and for the severe misbehavior of other customers. Our study expands this
when “consumers do not experience perceived ownership [they] avoid research stream into the antecedents of customer misbehavior by
identification with the accessed object of consumption” (Bardhi & exploring service providers’ interaction modes and frequencies as po­
Eckhardt, 2012, p. 849). Ownership creates legal, financial, social, and tential drivers of customer misbehavior. In contrast to the extant
performance risks and responsibilities (Moeller & Wittkowski, 2010; research that focuses on business-to-peer settings, this study seeks to
Schaefers et al., 2016a) that customers can avoid when using services in understand customer misbehavior in P2P contexts.
the sharing economy. The lack of a sense of responsibility for the
accessed resources may increase customers’ willingness to misuse them 2.3. Misbehavior concealment vs. Other forms of misbehavior
(Bardhi & Eckhardt, 2012; Schaefers et al., 2016b). In addition, two
particular characteristics of P2P accommodation services in the sharing According to Harris and Reynolds’ (2003) definition of customer
economy aggravate customer misbehavior risks. First, customers often misbehavior, customers may engage in such conduct overtly or covertly.
access accommodation assets without direct supervision from the owner To date, little research has investigated purposeful, non-coincidental
or platform. The lack of supervision while consuming the service in­ attempts to conceal misbehavior as a specific form of covert misbe­
creases the likelihood of misbehavior (Schaefers et al., 2016b) and re­ havior. This type of conduct has mainly been studied in organizational
duces the perceived risk of misbehavior being detected (Fisk et al., 2010; contexts to understand the actions employees might take after engaging
Wirtz & Kum, 2004). Second, people behave differently “at home” vs. in misbehavior (Bonner et al., 2017; Welsh et al., 2015), including
when they are “away.” From a sociological perspective, consumers tend concealment (Kundro & Nurmohamed, 2021; Mazar et al., 2008).
to perceive tourism as a permissive realm in which they can temporarily Concealment differs from other types of misbehavior to the extent that
suspend their adherence to social norms and values and engage in transgressors are aware of the original misbehavior yet take deliberate
deviant behavior (Turner & Ash, 1975; Uriely et al., 2011). Tourist actions to cover it up; it is a direct response to a prior transgression
spaces provide them with a sense of being less restrained, making them (Kundro & Nurmohamed, 2021). Whereas other misbehaviors can occur
more willing to take risks and be adventurous (Uriely & Belhassen, intentionally or unintentionally (Harris & Reynolds, 2003), concealment
2006; Wickens, 1997). Empirical evidence affirms that tourists exhibit is closely associated with personal gain as it enables transgressors to
less social control when “away” than when “at home,” along with a avoid detection and sanctions (Ashforth & Anand, 2003; Kundro &
reduced sense of psychological closeness, which can lead to misbehavior Nurmohamed, 2021).
(Wan et al., 2021). Bonner et al. (2017) suggest that people react to feelings of shame
Research into the antecedents of customer misbehavior in general and worry about reputational damage after misbehaving with self-
further identifies customer’s personal characteristics, such as morality, protecting behaviors. As most people are concerned with maintaining
emotions, attitudes (Babin & Babin, 1996), sensation seeking, aggres­ a positive self-concept, they seek to balance the tension between per­
siveness (Daunt & Harris, 2011), and demographic features (Babin & sonal gains from misbehaving and compatibility with their self-concept;
Babin, 1996; Daunt & Harris, 2011), as drivers of misbehavior. Situa­ for example, by rationalizing their actions (Mazar et al., 2008; Welsh
tional aspects can also affect misbehavior. First, the servicescape’s et al., 2015). Hence, misbehavior concealment poses a moral dilemma
physical and social characteristics can influence misbehavior intentions between doing the right thing and one’s self-interest. In addition, un­
because customers use these inputs to weigh potential gains against the ethical behaviors become progressively worse when people justify their
risks of misbehavior. Critical factors in these evaluations include the conduct, minimize their personal responsibility by attributing it to
target’s value, accessibility, vulnerability, the presence of service external factors, depersonalize victims, and avoid self-censure (Bandura,
personnel, social density (Daunt & Greer, 2015; Daunt & Harris, 2012), 1999; Welsh et al., 2015).
and repeat purchase intention (Wirtz & Kum, 2004). Second, variations As illustrated in the two exemplary opening posts, misbehavior
in the customer–firm relationship can trigger customer misbehavior, concealment represents a relevant problem for P2P accommodation
especially in the form of acts of revenge against the company (Daunt & services. Such instances typically manifest as property damage, which
Harris, 2012) or opportunistic behaviors. Wirtz and Kum (2004) suggest may occur deliberately or accidentally. In either case, it violates
that dissatisfied or moderately satisfied customers are more likely than generally accepted norms of conduct (Fullerton & Punj, 2004): if a
their satisfied counterparts to take advantage of service guarantees. customer damages property, normative expectations require them to
Similarly, when customers perceive that they have been treated unfairly, report it, take accountability, and pay for any necessary repairs. If cus­
they are more likely to exhibit opportunistic behavior (e.g., making false tomers purposefully avoid such accountability, the platform or host
claims) during service recovery processes (Wirtz & McColl-Kennedy, must bear the financial burden of the damage. Thus, concealment, rather
2010). than property damage itself, which may be accidental, implies a per­
Research on the specific antecedents of customer misbehavior in the sonal benefit derived from avoiding the financial and reputational costs
sharing economy context is scarce. Initial research suggests that it might that the original misbehavior might impose. In this sense, we focus on
be driven by situational aspects (e.g., encountering previous misbe­ customers’ misbehavior concealment intentions, which we define as a
havior by other customers) or service elements related to the platform specific form of misbehavior in which customers purposefully, not
(e.g., low ratings) (Schaefers et al., 2016b; Kim et al., 2021). Recent coincidentally, refuse to assume financial responsibility for the inten­
studies have explored how firms can prevent customer misbehavior from tional or unintentional damages they incur—whether by denying them,
spreading. In sharing services that lack service personnel, Schaefers failing to report them, or concealing them—resulting in detrimental
et al., (2016b) found that both the strength of the accessed product’s impacts on firms and service providers.
brand and the identifiability (cf. anonymity) of the owner reduced the Our review of the relevant literature reveals a pertinent research gap.
contagion of previous misbehavior in the context of car-sharing services. Specifically, given the intrinsic characteristics of services in the sharing
Managers might invest in brand building and more personal relation­ economy, which are not directly comparable with other settings in
ships with their customers to reduce the spread of misbehavior which customer misbehavior has been studied (e.g., retail, hotels,

6
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

restaurants), we find a void of empirical insights related to (1) the hypotheses about the impacts of the interaction mode on guests’
outcomes of personal vs. technology-enabled service interactions in the misbehavior concealment intentions (H1), mediated by their social
sharing economy, (2) the service provider–related antecedents of identification (H2), defined as the “positive emotional evaluation of the
customer misbehavior in sharing economy services, and (3) misbehavior relationship between the self and the in-group” (Postmes et al., 2013, p.
concealment intentions as a specific manifestation of customer misbe­ 599). We also hypothesize a moderating effect of virtual community
havior that is relevant to sharing economy settings. Therefore, we membership on guests’ misbehavior concealment intentions mediated
research the peer- and platform-side drivers of customer misbehavior by social identification (H3).
concealment intentions. By interweaving the three streams of research, According to attribution theory (Heider, 1958; Weiner, 1985), peo­
we propose that the mode of interaction and interaction frequency be­ ple make causal attributions to understand the behaviors and events
tween hosts and guests, as well as guests’ membership in a platform- surrounding them, which determine their affect and behavior. After a
provided community, affect customers’ likelihood of concealing their cognitive interpretation of a particular behavior or event, people attri­
misbehavior. bute its causes internally (to the self) or externally (to other people, the
environment, and situational factors), judge whether these causes are
3. Conceptual model and hypotheses temporal, and consider the extent to which they can control them
(Weiner, 1985). Depending on how a guest interprets the increasing
Based on social identity (Tajfel & Turner, 1986) and attribution interaction frequency initiated by the host, differential attributions and
(Heider, 1958; Weiner, 1985) theory, we derived hypotheses about the affective or behavioral responses may arise. We use attribution theory
effects of host–guest interaction mode and frequency on misbehavior (Heider, 1958; Weiner, 1985) to predict the impacts of increasing
concealment intentions, respectively. Fig. 1 illustrates the conceptual interaction frequency on guests’ misbehavior concealment intentions
model and provides an overview of the empirical program. (H4, H5, H6 and H4alt, H5alt, H6alt), mediated by either social identifica­
tion (H7, H8, H9) or irritation (H7alt, H8alt, H9alt), which we define as a
negative affective response caused by repeated exposure to a particular
3.1. Theoretical background (often unsolicited) stimulus (Morimoto & Chang, 2006; Pelsmacker &
Van den Bergh, 1998; Van Diepen et al., 2009).
According to social identity theory (Tajfel & Turner, 1986), people
define themselves in terms of the social groups to which they belong,
such that social identification emerges when they categorize themselves 3.2. Effects of face-to-face interaction on customers’ misbehavior
into a group, based on their emotional ties and develop a sense of concealment intentions
belonging, reflecting group affiliation. The salience of group identity
guides their personal attitudes and behaviors, including adherence to in- Despite the vast opportunities associated with the use of technology
group norms (Tajfel & Turner, 1986; see also Hogg & Reid, 2006). We in service delivery, extant research highlights the importance of non-
use social identity theory (Tajfel & Turner, 1986) to derive our verbal communication during service interactions and its impact on

Fig. 1. Conceptual model and hypotheses.

7
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

customers’ service evaluations and behavioral intention (e.g., Gabbott & of a face-to-face interaction on guests’ misbehavior concealment in­
Hogg, 2001; Sundaram & Webster, 2000). Non-verbal cues, defined as tentions, mediated by social identification, may be greater among guests
facial (e.g., eye contact, smile), bodily (e.g., personal distance), and who are not members of the virtual community. Accordingly, we
gestural (e.g., touch, wave) hints (Marinova, Singh, & Singh, 2018), postulate the following:
account for approximately 70 % of the content of a message (Mehrabian,
2017) and are actively perceived and processed in face-to-face in­ H3: Guests’ virtual community membership moderates the negative
teractions (Marinova, Singh, & Singh, 2018). Situational and relational indirect effect of the interaction mode on misbehavior concealment
cues entailed in non-verbal communication can significantly modify intentions, mediated by increased social identification. Specifically,
interaction dynamics and outcomes (Ahearne et al., 2022). Media with the negative indirect effect is weakened for members and strength­
few or no non-verbal cues may transmit the perception of not being in a ened for non-members of the virtual community, respectively, owing
real social situation, resulting in the reduced influence of social norms to the weaker (stronger) positive effects of a face-to-face (vs. online-
and an increased likelihood of people engaging in anti-normative be­ only) interaction on social identification for members (non-
haviors (Kahai & Cooper, 2003; Kiesler et al., 1984). However, face-to- members).
face interaction, with its inherent richness and situational and relational
cues, should favor compliance with social norms and reduce the likeli­ 3.4. Effects of face-to-face interaction frequency on customers’
hood of concealing initial misbehavior more than online-only in­ misbehavior concealment intentions
teractions. Accordingly, we propose the following:
Generally, frequent interactions with service providers allow cus­
H1: A face-to-face (vs. online-only) interaction between hosts and tomers to gain more information about a product or service, reducing
guests has a negative direct effect on guests’ misbehavior conceal­ uncertainty and building trust and loyalty (Palmatier et al., 2006).
ment intentions. Greater contact frequency also enhances customer-reported relationship
strength (Dagger et al., 2009), closeness, and commitment (Crosby et al.,
We propose that increased social identification is the psychological 1990). However, frequent service provider-initiated interactions are not
mechanism underlying this hypothesized direct effect. Specifically, non- always perceived favorably or appreciated; customers express distinct
verbal cues elicit cognitive and emotional states (Sproull & Kiesler, preferences for closeness and attachment, and some are not interested in
1986); thus, face-to-face interactions can facilitate the formation of building close commercial relationships (Mende et al., 2013). We
positive impressions (Okdie et al., 2011) and the development of group theorize that whether customers respond favorably or unfavorably to
identification (Bouas & Arrow, 1996; Guegan et al., 2017). In addition, frequent face-to-face interactions depends on their attribution of the
face-to-face interactions should favor group or social identification over service provider’s intentions. In line with attribution theory (Heider,
online-only interactions. Such social identification, in turn, may reduce 1958; Weiner, 1985), guests seek to understand the reasons for or in­
guests’ intentions to conceal their misbehavior and mediate the rela­ tentions behind hosts’ frequent contact. The identified causes might
tionship between the host–guest interaction mode and guests’ behav­ evoke positive (e.g., relationship interest and support) or negative (e.g.,
ioral intentions. In line with social identity theory (Tajfel & Turner, control and surveillance) attributions. Depending on whether guests
1986), if guests identify closely with the platform and its users, including regard greater interaction frequency positively or negatively, different
the host, their behavior should align with group norms, which include effects should arise on their misbehavior concealment intentions.
expectations that they take responsibility for any damage rather than On the one hand, if a guest attributes frequent interactions initiated
trying to conceal their misbehavior. Thus, we propose the following: by the host (e.g., genuine effort to deliver quality service and develop a
social relationship beyond commercial exchange) positively, it should
H2: A face-to-face (vs. online-only) interaction between hosts and enhance the host–guest relational strength and contribute to guest norm
guests has a negative indirect effect on guests’ misbehavior compliance. In social relationships, people try to avoid situations in
concealment intentions, mediated by increased social identification. which they might deceive their counterparts (Shalvi et al., 2011) or
infringe on norms to avoid compromising their social and self-image
3.3. Contingency role of a virtual community membership (Gross & Vostroknutov, 2022). Thus, social and self-image concerns
are relevant determinants of norm compliance (Gross & Vostroknutov,
Virtual communities are social aggregations in online environments 2022). In this sense, we hypothesize that more frequent interactions
that are built on common interests or experiences (Leal et al., 2014). strengthen the host–guest relationship, which contributes to guests’
Members come together in these spaces to exchange information or compliance with generally accepted norms of conduct (i.e., reduces their
engage in discussions about a shared interest guided by implicit or misbehavior concealment intentions). To test the impact of increased
explicit codes of conduct and shared values (Wu et al., 2010). Re­ interaction frequency, we compared one against two, two against three,
lationships among virtual community members are grounded in reci­ and three against four face-to-face interactions.
procity, trust, and commitment (Chan & Li, 2010; Leal et al., 2014; Wu
et al., 2010). Firms can leverage these communities to strengthen their H4: Two (vs. one) face-to-face interactions between hosts and guests
brand image, broaden marketing capabilities, establish relationships have negative direct effects on guests’ misbehavior concealment
with customers, enhance customer loyalty, and develop products and intentions.
services (Kim et al., 2004). H5: Three (vs. two) face-to-face interactions between hosts and
In line with social identity theory (Tajfel & Turner, 1986), we hy­ guests have negative direct effects on guests’ misbehavior conceal­
pothesize that guests who are members of a virtual community catego­ ment intentions.
rize themselves as part of that social group, adopt its norms, and identify H6: Four (vs. three) face-to-face interactions between hosts and
with other community members, including both guests and hosts. Thus, guests have negative direct effects on guests’ misbehavior conceal­
they exhibit greater social identification than guests who are not ment intentions.
members of the virtual community, such that the host–guest interaction
mode (face-to-face vs. online-only) should have little additional impact. If positive attributions of more frequent host-initiated interactions
In contrast, among non-members who do not experience social identi­ prevail, we also posit that repeated face-to-face interactions and their
fication due to community membership, the host–guest interaction associated relational strength should favor social identification more
mode could be decisive for eliciting social identification. We hypothe­ than a single face-to-face interaction. Thus, social identification should
size a moderated mediation effect, such that the negative indirect effect increase as face-to-face interactions between hosts and guests become

8
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

more frequent and mediate the direct relationship. 4. Study 1: Face-to-face interactions and virtual community
membership
H7: Two (vs. one) face-to-face interactions between hosts and guests
have negative indirect effects on guests’ misbehavior concealment In Study 1, we investigated the host–guest interaction mode to
intentions, mediated by increased social identification. establish whether a face-to-face interaction, compared with an online-
H8: Three (vs. two) face-to-face interactions between hosts and only interaction, affects guests’ intentions to conceal their misbe­
guests have negative indirect effects on guests’ misbehavior havior (H1). Drawing on social identity theory (Tajfel & Turner, 1986),
concealment intentions, mediated by increased social identification. we investigate the potential mediating role of social identification (H2).
H9: Four (vs. three) face-to-face interactions between hosts and We also explore the effectiveness of virtual community membership as a
guests have negative indirect effects on guests’ misbehavior platform design mechanism that may moderate the effects of the inter­
concealment intentions, mediated by increased social identification. action mode on social identification (H3).

On the other hand, guests may attribute their host seeking frequent
4.1. Research design, participants, and procedure
interactions with them as his/her intention to control them. Customers’
negative emotional and cognitive evaluations of service provision are
We conducted a scenario-based experiment to examine the effect of
associated with dysfunctional and retaliatory behaviors (Daunt & Har­
the host–guest interaction mode on P2P accommodation services in the
ris, 2012; Reynolds & Harris, 2009). This relationship has been exten­
sharing economy and the moderating impact of membership in a virtual
sively studied in prior empirical research (Daunt & Harris, 2012; Mills,
community. In a 2 × 2 between-subjects factorial design, we manipu­
1981; Wirtz & Kum, 2004), with results indicating that customers’
lated the guests’ mode of interaction with the host (face-to-face, online-
negative evaluations of a service feature enhance their likelihood of
only) and membership in a virtual community (membership, no mem­
engaging in misbehavior toward the service provider (Fisk et al., 2010;
bership). For data collection, we used Prolific and recruited a sample of
Harris & Reynolds, 2003; Wirtz & Kum, 2004). Thus, following a
233 U.S. participants. The mean age was 38.6 years, and 54.8 % were
negative attribution of host-initiated interaction frequency, the more
women. Regarding their experience with P2P accommodation services,
face-to-face encounters that take place, the greater the likelihood that
75.2 % indicated that they currently used at least one platform and, on
the guest engages in dysfunctional behavior, particularly, misbehavior
average, had stayed 4.38 nights in non-hotel accommodations per year.
concealment. Based on this, we propose the following:
Respondents were randomly assigned to one of four experimental
groups. Each participant received a scenario and a questionnaire. The
H4alt: Two (vs. one) face-to-face interactions between hosts and
scenario described a fictitious online platform, sharingmyplace.com, as
guests have positive direct effects on guests’ misbehavior conceal­
an intermediary between homeowners (i.e., hosts) and customers (i.e.,
ment intentions.
guests) that helps travelers book overnight stays at privately owned
H5alt: Three (vs. two) face-to-face interactions between hosts and
vacation accommodations. The scenario also introduced a “Sharing
guests have positive direct effects on guests’ misbehavior conceal­
Buddies” virtual community as a complimentary, optional service of the
ment intentions.
platform that enables hosts and guests to exchange information, travel
H6alt: Four (vs. three) face-to-face interactions between hosts and
advice, and photos and thus develop closer personal relationships. We
guests have positive direct effects on guests’ misbehavior conceal­
asked participants to imagine that they were customers of the platform
ment intentions.
who had booked an apartment for an upcoming holiday.
To manipulate the interaction mode, we assigned participants to
If such negative attributions dominate, they may spark irritation
either a personal interaction (face-to-face) or an exclusively online
among guests. Specifically, overly frequent contacts from service pro­
interaction (online-only) with their host. The former group learned that
viders, especially when unsolicited, may evoke feelings of irritation and
the host provided all the relevant information about the apartment
perceptions of a loss of control or intrusiveness (Godfrey et al., 2011;
during their personal encounter, whereas the latter received the same
Morimoto & Chang, 2006). Following attribution theory (Heider, 1958;
information via e-mail. To signal the opportunity to conceal misbe­
Weiner, 1985), guests seek to understand the causes of such repeated
havior, all scenarios noted that the host would be out of town during the
contact from the host and characterize them as external (i.e., at the
check-out process and thus could not check the state of the apartment at
host’s discretion). In this sense, we predict that guests might perceive
the end of the stay. We manipulated community membership by indi­
repeated face-to-face interactions initiated by hosts as unsolicited and
cating to participants that they were either members of the “Sharing
intrusive, evoking irritation as a negative emotional manifestation of the
Buddies” virtual community and sought social exchanges with other
threat to their freedom to engage in a social relationship. That is,
members to improve their travel experience (membership) or that,
although initial face-to-face interactions may enhance guests’ social
although they were aware of the virtual community, they were not
identification, incrementally increasing interaction frequency can lead
members (no membership). In all scenarios, the host was identified as a
to the opposite effect that instead of strengthening the relationship,
member of the virtual community. After reading the scenarios, all par­
prompts customer irritation and misbehavior concealment intentions.
ticipants completed a questionnaire to gauge their misbehavior
This is because customers rationalize their misbehavior based on their
concealment intentions and social identification with the host.
negative emotional perceptions of repeated face-to-face interactions.
Accordingly, we put forward the following hypotheses:
4.2. Manipulation and realism checks
H7alt: Two (vs. one) face-to-face interactions between hosts and
guests have positive indirect effects on guests’ misbehavior The manipulation checks supported the effectiveness of our treat­
concealment intentions, mediated by increased irritation. ments. To test the manipulation of the interaction mode, we asked
H8alt: Three (vs. two) face-to-face interactions between hosts and participants in all experimental groups to indicate whether they met
guests have positive indirect effects on guests’ misbehavior their host personally upon arrival (Mface-to-face = 3.34, SD = 2.61; Monline
concealment intentions, mediated by increased irritation. = 1.35, SD = 0.99; t = 7.83, p <.01), and whether they were members of
H9alt: Four (vs. three) face-to-face interactions between hosts and a virtual community (Mmember = 5.95, SD = 1.71; Mnonmember = 2.15, SD
guests have positive indirect effects on guests’ misbehavior = 1.91; t = 16.2, p <.01). The results of the realism checks also indicated
concealment intentions, mediated by increased irritation. that participants could easily imagine the described situation (M = 6.30,
SD = 0.90) and envision themselves within it (M = 6.26, SD = 0.95).

9
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

4.3. Measures and validity assessment number of overnight stays in non-hotel accommodations, and the
number of platforms participants used. Our results show no significant
We used a multi-item scale (see Table 2) to measure social identifi­ direct effect of a face-to-face (vs. online-only) host–guest interaction on
cation (Doosje et al., 1995, 1998). We measured participants’ misbe­ guests’ misbehavior concealment intentions (b = 0.03, SE = 0.25, p
havior concealment intentions with a single item (Bergkvist & Rossiter, =.91, R2 = 0.03). Thus, H1 was not supported. To test our mediation and
2007; Fuchs & Diamantopoulos, 2009): “I would not notify [the host] moderation hypotheses, we analyzed the data using PROCESS (version
about a damage I made in the apartment.” As recommended by prior 4.0, Hayes, 2021), which estimates direct and indirect effects in single
customer misbehavior research (Daunt & Harris, 2011, 2012; Schaefers and multiple mediation models based on the ordinary least squares
et al., 2016b), we controlled for social desirability effects (Paulhus, regression and bootstrapping methods (Hayes, 2018) and found a
1991) using a multi-item scale (Hart et al., 2015; Table 2). Similarly, we negative indirect effect of a face-to-face (vs. an online-only) interaction
included respondents’ demographics (i.e., age and gender) and experi­ on guests’ misbehavior concealment intentions, mediated by an increase
ence with P2P accommodation services (i.e., number of overnight stays in social identification, which supports H2 (b = -0.13, 95 % confidence
in non-hotel accommodations per year and the number of platforms they interval [CI] = [-0.27, -0.02]; PROCESS Model 4). That is, we found a
currently used) as covariates in our model. All items were rated using a significant indirect-only mediation effect (Zhao et al., 2010), such that
seven-point Likert-type scale (1 = strongly disagree, 7 = strongly agree). social identification fully mediates the impact of the host–guest inter­
We assessed the convergent and discriminant validity of the social action mode on guests’ misbehavior concealment intentions, but no
identification and social desirability constructs using AMOS 28.0. Both direct effect appears. Table 3 presents the results.
scales exhibited convergent validity according to the factor loadings, Subsequently, we analyzed the moderating effect of virtual com­
Cronbach’s α, composite reliability (CR), and average variance extracted munity membership on the face-to-face (vs. online-only) host–guest
(AVE) scores, which exceeded the common thresholds (Table 2). Fornell interaction–social identification–misbehavior concealment intentions
and Larcker’s (1981) criterion was applied to test discriminant validity. relationship using PROCESS Model 7. We found a significant negative
interaction effect between virtual community membership and host–­
guest interaction mode on social identification (b = -0.67, p =.04). In
4.4. Results
line with H3, we also found a positive moderated mediation of the
interaction between virtual community membership and the host–guest
We conducted a linear regression analysis to test for the predicted
interaction mode on guests’ misbehavior concealment intentions
direct effect, controlling for the effects of age, gender, social desirability,

Table 2
Construct measures and validity assessment (Studies 1 and 2)
a b
Construct Items Factor M (SD) Cronbach’s α CR AVE Construct Correlations
b b b b b
Loading
MC SI IR SD

Misbehavior Concealment I would not notify (Host) about a 2.13 N.A. -0.19 0.04
Intentions (MC) damage I made in the apartment. (1.86)
3.04 N.A. -0.30 0.22 -0.16
(2.04)
Social Identification (SI) I identify myself with (Host). 0.79 4.15 0.78 0.79 0.57 -0.19 N.A. 0.27
(adapted from Doosje et al., (1.27)
1995, 1998) 0.79 4.77 0.80 0.81 0.60 -0.30 N.A. -0.17 0.27
(1.35)
I feel committed to (Host). 0.90
0.88
I am glad to be a guest at (Host’s) 0.53
apartment.
0.62
Irritation (IR)c It irritates me that (Host) comes by the 0.89 2.79 0.96 0.97 0.80 0.22 -0.17 N.A. -0.16
(adapted from Akaah et al., apartment without my consent. (1.70)
1995; Van Diepen et al., 2009) The frequency at which (Host) comes by 0.93
the apartment annoys me.
I find the interaction with (Host) 0.90
annoying.
I am bored by the interaction with 0.81
(Host).
The way (Host) comes by the apartment 0.92
is an invasion of my privacy.
It irritates me how often (Host) comes by 0.95
the apartment.
It would be better to reduce the 0.86
frequency of (Host) coming by the
apartment.
Social Desirability (SD) I always know why I like things. 0.56 4.88 0.73 0.74 0.50 0.04 0.27 N.A.
(adapted from Hart et al., 2015) (1.18)
0.58 5.20 0.76 0.77 0.53 -0.16 0.27 -0.16 N.A.
(1.18)
I am a completely rational person. 0.84
0.80
I am very confident of my judgements. 0.69
0.79

Notes: M = mean, SD = standard deviation, CR = composite reliability, AVE = average variance extracted, N.A. = not applicable.
a
Items were measured on a 7-point Likert-type scale ranging from 1 = strongly disagree to 7 = strongly agree.
b
The number/numbers in the first/second line refer/refers to Study 1/Study 2.
c
Study 2 only.

10
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

Table 3
Results of Study 1 (PROCESS Models 4 and 7)
Coefficient LLCI ULCI Hypothesis
(SE) Test

Indirect effects on misbehavior concealment intentions

Face-to-face interaction (vs. online-only) → social identification → misbehavior concealment intentions -0.13 -0.27 -0.02 H2: Supported
(0.06)
Face-to-face interaction (vs. online-only) x virtual community membership → social identification → misbehavior 0.20 0.01 0.49 H3: Supported
concealment intentions
(0.12)

Effects on social identification

Face-to-face interaction (vs. online-only) → social identification 0.42** 0.10 0.74
(0.16)
Face-to-face interaction (vs. online-only) × virtual community membership → social identification -0.67* -1.3 -0.04
(0.32)

Effects on misbehavior concealment intentions

Face-to-face interaction (vs. online-only) → misbehavior concealment intentions 0.15 -0.34 0.65
(0.25)
Social identification → misbehavior concealment intentions -0.30** -0.50 -0.01
(0.10)

Controls
Age → social identification 0.01 -0.01 0.01
(0.01)
Gender → social identification -0.37** -0.64 -0.10
(0.14)
Social desirability → social identification 0.25*** 0.11 0.39
(0.07)
Number of overnight stays in non-hotel accommodations → social identification 0.02 -0.01 0.04
(0.01)
Number of platforms in use → social identification -0.02 -0.19 0.14
(0.08)
Age → misbehavior concealment intentions -0.01 -0.03 0.01
(0.01)
Gender → misbehavior concealment intentions 0.22 -0.20 0.65
(0.21)
Social desirability → misbehavior concealment intentions 0.19 -0.03 0.40
(0.11)
Number of overnight stays in non-hotel accommodations → misbehavior concealment intentions -0.01 -0.03 0.03
(0.02)
Number of platforms in use → misbehavior concealment intentions -0.07 -0.33 0.19
(0.13)

Notes: Coefficients represent unstandardized estimates. Two-tailed test for hypothesized effects. SE = Standard Error; LLCI = lower limit confidence interval; ULCI =
upper limit confidence interval.
R2 Social Identification = 11.2 % (main mediation model), 13 % (moderated mediation model) / R2 Misbehavior Concealment Intentions = 6.21 % (main mediation
model), 6.21 % (moderated mediation model).
*p <.05. **p <.01. ***p <.001.

through social identification (index of moderated mediation = 0.20, 95 guests’ misbehavior concealment intentions, mediated by social identi­
% CI [0.01, 0.49]). Specifically, the conditional indirect effect was only fication, was significant only for non-members of the virtual community.
significant for guests who are not members of the virtual community (b
= -0.23, 95 % CI [-0.47, -0.06]), whereas the effect for members of the 5. Study 2: Bright and dark side effects of face-to-face
virtual community was not significant (b = -0.03, 95 % CI [-0.17, 0.12]). interactions

4.5. Discussion of findings The results of Study 1 implied that interacting face-to-face with
guests is advantageous for hosts. To consider the potential implications
Study 1 revealed differences in guests’ misbehavior concealment of more or less frequent face-to-face interactions, in Study 2, we
intentions depending on how they interacted with their host (face-to- empirically examined the direct effects of multiple host–guest face-to-
face or online-only): guests who experienced a face-to-face interaction face interactions on guests’ misbehavior concealment intentions (H4,
exhibit stronger social identification and thus weaker intentions to H4alt, H5, H5alt, H6, and H6alt) and empirically tested the mediating roles
conceal their misbehavior compared with guests who interacted exclu­ of social identification (H7, H8, and H9) and irritation (H7alt, H8alt, and
sively online. The data also revealed that membership in a virtual H9alt) sparked by repeated face-to-face interactions initiated by the host.
community moderates the effect of the host–guest interaction mode on
social identification. As social identification is evoked by virtual com­ 5.1. Research design, participants, and procedure
munity membership, the interaction mode is not relevant to virtual
community members and does not affect their adherence to group To analyze the effect of face-to-face interaction frequency on guests’
norms. Instead, the host–guest interaction mode exerts a relevant effect intention to conceal misbehaviors, we conducted a scenario-based
and elicits social identification for guests who are not members of the experiment. In a unifactorial experimental design with five treatment
virtual community. The negative effect of a face-to-face interaction on groups (interaction frequency), we manipulated the frequency (one

11
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

check-in, one check-out, and two, three, or four interactions) of face-to- so that the host does not find it right away?”.
face host–guest interactions. We did not go beyond four interactions We performed a confirmatory factor analysis for social identification,
because we deemed a larger number of encounters unrealistic in the irritation, and social desirability using AMOS 28.0. All scales exhibited
hospitality context. Therefore, we conducted a conservative test for the convergent validity (Table 2). We also confirmed Fornell and Larcker’s
proposed effects. Building on the findings of Study 1 and to gain insights (1981) discriminant validity criteria.
into the best timing for a single face-to-face interaction, we included two
groups (check-in or check-out interaction). We collected data from 5.4. Results
Prolific, comprising 457 U.S. respondents randomly assigned to one of
the five experimental groups. The mean age was 43.4 years, and 36.1 % Similar to Study 1, we conducted linear regression analyses to test
were women. Similar to Study 1, we checked their experience with P2P the proposed direct effects and the parallel mediation of social identi­
accommodation services, noting that 98.5 % were current users of at fication and irritation using PROCESS Model 4 (version 4.0, Hayes,
least one platform, and they stayed, on average, 6.21 nights in non-hotel 2021). We performed pairwise comparisons (i.e., two vs. one, three vs.
accommodations per year. two, and four vs. three) using individual regression models.
Similar to those in Study 1, participants received a scenario As aforementioned, we predicted the direct negative effects of
description and a questionnaire. The scenario once again describes the increasing the frequency of host–guest face-to-face interactions on
fictitious online platform sharingmyplace.com. In this case, we asked guests’ misbehavior concealment intentions in H4, H5, and H6. On the
respondents to envision themselves as customers of the platform who other hand, we hypothesized potential positive, direct effects in H4alt,
had booked a two-week stay in an apartment. To manipulate interaction H5alt, and H6alt. We pooled the two groups with a single face-to-face
frequency, participants assigned to the different treatment groups read interaction (check-in or check-out) into one group and compared it to
that they had (1) one face-to-face interaction with their host at arrival; a group with two interactions. The results show no significant direct
(2) one face-to-face interaction with their host at departure; (3) two effect of the two (vs. one) face-to-face interactions (b = -0.43, SE = 0.27,
face-to-face interactions with their host: once at arrival and another at p =.12, R2 = 0.04) on guests’ misbehavior concealment intentions.
departure; (4) three face-to-face interactions with their host: at arrival, Similarly, no direct effects emerged from the analysis of three (vs. two)
at departure, and once in between when the host came by the apartment face-to-face interactions (b = -0.10, SE = 0.31, p =.74, R2 = 0.08) nor of
to socialize and offer assistance; or (5) four face-to-face interactions with four (vs. three) face-to-face interactions (b = -0.30, SE = 0.29, p =.29,
their host: at arrival, at departure, and twice in between when the host R2 = 0.10). Thus, we found no empirical support for any of the hy­
came by the apartment to socialize and offer assistance. potheses related to the direct effects of increasing interaction frequency
Finally, to measure the guests’ misbehavior concealment intentions, on misbehavior concealment intentions (H4–H6 or H4alt–H6alt).
all scenarios indicated that the guest and their travel companions had Second, to build on Study 1′s insights regarding the social
left stains on the mattresses due to eating in the bedrooms, which identification-enhancing effects of face-to-face interactions, we hy­
violated the house rules. To simulate the opportunity to conceal their pothesized that an increasing number of personal interactions would
misbehavior, we indicated that additional dark-colored bedding sets increase social identification, which might reduce guests’ misbehavior
were available in the apartment, so they could choose to cover up the concealment intentions. Specifically, in H7, H8, and H9, we predicted the
stains by changing the sheets. After reading the scenarios, we measured negative indirect effects of two (vs. one), three (vs. two), and four (vs.
participants’ misbehavior concealment intentions, social identification, three) face-to-face interactions on misbehavior concealment intentions,
and irritation using a questionnaire. To replicate our approach in Study mediated by an increase in social identification. We found no significant
1, we controlled for the effects of socially desirable answers, de­ indirect effects from the analyses of two (vs. one) (b = -0.49, 95 % CI =
mographics (i.e., age and gender), and experience with P2P accommo­ [-1.01, 0.03]), three (vs. two) (b = -0.18, 95 % CI = [-0.78, 0.42]), or
dation services (i.e., the number of overnight stays in non-hotel four (vs. three) (b = -0.01, 95 % CI = [-0.24, 0.08]) face-to-face in­
accommodations per year and the number of platforms they used). teractions. Therefore, we cannot confirm the predicted bright side ef­
fects (H7–H9) of the increasing frequency of host–guest personal
5.2. Manipulation and realism checks interactions. Table 4 presents the detailed results.
Third, Study 2 offered interesting insights regarding the mediating
The manipulations worked as intended. We asked participants to role of irritation. In H7alt, H8alt, and H9alt, we proposed that if customers
indicate whether they encountered their host face-to-face once (Mone = perceive the frequency of face-to-face encounters initiated by their host
6.04, SD = 1.74; Mother = 2.14, SD = 2.03; t = 22.07, p <.01), twice negatively, irritation might arise as a negative affective response, which
(Mtwo = 5.38, SD = 2.27; Mother = 2.11, SD = 1.94; t = 12.46, p <.01), would increase their misbehavior concealment intentions. Although we
three times (Mthree = 5.70, SD = 2.12; Mother = 1.93, SD = 1.72; t = did not find any indirect effects of two (vs. one) face-to-face interactions
15.34, p <.01), or four times (Mfour = 4.98, SD = 2.51, Mother = 1.65, SD (b = 0.09, 95 % CI = [-0.02, 0.22]) mediated by irritation (H7alt), in line
= 1.43, t = 12.32, p <.01). Realism checks suggested that participants with H8alt, we found a positive indirect effect of three (vs. two) face-to-
could easily imagine the situation described in the scenarios (M = 6.12, face interactions on misbehavior concealment intentions mediated by an
SD = 1.27) and envision themselves therein (M = 6.03, SD = 1.33). increase in irritation (b = 0.15, 95 % CI = [0.01, 0.37]). Regarding the
predicted positive effect of four (vs. three) face-to-face interactions on
5.3. Measures and validity assessment guests’ misbehavior concealment intentions, we did not find empirical
evidence to support H9alt (b = -0.06, 95 % CI = [-0.24, 0.08]). Table 4
To replicate the measures used in Study 1, we used the same multi- lists the detailed results.
item scale for social identification. In addition, irritation was
measured as a mediator (Akaah et al., 1995; Van Diepen et al., 2009). 5.5. Robustness check and additional findings
Using the same single item from Study 1, we measured participants’
misbehavior concealment intentions and used the same multi-item scale To ascertain whether our model is robust to alternative operation­
to control for social desirability. All items were rated using seven-point alizations of misbehavior concealment intentions, we conducted a par­
Likert-type scales (1 = strongly disagree, 7 = strongly agree; see allel mediation analysis with social identification and irritation for each
Table 2). To check the robustness of our measurement of misbehavior of our pairwise comparisons using the binary outcome variable (yes, no)
concealment intentions, we included a binary (yes, no) outcome vari­ for whether they would conceal the damage. The results of the logistic
able: “Would you conceal the damage you caused in the apartment (i.e., regression were consistent with the findings of the single-item mea­
use the dark-colored bedding to cover up the stains on the mattresses), surement (see Table 5).

12
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

Table 4
Results of Study 2 (PROCESS Model 4)
Coefficient (SE) LLCI ULCI Hypothesis Test

1. Model A: Two (vs. one) face-to-face interactions

Indirect effects on misbehavior concealment intentions
Two (vs. one) face-to-face interactions → social identification → misbehavior concealment intentions -0.03 -0.16 0.08 H7: Rejected
(0.06)
Two (vs. one) face-to-face interactions → irritation → misbehavior concealment intentions 0.10 -0.02 0.22 H7 alt: Rejected
(0.06)

Effects on social identification

Two (vs. one) face-to-face interactions → social identification 0.10 -0.24 0.44
(0.17)
Effects on irritation
Two (vs. one) face-to-face interactions → irritation 0.35 -0.04 0.75
(0.20)
Effects on misbehavior concealment intentions
Two (vs. one) face-to-face interactions → misbehavior concealment intentions -0.49 -1.01 0.03
(0.27)
Social identification → misbehavior concealment intentions -0.31*** -0.50 -0.13
(0.09)
Irritation → misbehavior concealment intentions 0.27*** 0.12 0.43
(0.08)

Controls
Age → social identification 0.02* 0.01 0.03
(0.01)
Gender → social identification 0.13 -0.18 0.45
(0.16)
Social desirability → social identification 0.22*** 0.09 0.35
(0.06)
Number of overnight stays in non-hotel accommodations → social identification 0.01 -0.01 0.02
(0.01)
Number of platforms in use → social identification 0.08 -0.05 0.22
(0.07)
Age → irritation -0.01 -0.02 0.01
(0.01)
Gender → irritation 0.53** 0.16 0.90
(0.19)
Social desirability → irritation -0.27*** -0.41 -0.11
(0.08)
Number of overnight stays in non-hotel accommodations → irritation -0.01 -0.03 0.01
(0.01)
Number of platforms in use → irritation 0.23** 0.07 0.38
(0.08)
Age → misbehavior concealment intentions -0.01 -0.02 0.02
(0.01)
Gender → misbehavior concealment intentions -0.38 -0.87 0.11
(0.25)
Social desirability → misbehavior concealment intentions -0.01 -0.20 0.20
(0.10)
Number of overnight stays in non-hotel accommodations → misbehavior concealment intentions -0.02 -0.04 0.01
(0.01)
Number of platforms in use → misbehavior concealment intentions -0.09 -0.30 0.11
(0.10)
2. Model B: Three (vs. two) face-to-face interactions

Indirect effects on misbehavior concealment intentions

Three (vs. two) face-to-face interactions → social identification → misbehavior concealment intentions -0.07 -0.23 0.06 H8: Rejected
(0.07)
Three (vs. two) face-to-face interactions → irritation → misbehavior concealment intentions 0.15 0.01 0.37 H8 alt: Supported
(0.09)

Effects on social identification

Three (vs. two) face-to-face interactions → social identification 0.22 -0.19 0.62
(0.21)
Effects on irritation
Three (vs. two) face-to-face interactions → irritation 0.70** 0.19 1.21
(0.26)

Effects on misbehavior concealment intentions

Three (vs. two) face-to-face interactions → misbehavior concealment intentions -0.18 -0.78 0.42
(0.30)
Social identification → misbehavior concealment intentions -0.31** -0.54 -0.09
(0.11)
(continued on next page)

13
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

Table 4 (continued )
Coefficient (SE) LLCI ULCI Hypothesis Test

Irritation → misbehavior concealment intentions 0.21* 0.03 0.39

(0.09)

Controls
Age → social identification 0.02** 0.01 0.04
(0.01)
Gender → social identification 0.17 -0.21 0.55
(0.19)
Social desirability → social identification 0.19* 0.01 0.37
(0.09)
Number of overnight stays in non-hotel accommodations → social identification 0.01 -0.01 0.02
(0.01)
Number of platforms in use → social identification 0.17* 0.01 0.33
(0.08)
Age → irritation -0.01 -0.33 0.01
(0.01)
Gender → irritation 0.13 -0.34 0.61
(0.12)
Social desirability → irritation -0.32** -0.55 -0.09
(0.12)
Number of overnight stays in non-hotel accommodations → irritation -0.01 -0.03 0.01
(0.01)
Number of platforms in use → irritation 0.05 -0.15 0.25
(0.10)
Age → misbehavior concealment intentions -0.01 -0.04 0.01
(0.01)
Gender → misbehavior concealment intentions -0.44 -0.99 0.11
(0.28)
Social desirability → misbehavior concealment intentions -0.25 -0.51 0.02
(0.14)
Number of overnight stays in non-hotel accommodations → misbehavior concealment intentions 0.01 -0.01 0.03
(0.01)
Number of platforms in use → misbehavior concealment intentions 0.13 -0.11 0.36
(0.12)
3. Model C: Four (vs. three) face-to-face interactions

Indirect effects on misbehavior concealment intentions

Four (vs. three) face-to-face interactions → social identification → misbehavior concealment intentions -0.01 -0.10 0.10 H9: Rejected
(0.05)
Four (vs. three) face-to-face interactions → irritation → misbehavior concealment intentions -0.06 -0.24 0.08 H9 alt: Rejected
(0.08)

Effects on social identification

Four (vs. three) face-to-face interactions → social identification 0.01 -0.35 0.37
(0.18)

Effects on irritation
Four (vs. three) face-to-face interactions → irritation -0.21 -0.73 0.30
(0.26)

Effects on misbehavior concealment intentions

Four (vs. three) face-to-face interactions → misbehavior concealment intentions -0.24 -0.78 0.29
(0.27)
Social identification → misbehavior concealment intentions -0.24 -0.48 0.01
(0.12)
Irritation → misbehavior concealment intentions 0.28** 0.11 0.45
(0.08)

Controls
Age → social identification 0.02* 0.01 0.03
(0.01)
Gender → social identification -0.01 -0.39 0.37
(0.02)
Social desirability → social identification 0.34*** 0.17 0.51
(0.09)
Number of overnight stays in non-hotel accommodations → social identification -0.01 -0.02 0.01
(0.01)
Number of platforms in use → social identification 0.26** 0.09 0.43
(0.08)
Age → irritation -0.01 -0.03 0.01
(0.01)
Gender → irritation -0.26 -0.80 0.28
(continued on next page)

14
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

Table 4 (continued )
Coefficient (SE) LLCI ULCI Hypothesis Test

(0.27)
Social desirability → irritation -0.24 -0.15 0.33
(0.12)
Number of overnight stays in non-hotel accommodations → irritation -0.01 -0.03 0.01
(0.01)
Number of platforms in use → irritation 0.09 -0.15 0.33
(0.12)
Age → misbehavior concealment intentions -0.02 -0.04 0.01
(0.01)
Gender → misbehavior concealment intentions -0.10 -0.66 0.46
(0.01)
Social desirability → misbehavior concealment intentions -0.25 -0.07 0.45
(0.13)
Number of overnight stays in non-hotel accommodations → misbehavior concealment intentions 0.01 -0.01 0.03
(0.13)
Number of platforms in use → misbehavior concealment intentions 0.19 -0.07 0.44
(0.01)

Notes: Coefficients represent unstandardized estimates. Two-tailed test for hypothesized effects. SE = Standard Error; LLCI = lower limit confidence interval; ULCI =
upper limit confidence interval.
Model A: R2 Social Identification = 8.4 % / R2 Irritation = 10 % / R2 Misbehavior Concealment intentions = 11.6 %.
Model B: R2 Social Identification = 9.4 % / R2 Irritation = 11.7 % / R2 Misbehavior Concealment intentions = 16.9 %.
Model C: R2 Social Identification = 17.5 % / R2 Irritation = 4 % / R2 Misbehavior Concealment intentions = 20.6 %.
*p <.05. **p <.01. ***p <.001.

Moreover, to gain insights into the potential effects of the timing of a 6. General discussion
single face-to-face interaction, we conducted a parallel mediation
analysis of such interaction at check-in (vs. check-out) through social This study sought to explore the impact of face-to-face interactions
identification and irritation in PROCESS Model 4. We found a significant among hosts and guests in the P2P sharing economy service context on
indirect effect on misbehavior concealment intentions of a single face-to- guests’ intention to conceal their misbehavior, with a particular focus on
face interaction at check-in (vs. check-out), mediated by increased social the psychological mechanisms that might underlie this effect as well as
identification (b = -0.16, 95 % CI = [-0.37, -0.02]) but no significant potential contingencies. In two scenario-based experiments, we inves­
indirect effect through irritation (b = -0.10, 95 % CI = [-0.27, 0.05]). tigated the bright and dark side effects of host–guest face-to-face in­
The linear regression analysis revealed no direct effects (b = -0.06, SE = teractions in the context of P2P accommodation services. Social
0.31, p =.85, R2 = 0.03). Thus, as the findings in Table 6 indicate, the identification and irritation emerged as the psychological mechanisms
timing of a single host–guest face-to-face interaction is relevant. An through which interaction mode and frequency can affect guests’ in­
interaction at check-in rather than at check-out favors guests’ social tentions to conceal misbehavior. We also identified virtual community
identification, thereby reducing their intention to conceal their membership as a contingency of the relationship between the host–guest
misbehavior. interaction mode and guests’ social identification.

5.6. Discussion 6.1. Theoretical implications

Given the fact that there were no differences in social identification This study contributes to service research in four ways. First, we
due to frequent face-to-face interactions, such efforts appear to make no contribute to service encounter research, specifically regarding service
difference to guests’ misbehavior concealment intentions. Contrary to interactions, by stressing the relevance of interpersonal communication
our predictions in H7–H9, repetitive face-to-face interactions initiated by and the richness of non-verbal cues achieved through different
the host do not strengthen relational bonds or reduce guests’ intention to communication media (face-to-face and online-only) in predicting cus­
conceal their misbehavior. Thus, the data do not provide empirical ev­ tomers’ behavioral intentions. Specifically, in Study 1, we found that a
idence for the bright side effects of repeated face-to-face interactions. face-to-face interaction leads to an increase in social identification and,
However, our nuanced findings revealed that a single face-to-face thus, weaker misbehavior concealment intentions compared with an
interaction at check-in prompted greater social identification than a online-only interaction. We concur with studies that assess the effects of
face-to-face interaction during check-out. interaction modes (e.g., Luo & Zhang, 2016; Moon et al., 2019; Okdie
Additionally, we predicted that guests’ irritation levels would in­ et al., 2011) and affirm the benefits of face-to-face interactions. Simi­
crease in response to more frequent face-to-face interactions, leading to larly, we build on previous research on the outcomes of different forms
increased misbehavior concealment intentions. Contrary to our initial of customer–firm interaction in services that involve both human and
prediction, this negative effect was not incremental (Table 4). Guests technological elements (Makarem et al., 2009) and suggest that despite
who had two face-to-face interactions with their host did not express any the convenience technology offers, interpersonal interaction remains a
more irritation or misbehavior concealment intentions than those with critical factor for successful service delivery. In addition to comparing
one interaction, nor did those who experienced four rather than three the effects of face-to-face and online-only interactions on misbehavior
encounters. Instead, there appears to be a threshold for three in­ concealment intentions, in Study 2, we examine the outcomes of
teractions, such that once they have three face-to-face interactions, repeated face-to-face interactions and empirically show that too many
guests become more irritated and express greater intention to conceal personal interactions can lead to irritation instead of enhancing social
their misbehavior. In other words, the dark side effects of frequent face- identification, and thereby increase guests’ misbehavior concealment
to-face interactions emerge with the occurrence of three face-to-face intentions. Specifically, we find no significant indirect effect on misbe­
encounters. havior concealment intentions mediated by social identification when
comparing two vs. one, three vs. two, or four vs. three face-to-face in­
teractions, for which we do not observe a bright side effect of increasing

15
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

Table 5
Robustness check Study 2: Misbehavior concealment intentions as binary outcome variable (PROCESS Model 4)
Coefficient (SE) LLCI ULCI

1. Model A: Two (vs. one) face-to-face interactions

Indirect effects on misbehavior concealment intentions (binary outcome variable)
Two (vs. one) face-to-face interactions → social identification → misbehavior concealment intentions 0.03 -0.94 0.18
(0.07)
Two (vs. one) face-to-face interactions → irritation → misbehavior concealment intentions -0.13 -0.35 0.02
(0.09)

Effects on social identification

Two (vs. one) face-to-face interactions → social identification 0.10 -0.24 0.44
(0.17)

Effects on irritation
Two (vs. one) face-to-face interactions → irritation 0.35 -0.04 0.75
(0.20)

Effects on misbehavior concealment intentions (binary outcome variable)

Two (vs. one) face-to-face interactions → misbehavior concealment intentions 0.23 -0.40 0.87
(0.32)
Social identification → misbehavior concealment intentions 0.34** 0.12 0.56
(0.11)
Irritation → misbehavior concealment intentions -0.39*** -0.57 -0.20
(0.09)

Controls
Age → social identification 0.02* 0.01 0.03
(0.01)
Gender → social identification 0.13 -0.18 -0.45
(0.16)
Social desirability → social identification 0.22*** 0.09 0.35
(0.06)
Number of overnight stays in non-hotel accommodations → social identification 0.01 -0.01 0.02
(0.01)
Number of platforms in use → social identification 0.09 -0.04 0.22
(0.07)
Age → irritation -0.01 -0.02 0.01
(0.01)
Gender → irritation 0.53** 0.16 0.90
(0.19)
Social desirability → irritation -0.26*** -0.41 -0.11
(0.08)
Number of overnight stays in non-hotel accommodations → irritation -0.01 -0.03 0.01
(0.01)
Number of platforms in use → irritation 0.23** -0.07 0.38
(0.08)
Age → misbehavior concealment intentions -0.04*** 0.01 0.06
(0.01)
Gender → misbehavior concealment intentions -0.03 -0.63 0.57
(0.31)
Social desirability → misbehavior concealment intentions -0.11 -0.36 0.14
(0.13)
Number of overnight stays in non-hotel accommodations → misbehavior concealment intentions 0.01 -0.04 0.05
(0.02)
Number of platforms in use → misbehavior concealment intentions 0.21 -0.08 0.50
(0.15)
2. Model B: Three (vs. two) face-to-face interactions

Indirect effects on misbehavior concealment intentions (binary outcome variable)

Three (vs. two) face-to-face interactions → social identification → misbehavior concealment intentions 0.07 -0.08 0.29
(0.09)
Three (vs. two) face-to-face interactions → irritation → misbehavior concealment intentions -0.19 -0.51 -0.02
(0.13)

Effects on social identification

Three (vs. two) face-to-face interactions → social identification 0.21 -0.19 0.62
(0.20)

Effects on irritation
Three (vs. two) face-to-face interactions → irritation 0.70** 0.19 1.21
(0.26)

(continued on next page)

16
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

Table 5 (continued )
Coefficient (SE) LLCI ULCI

Effects on misbehavior concealment intentions (binary outcome variable)

Three (vs. two) face-to-face interactions → misbehavior concealment intentions 0.27 -0.53 1.06
(0.40)
Social identification → misbehavior concealment intentions 0.35* 0.06 0.65
(0.15)
Irritation → misbehavior concealment intentions -0.27* -0.50 -0.05
(0.11)

Controls
Age → social identification 0.02 -0.19 0.62
(0.21)
Gender → social identification 0.17** 0.01 0.04
(0.19)
Social desirability → social identification 0.19* 0.01 0.37
(0.09)
Number of overnight stays in non-hotel accommodations → social identification 0.01 -0.01 0.01
(0.01)
Number of platforms in use → social identification 0.17* 0.01 0.33
(0.08)
Age → irritation -0.01 -0.03 0.01
(0.01)
Gender → irritation 0.13 -0.34 0.61
(0.24)
Social desirability → irritation -0.32** -0.55 -0.09
(0.12)
Number of overnight stays in non-hotel accommodations → irritation -0.01 -0.03 0.01
(0.01)
Number of platforms in use → irritation 0.05 -0.15 0.25
(0.10)
Age → misbehavior concealment intentions 0.03 -0.01 0.06
(0.01)
Gender → misbehavior concealment intentions 0.18 -0.53 0.91
(0.37)
Social desirability → misbehavior concealment intentions 0.33 -0.02 0.67
(0.18)
Number of overnight stays in non-hotel accommodations → misbehavior concealment intentions -0.01 -0.04 0.01
(0.01)
Number of platforms in use → misbehavior concealment intentions -0.09 -0.40 0.22
(0.16)
3. Model C: Four (vs. three) face-to-face interactions

Indirect effects on misbehavior concealment intentions (binary outcome variable)

Four (vs. three) face-to-face interactions → social identification → misbehavior concealment intentions 0.01 -0.17 0.21
(0.09)
Four (vs. three) face-to-face interactions → irritation → misbehavior concealment intentions 0.04 -0.06 0.23
(0.07)

Effects on social identification

Four (vs. three) face-to-face interactions → social identification 0.01 -0.35 0.37
(0.18)

Effects on irritation
Four (vs. three) face-to-face interactions → irritation -0.21 -0.73 0.30
(0.26)

Effects on misbehavior concealment intentions (binary outcome variable)

Four (vs. three) face-to-face interactions → misbehavior concealment intentions 0.01 -0.74 0.76
(0.38)
Social identification → misbehavior concealment intentions 0.42* 0.08 0.76
(0.17)
Irritation → misbehavior concealment intentions -0.19 -0.43 0.04
(0.12)

Controls
Age → social identification 0.02* 0.01 0.03
(0.18)
Gender → social identification -0.01 -0.39 0.37
(0.19)
Social desirability → social identification -0.34*** 0.17 0.51
(0.09)
Number of overnight stays in non-hotel accommodations → social identification -0.01 -0.02 0.01
(0.01)
(continued on next page)

17
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

Table 5 (continued )
Coefficient (SE) LLCI ULCI

Number of platforms in use → social identification 0.26** 0.09 0.43

(0.08)
Age → irritation 0.01 -0.73 0.30
(0.01)
Gender → irritation -0.26 -0.80 0.28
(0.27)
Social desirability → irritation -0.24 -0.48 0.01
(0.12)
Number of overnight stays in non-hotel accommodations → irritation -0.01 -0.03 0.01
(0.01)
Number of platforms in use → irritation 0.09 -0.15 0.33
(0.12)
Age → misbehavior concealment intentions 0.03* 0.01 0.06
(0.01)
Gender → misbehavior concealment intentions 0.38 -0.39 1.15
(0.39)
Social desirability → misbehavior concealment intentions 0.34 -0.01 0.70
(0.18)
Number of overnight stays in non-hotel accommodations → misbehavior concealment intentions -0.01 -0.04 0.01
(0.01)
Number of platforms in use → misbehavior concealment intentions -0.30 -0.65 0.05
(0.18)

Notes: Coefficients represent unstandardized estimates. Two-tailed test for hypothesized effects. SE = Standard Error; LLCI = lower limit confidence interval; ULCI =
upper limit confidence interval.
Model A 1: R2 Social Identification = 8.4 % / R2 Irritation = 10 %.
Model B: R2 Social Identification = 9.4 % / R2 Irritation = 11.7 %.
Model C: R2 Social Identification = 17.5 % / R2 Irritation = 4.6 %.
*p <.05. **p <.01. ***p <.001.

the number of face-to-face encounters. However, we observe a dark side face-to-face interactions.
effect, such that three (vs. two) face-to-face interactions have a positive Fourth, our research offers a novel approach to investigating
indirect effect on misbehavior concealment intentions, mediated by an customer misbehavior by analyzing concealment intentions as a specific
increase in irritation. Interestingly, once sparked, irritation appeared to manifestation of customer misbehavior, which is prevalent in sharing
be stable rather than increasing linearly. Hence, we did not find signif­ economy settings and can impose an equivalent burden on firms and
icant effects on irritation and, subsequently, on misbehavior conceal­ service providers relative to other types of customer misbehavior (e.g.,
ment intentions when comparing four vs. three face-to-face interactions. property damage). Few empirical studies address this particular type of
Second, we theoretically argue and empirically distill the mediating misbehavior from an organizational perspective, and those that do,
roles of social identification and irritation as parallel psychological mainly refer to employee misbehavior concealment or service sabotage
mechanisms. Therefore, we extend prior knowledge on social identity (Harris & Ogbonna, 2002, 2010). Thus, we address an important
theory (Tajfel & Turner, 1986). Specifically, based on the empirical customer perspective that has thus far been underrepresented in the
findings of Study 1, we introduce the mode of interaction as an addi­ extant research.
tional factor to enhance group cohesion, elicit differential group alle­
giance levels, and foster norm compliance. Moreover, our results 6.2. Managerial implications
underscore the relevance of attributions in driving customer reactions.
We found in Study 2 that interaction frequency can elicit both favorable Our findings suggest two main managerial implications. First, in
and unfavorable affective and behavioral responses, with unfavorable view of the insights provided by Study 2, hosts should account for the
responses prevailing in our empirical setting. Concurring with previous contradictory psychological mechanisms that may arise from repeated
marketing and service research (e.g. Eggert, Steinhoff, & Garnefeld, face-to-face interactions with guests. Compared with online-only in­
2015; Mohr & Bitner, 1995), we stress the relevance of attribution teractions, face-to-face interactions elicit social identification, which
theory (Heider, 1958; Weiner, 1985) as a valuable approach for theo­ reduces guests’ intentions to conceal their misbehavior. This effect is
rizing the outcomes of encounters between customers and service particularly evident when the face-to-face interaction occur at check-in
providers. rather than at check-out, as shown in our exploration of the potential
Third, we extend previous work on customer misbehavior in the effects of the timing of a single face-to-face interaction. However, too
sharing economy; notably, we take the perspective of the service pro­ many face-to-face interactions can irritate guests without increasing
vider (Danatzis & Möller-Herm, 2023; Schaefers et al., 2016b, Srivas­ their social identification. In contrast to the extant literature (e.g.,
tava et al., 2022). Prior studies have mostly addressed customer Dagger et al., 2009), our findings suggest that repeated personal in­
misbehavior from a customer standpoint, such as by investigating per­ teractions initiated by a service provider can evoke negative emotions
sonality traits (Babin & Babin, 1996; Daunt & Harris, 2011) or situa­ rather than strengthen relational ties with customers and negatively
tional drivers (e.g., Daunt & Greer, 2015; Daunt & Harris, 2012) that affect their behavioral intentions. According to our data, irritation arises
may prompt such actions. However, service providers and platforms once three face-to-face encounters occur, suggesting a cut-off value of
may also influence customers’ behavioral intentions, and outlining such two interactions beyond which unfavorable affective and behavioral
effects is relevant for theory and practice because it sheds light on in­ responses occur. These findings may reflect overkill, such that the extent
struments within platforms and hosts’ control that they can use to pre­ of interaction feels intrusive to customers (Mende & Bolton, 2011;
vent customer misbehavior. Our study also addresses calls to investigate Mende, Bolton, & Bitner, 2013). Based on this negative perception of the
the potential dark side of the sharing economy (Buhalis et al., 2020; situation, customers may rationalize the act of concealing their misbe­
Eckhardt et al., 2019) and shows how customer misbehavior can be havior. When service providers put too much effort into building and
prevented through a social strategy (Fombelle et al., 2020), namely, strengthening relationships, their attempts may backfire and cause

18
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

Table 6
Additional findings from Study 2: Timing of a single face-to-face interaction (PROCESS Model 4)
Coefficient (SE) LLCI ULCI

Indirect effects on misbehavior concealment intentions (single-item measure)

Single interaction at check-in (vs. at check-out) → social identification → misbehavior concealment intentions -0.16 -0.37 -0.02
(0.09)
Single interaction at check-in (vs. at check-out) → irritation → misbehavior concealment intentions -0.10 -0.27 0.05
(0.08)

Effects on social identification

Single interaction at check-in (vs. at check-out) → social identification -0.47* 0.10 0.84
(0.18)
Effects on irritation
Single interaction at check-in (vs. at check-out) → irritation -0.31 -0.74 0.13
(0.22)

Effects on misbehavior concealment intentions (single-item measure)

Single interaction at check-in (vs. at check-out) → misbehavior concealment intentions 0.21 -0.39 0.81
(0.30)
Social identification → misbehavior concealment intentions -0.35** -0.58 -0.11
(0.12)
Irritation → misbehavior concealment intentions 0.32** 0.12 0.52
(0.10)

Controls
Age → social identification 0.01* 0.01 0.03
(0.19)
Gender → social identification -0.05 -0.43 0.34
(0.19)
Social desirability → social identification 0.31*** 0.16 0.46
(0.07)
Number of overnight stays in non-hotel accommodations → social identification 0.01 -0.01 0.02
(0.01)
Number of platforms in use → social identification 0.13 -0.03 0.29
(0.08)
Age → irritation -0.01 -0.02 0.01
(0.01)
Gender → irritation 0.63** 0.18 1.09
(0.23)
Social desirability → irritation -0.25** -0.43 -0.08
(0.09)
Number of overnight stays in non-hotel accommodations → irritation -0.01 -0.03 0.01
(0.01)
Number of platforms in use → irritation 0.37*** 0.17 0.56
(0.10)
Age → misbehavior concealment intentions 0.01 -0.01 0.03
(0.01)
Gender → misbehavior concealment intentions -0.19 -0.82 0.43
(0.32)
Social desirability → misbehavior concealment intentions 0.07 -0.18 0.32
(0.13)
Number of overnight stays in non-hotel accommodations → misbehavior concealment intentions -0.02 -0.05 0.01
(0.01)
Number of platforms in use → misbehavior concealment intentions -0.18 -0.45 0.09
(0.14)

Notes: Coefficients represent unstandardized estimates. Two-tailed test for hypothesized effects. SE = Standard Error; LLCI = lower limit confidence interval; ULCI =
upper limit confidence interval.
R2 Social Identification = 16.7 % / R2 Irritation = 14.4 % / R2 Misbehavior Concealment Intentions = 12.4 %.
*p <.05. **p <.01. ***p <.001.

adverse effects. Based on these findings, we recommend that hosts limit 7. Limitations and further research
themselves to initiating no more than two face-to-face interactions with
guests. We offer an initial examination of the effects of the interaction mode
Second, P2P platforms should establish virtual communities to on customer misbehavior concealment intentions in the sharing econ­
engage users as a natural mechanism to foster social identification and omy. The limitations of our study suggest four avenues for future
reduce guests’ intention to conceal their misbehavior. However, ac­ research.
cording to our moderation analysis results in Study 1, if guests are not First, further research should examine the effectiveness of other
members of such communities, face-to-face interactions can be partic­ platform-related instruments beyond virtual communities (e.g., deter­
ularly relevant, such that their impact on social identification is only rence mechanisms, two-sided review systems, and cleaning fees) in
significant for guests featuring rather weak-tie relationships with the regulating and preventing concealment. These approaches could deliver
platform (e.g., new users). interesting findings regarding other forms of customer misbehavior,
especially when customers are fully aware of the consequences of
misbehavior (e.g., fines, bad reviews, and platform suspensions) and

19
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

their options for dealing with them (e.g., insurance policies). Airbnb, for Funding
example, has acted on this issue, introducing its “AirCover for Hosts”
policy, which provides up to US $1 million in compensation for property This research did not receive any specific grant from funding
damage and offers liability insurance for guests during their stay agencies in the public, commercial, or not-for-profit sectors.
(Airbnb, 2022).
Second, we call for more research into the elements of service pro­ CRediT authorship contribution statement
vision that may enhance or attenuate customers’ misbehavior conceal­
ment intentions. In P2P accommodation services, these elements might Edna Ozuna: Writing – review & editing, Writing – original draft,
include welcome gifts, special arrangements for arrival and departure (e. Validation, Methodology, Investigation, Formal analysis, Conceptuali­
g., early check-in and late check-out), cleanliness of the accommodation, zation. Lena Steinhoff: Writing – review & editing, Writing – original
luxury of the furniture, or even the specific format of online interaction. draft, Supervision, Methodology, Conceptualization.
In our manipulation of online-only interactions, we considered
communication via e-mail exclusively. Further research should investi­
gate the outcomes of other forms of synchronous or asynchronous online Declaration of competing interest
communication such as video calls and instant messaging. In addition,
despite the internal validity of our scenario-based experiments, their The authors declare that they have no known competing financial
external validity may be limited. Therefore, we call for field studies that interests or personal relationships that could have appeared to influence
observe actual customer misbehavior and link them with interaction the work reported in this paper.
modes between hosts and guests.
Third, future studies should explore the role of customer-related References
aspects that might also influence misbehavior. In studying social iden­
tification, we do not consider the potential impact of host–guest simi­ Ahearne, M., Atefi, Y., Lam, S. K., & Pourmasoudi, M. (2022). The future of buyer–seller
interactions: A conceptual framework and research agenda. J Acad Mark Sci, 50,
larity in terms of shared values, culture, opinions, ethnic background, or
22–45. https://doi.org/10.1007/s11747-021-00803-0
political views. Montoya and Briggs (2013) provide an interesting Airbnb. (2022, May 11). Everything you need to know about AirCover for Hosts. Retrieved
approach to investigating how employee–customer shared ethnicity af­ July 18, 2022, from https://www.airbnb.com/resources/hosting-homes/a/ever
ything-you-need-to-know-about-aircover-for-hosts-469?_set_bev_on_new_domain
fects face-to-face service interactions according to social identity theory.
=1657997268_NGY5YjgzMmM3MmIz&locale=en.
Their findings suggest that shared ethnicity affects customer expecta­ Akaah, I. P., Korgaonkar, P. K., & Lund, D. (1995). Direct marketing attitudes. J Bus Res,
tions of the relational exchange and preferential treatment benefits 34(3), 211–219. https://doi.org/10.1016/0148-2963(94)00119-Y
(Montoya & Briggs, 2013). Other aspects of customer–service provider American Hotel and Lodging Association. (2017). Hosts with Multiple Units - A Key Driver
of Airbnb Growth. Retrieved October 10, 2022, from .
congruency can be addressed to determine whether similarity might Ashforth, B. E., & Anand, V. (2003). The normalization of corruption in organizations.
attenuate misbehavior in general. Moreover, in investigating irritation, Res Organ Behav, 25, 1–52. https://doi.org/10.1016/S0191-3085(03)25001-2
we do not account for the potential effects of customers’ unique rela­ Babin, L. J., & Babin, L. A. (1996). Effects of moral cognitions and consumer emotions on
shoplifting intentions. Psychol Mark, 13(8), 785–802. https://doi.org/10.1002/
tional orientations, which influence their responses to relational efforts, (SICI)1520-6793(199612)13:8<785::AID-MAR5>3.0.CO;2-B
because not all customers want to engage in a strong relationship with a Bandura, A. (1999). Moral disengagement in the preparation of inhumanities. Pers Soc
service provider (Mende & Bolton, 2011; Palmatier et al., 2008). That is, Psychol Rev, 3(3), 193–209. https://doi.org/10.1207/s15327957pspr0303_3
Bardhi, F., & Eckhardt, G. M. (2012). Access-based consumption: The case of car sharing.
some guests enjoy this possibility, especially considering the unique J Consum Res, 39(4), 881–898. https://doi.org/10.1086/666376
value proposition of Airbnb, which aims to “offer personal and unique Belanche, D., Casaló, L. V., Flavián, C., & Schepers, J. (2020). Robots or frontline
experiences for guests, provided by locals who know the area,” (Ting, employees? Exploring customers’ attributions of responsibility and stability after
service failure or success. J Serv Manag, 31(2), 267–289. https://doi.org/10.1108/
2018) while others do not. Additional studies should investigate the
JOSM-05-2019-0156
potential impact of relational orientation on social identification and Benoit, S., Baker, T. L., Bolton, R. N., Gruber, T., & Kandampully, J. (2017). A triadic
irritation. Furthermore, we determined that irritation arises in more framework for collaborative consumption (CC): Motives, activities and resources and
capabilities of actors. J Bus Res, 79, 219–227. https://doi.org/10.1016/j.
than two face-to-face encounters; however, previous research has sug­
jbusres.2017.05.004
gested that greater interaction frequency is a tool to enhance relational Bergkvist, L., & Rossiter, J. R. (2007). The predictive validity of multiple-item versus
strength (Dagger et al., 2009). These contrasting findings might reflect single-item measures of the same constructs. J Mark Res, 44(2), 175–184. https://
the industry context: three face-to-face interactions might seem irri­ doi.org/10.1509/jmkr.44.2.175
Bitner, M. J., Booms, B. H., & Tetreault, M. S. (1990). The service encounter: Diagnosing
tating for accommodation services, but the same number of encounters favorable and unfavorable incidents. J Mark, 54(1), 71–84. https://doi.org/
could be appealing in other service settings. By testing the effects of 10.2307/1252174
interaction frequency in other industries, future studies can assess the Bonner, J. M., Greenbaum, R. L., & Quade, M. J. (2017). Employee unethical behavior to
shame as an indicator of self-image threat and exemplification as a form of self-
applicability of our findings and the proposed threshold. image protection: The exacerbating role of supervisor bottom-line mentality. J Appl
Fourth, we employed attribution theory (Heider, 1958; Weiner, Psychol, 102(8), 1203–1221. https://doi.org/10.1037/apl0000222
1985) in our hypothesis development to predict the effects of increasing Bouas, K. S., & Arrow, H. (1996). The development of group identity in computer and
face-to-face groups with membership change. Computer Supported Cooperative Work
interaction frequency on misbehavior concealment intentions, but we (CSCW), 4, 153–178. https://doi.org/10.1007/BF00749745
did not explicitly measure positive and negative attributions in our ex­ Buhalis, D., Andreu, L., & Gnoth, J. (2020). The dark side of the sharing economy:
periments. We hypothesized that increasing the interaction frequency Balancing value co-creation and value co-destruction. Psychol Mark, 37(5), 689–704.
https://doi.org/10.1002/mar.21344
would prompt differential attributions, which, in turn, would lead to
Busch, C., Demary, V., Engels, B., Haucap, J., Kehder, C., Loebert, I., & Rusche, C. (2018).
affective and/or behavioral responses such as social identification and Sharing economy im wirtschaftsraum deutschland. [Sharing Economy in the Economic
irritation. Previous research has similarly applied attribution theory Area of Germany]. German Federal Ministry for Economic Affairs and Climate Action.
Carville, O. (2021, June 15). Airbnb is spending millions of dollars to make nightmares
(Heider, 1958; Weiner, 1985) to explore emotions and behaviors
go away. Bloomberg Businessweek. https://www.bloomberg.com/news/features
resulting from differential attributions to a given stimulus (e.g., Eggert, /2021-06-15/airbnb-spends-millions-making-nightmares-at-live-anywhere-rentals-
Steinhoff, & Garnefeld, 2015; Mohr & Bitner, 1995). Further research go-away.
should shed light on this issue by explicitly measuring customers’ pos­ Chan, K. W., & Li, S. Y. (2010). Understanding consumer-to-consumer interactions in
virtual communities: The salience of reciprocity. J Bus Res, 67(5), 1033–1040.
itive and negative attributions as antecedents of downstream psycho­ https://doi.org/10.1016/j.jbusres.2008.08.009
logical mechanisms such as social identification or irritation. Crosby, L. A., Evans, K. R., & Cowles, D. (1990). Relationship quality in services selling:
An interpersonal influence perspective. J Mark, 54(3), 68–81. https://doi.org/
10.2307/1251817
Dagger, T. S., Danaher, P. J., & Gibbs, B. J. (2009). How often versus how long: The
interplay of contact frequency and relationship duration in customer-reported

20
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

service relationship strength. J Serv Res, 11(4), 371–388. https://doi.org/10.1177/ Jörling, M., Böhm, R., & Paluch, S. (2019). Service robots: Drivers of perceived
1094670508331251 responsibility for service outcomes. J Serv Res, 22(4), 404–420. https://doi.org/
Danatzis, I., & Möller-Herm, J. (2023). Stopping the spread: How blame attributions 10.1177/1094670519842334
drive customer-to-customer misbehavior contagion and what frontline employees Kahai, S. S., & Cooper, R. B. (2003). Exploring the core concepts of media richness
can do to curb it. J Serv Res, 26(3), 459–475. https://doi.org/10.1177/ theory: The impact of cue multiplicity and feedback immediacy on decision quality.
10946705221150441 Journal of Management Information Systems, 20(1), 263–299. https://doi.org/
Daunt, K. L., & Greer, D. A. (2015). Unpacking the perceived opportunity to misbehave: 10.1080/07421222.2003.11045754
The influence of spatio-temporal and social dimensions on consumer misbehavior. Kiesler, S., Siegel, J., & McGuire, T. W. (1984). Social psychological aspects of computer-
Eur J Mark, 49(9/10), 1505–1526. https://doi.org/10.1108/EJM-01-2014-0061 mediated communication. Am Psychol, 39(10), 1123–1134. https://doi.org/
Daunt, K. L., & Harris, L. C. (2011). Customers acting badly: Evidence from the 10.1037/0003-066X.39.10.1123
hospitality industry. J Bus Res, 64(10), 1034–1042. https://doi.org/10.1016/j. Kim, T., Anik, L., & Cian, L. (2021). Feedback as a two-way street: When and why rating
jbusres.2010.10.010 consumers fails. Mark Lett, 32, 351–362. https://doi.org/10.1007/s11002-021-
Daunt, K. L., & Harris, L. C. (2012). Exploring the forms of dysfunctional customer 09570-y
behaviour: A study of differences in servicescape and customer disaffection with Kim, W. G., Lee, C., & Hiemstra, S. J. (2004). Effects of an online virtual community on
service. J Mark Manag, 28(1–2), 129–153. https://doi.org/10.1080/ customer loyalty and travel product purchases. Tour Manag, 25(3), 343–355. https://
0267257X.2011.619149 doi.org/10.1016/S0261-5177(03)00142-0
Doosje, B., Branscombe, N. R., Spears, R., & Manstead, A. S. R. (1998). Guilty by Kundro, T. G., & Nurmohamed, S. (2021). Understanding when and why cover-ups are
association: When one’s group has a negative history. J Pers Soc Psychol, 75(4), punished less severely. Acad Manag J, 64(3), 873–900. https://doi.org/10.5465/
872–886. https://doi.org/10.1037/0022-3514.75.4.872 amj.2018.1396
Doosje, B., Ellemers, N., & Spears, R. (1995). Perceived intragroup variability as a Leal, G. P. A., Hor-Meyll, L. F., & de Paula Pessôa, L. A. G. (2014). Influence of virtual
function of group status and identification. J Exp Soc Psychol, 31(5), 410–436. communities in purchasing decisions: The participants’ perspective. J Bus Res, 67(5),
https://doi.org/10.1006/jesp.1995.1018 882–890. https://doi.org/10.1016/j.jbusres.2013.07.007
Eckhardt, G. M., Houston, M. B., Jiang, B., Lamberton, C., Rindfleisch, A., & Zervas, G. Lloyd’s of London. (April 10, 2018). Top risks consumers are worried about when using
(2019). Marketing in the sharing economy. J Mark, 83(5), 5–27. https://doi.org/ sharing economy services worldwide in 2018. [Graph]. In Statista. Retrieved July 17,
10.1177/0022242919861929 2022, from https://www.statista.com/statistics/882489/risks-consumers-are-worr
Eggert, A., Steinhoff, L., & Garnefeld, I. (2015). Managing the bright and dark sides of ied-about-when-using-sharing-economy-services-worldwide/.
status endowment in hierarchical loyalty programs. J Serv Res, 18(2), 210–228. Luo, Q., & Zhang, H. (2016). Building interpersonal trust in a travel-related virtual
https://doi.org/10.1177/1094670514566797 community: A case study on a Guangzhou couchsurfing community. Tour Manag, 54,
Ert, E., Fleischer, A., & Magen, N. (2016). Trust and reputation in the sharing economy: 107–121. https://doi.org/10.1016/j.tourman.2015.10.003
The role of personal photos in Airbnb. Tour Manag, 55, 62–73. https://doi.org/ Makarem, S., Mudambi, S. M., & Podoshen, J. S. (2009). Satisfaction in technology-
10.1016/j.tourman.2016.01.013 enabled service encounters. J Serv Mark, 23(3), 134–144. https://doi.org/10.1108/
Farrell, A., Souchon, A., & Durden, G. (2001). Service encounter conceptualization: 08876040910955143
Employees’ service behaviors and customers’ service quality perceptions. J Mark Marinova, D., Singh, S. K., & Singh, J. (2018). Frontline problem-solving effectiveness: A
Manag, 17, 577–593. https://doi.org/10.1362/026725701323366944 dynamic analysis of verbal and nonverbal cues. J Mark Res, 55(2), 178–192. https://
Fisk, R., Grove, S., Harris, L. C., Keeffe, D. A., Daunt, K. L., Russell-Bennett, R., & Wirtz, J. doi.org/10.1509/jmr.15.0243
(2010). Customers behaving badly: A state of the art review, research agenda and Mazar, N., Amir, O., & Ariely, D. (2008). The dishonesty of honest people: A theory of
implications for practitioners. J Serv Mark, 24(6), 417–429. https://doi.org/ self-concept maintenance. J Mark Res, 45(6), 633–644. https://doi.org/10.1509/
10.1108/08876041011072537 jmkr.45.6.63
Fombelle, P. W., Voorhes, C. M., Jenkins, M. R., Sidaoui, K., Benoit, S., Gruber, T., McAlexander, J. H., Schouten, J. W., & Koenig, H. F. (2002). Building brand community.
Gustafsson, A., & Abosag, I. (2020). Customer deviance: A framework, prevention J Mark, 66, 38–54. https://doi.org/10.1509/jmkg.66.1.38.18451
strategies, and opportunities for future research. J Bus Res, 116, 387–400. https:// McLeay, F., Osburg, V. S., Yoganathan, V., & Patterson, A. (2021). Replaced by a Robot:
doi.org/10.1016/j.jbusres.2019.09.012 Service Implications in the Age of the Machine. J Serv Res, 24(1), 104–121. https://
Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with doi.org/10.1177/1094670520933354
unobservable variables and measurement error. J Mark Res, 18(1), 39–50. https:// Mehrabian, A. (2017). Nonverbal communication. London: Routledge.
doi.org/10.1177/002224378101800104 Mende, M., & Bolton, R. N. (2011). Why attachment security matters: How customers’
Fuchs, C., & Diamantopoulos, A. (2009). Using single-item measures for construct attachment styles influence their relationships with service firms and service
measurement in management research: Conceptual issues and application employees. J Serv Res, 14(3), 285–301. https://doi.org/10.1177/
guidelines. Die Betriebswirtschaft, 69(2), 195–210. 1094670511411173
Fullerton, R. A., & Punj, G. (2004). Repercussions of promoting an ideology of Mende, M., Bolton, R., & Bitner, M. J. (2013). Decoding customer–firm relationships:
consumption: Consumer misbehavior. J Bus Res, 57(11), 1239–1249. https://doi. How attachment styles help explain customers’ preferences for closeness, repurchase
org/10.1016/S0148-2963(02)00455-1 intentions, and changes in relationship breadth. J Mark Res, 50(1), 125–142. https://
Gabbott, M., & Hogg, G. (2010). The role of non-verbal communication in service doi.org/10.1509/jmr.10.0072
encounters: A conceptual framework. J Mark Manag, 1–2, 5–26. https://doi.org/ Meuter, M. L., Ostrom, A. L., Roundtree, R. I., & Bitner, M. J. (2000). Self service
10.1362/0267257012571401 technologies: Understanding customer satisfaction with technology-based service
Giebelhausen, M., Robinson, S. G., Sirianni, N. J., & Brady, M. K. (2014). Touch versus encounters. J Mark, 64(3), 50–64. https://doi.org/10.1509/jmkg.64.3.50.18024
tech: When technology functions as a barrier or a benefit to service encounters. Mills, M. K. (1981). Deviance and dissatisfaction: An exploratory study. Adv Consum Res,
J Mark, 78(4), 113–124. https://doi.org/10.1509/jm.13.0056 8(1), 682–686.
Godfrey, A., Seiders, K., & Voss, G. B. (2011). Enough is enough! The fine line in Moeller, S., & Wittkowski, K. (2010). The burdens of ownership: Reasons for preferring
executing multichannel relational communication. J Mark, 75, 94–109. https://doi. renting. Managing Service Quality: An International Journal, 20(2), 176–191. https://
org/10.1509/jmkg.75.4.94 doi.org/10.1108/09604521011027598
Gross, J., & Vostroknutov, A. (2022). Why do people follow social norms? Curr Opin Mohr, L. A., & Bitner, M. J. (1995). The role of employee effort in satisfaction with
Psychol, 4, 1–6. https://doi.org/10.1016/j.copsyc.2021.08.016 service transactions. J Bus Res, 32(3), 239–252. https://doi.org/10.1016/0148-2963
Guegan, J., Segonds, F., Barré, J., Maranzana, N., Mantelet, F., & Buisine, S. (2017). (94)00049-K
Social identity cues to improve creativity and identification in face-to-face and Montoya, D. Y., & Briggs, E. (2013). Shared ethnicity effects on service encounters: A
virtual groups. Comput Hum Behav, 77, 104–147. https://doi.org/10.1016/j. study across three U.S. subcultures. J Bus Res, 66(3), 314–320. https://doi.org/
chb.2017.08.043 10.1016/j.jbusres.2011.08.011
Harris, L. C., & Ogbonna, E. (2002). Exploring service sabotage: The antecedents, types Moon, H., Miao, L., Hanks, L., & Line, N. D. (2019). Peer-to-peer interactions:
and consequences of frontline, deviant, antiservice behaviors. J Serv Res, 4(3), Perspectives of Airbnb guests and hosts. Int J Hosp Manag, 77, 405–414. https://doi.
163–183. https://doi.org/10.1177/1094670502004003001 org/10.1016/j.ijhm.2018.08.004
Harris, L. C., & Ogbonna, E. (2010). Hiding customer complaints: Studying the Morimoto, M., & Chang, S. (2006). Consumers’ attitudes toward unsolicited commercial
motivations and forms of service employees’ complaint concealment behaviours. Br e-mail and postal direct mail marketing methods. Intrusiveness, perceived loss of
J Manag, 21(2), 262–279. https://doi.org/10.1111/j.1467-8551.2008.00617.x control and irritation. J Interact Advert, 7(1), 1–11. https://doi.org/10.1080/
Harris, L. C., & Reynolds, K. L. (2003). The consequences of dysfunctional customer 15252019.2006.10722121
behavior. J Serv Res, 6(2), 144–161. https://doi.org/10.1177/1094670503257044 Mosaad, M., Benoit, S., & Jayawardhena, C. (2023). The dark side of the sharing
Hart, C. M., Ritchie, T. D., Hepper, E. G., & Gebauer, J. E. (2015). The balanced inventory economy: A systematic literature review of externalities and their regulation. J Bus
of desirable responding short form (BIDR-16). SAGE Open, 5(4), 1–9. https://doi. Res, 168, Article 114186. https://doi.org/10.1016/j.jbusres.2023.114186
org/10.1177/2158244015621113 Okdie, B. M., Guadagno, R. E., Bernieri, F. J., Geers, A. L., & Mclarney-Vesotski, A. R.
Hayes, A. F. (2018). Introduction to mediation, moderation, and conditional process analysis (2011). Getting to know you: Face-to-face versus online interactions. Comput Hum
(2nd ed.). New York: The Guilford Press. Behav, 27(1), 153–159. https://doi.org/10.1016/j.chb.2010.07.017
Hayes, A. F. (2021). The PROCESS macro for SPSS, SAS, and R (Version 4.0). Retrieved Palmatier, R. W., Dant, R. P., Grewal, D., & Evans, K. R. (2006). Factors influencing the
March 14, 2022, from https://www.processmacro.org/index.html. effectiveness of relationship marketing: A meta-analysis. J Mark, 70, 136–153.
Heider, F. (1958). The psychology of interpersonal relations. New York: Wiley. https://doi.org/10.1509/jmkg.70.4.136
Hogg, M. A., & Reid, S. A. (2006). Social identity, self-categorization, and the Palmatier, R. W., Scheer, L. K., Evans, K. R., & Arnold, T. J. (2008). Achieving
communication of group norms. Commun Theory, 16(1), 7–30. https://doi.org/ relationship marketing effectiveness in business-to-business exchanges. J Acad Mark
10.1111/j.1468-2885.2006.00003.x Sci, 36, 174–190. https://doi.org/10.1007/s11747-007-0078-5

21
E. Ozuna and L. Steinhoff Journal of Business Research 177 (2024) 114582

Paulhus, D. L. (1991). Measurement and control of response bias. In J. P. Robinson, Uriely, N., Ram, Y., & Malach-Pines, A. (2011). Psychoanalytic sociology of deviant
P. R. Shaver, & L. S. Wrightsman (Eds.), Measures of personality and social tourist behavior. Ann Tour Res, 38(3), 1051–1069. https://doi.org/10.1016/j.
psychological attitudes (pp. 17–59). Academic Press. https://doi.org/10.1016/B978- annals.2011.01.014
0-12-590241-0.50006-X. Van Diepen, M., Donkers, B., & Franses, P. H. (2009). Does irritation induced by
Pelsmacker, P., & Van den Bergh, J. (1998). Advertising content and irritation: A study of charitable direct mailings reduce donations? Int J Res Mark, 26, 180–188. https://
226 TV commercials. J Int Consum Mark, 10(4), 5–27. https://doi.org/10.1300/ doi.org/10.1016/j.ijresmar.2009.03.007
J046v10n04_02 Varca, P. E. (2009). Emotional empathy and front line employees: Does it make sense to
Pera, R., Viglia, G., Grazzini, L., & Dalli, D. (2019). When empathy prevents negative care about the customer? J Serv Mark, 23(1), 51–56.
reviewing behavior. Ann Tour Res, 75, 265–278. https://doi.org/10.1016/j. Wan, L. C., Hui, M. K., & Qiu, Y. (2021). Tourist misbehavior: Psychological closeness to
annals.2019.01.005 fellow consumers and informal social control. Tour Manag, 83, Article 104258.
Perren, R., & Kozinets, R. V. (2018). Lateral exchange markets: How social platforms https://doi.org/10.1016/j.tourman.2020.104258
operate in a networked economy. J Mark, 82, 20–36. https://doi.org/10.1509/ Weiner, B. (1985). An attributional theory of achievement motivation and emotion.
jm.14.0250 Psychol Rev, 92(4), 548–573. https://doi.org/10.1037/0033-295X.92.4.548
Postmes, T., Haslam, S. A., & Jans, L. (2013). A single-item measure of social Welsh, D. T., Ordóñez, L. D., Snyder, D. G., & Christian, M. S. (2015). The slippery slope:
identification: Reliability, validity, and utility. Br J Soc Psychol, 52, 597–617. How small ethical transgressions pave the way for larger future transgressions. J Appl
https://doi.org/10.1111/bjso.12006 Psychol, 100(1), 114–127. https://doi.org/10.1037/a0036950
Reynolds, K. L., & Harris, L. C. (2009). Dysfunctional customer behavior severity: An Wickens, E. (1997). Licensed for thrill: Risk-taking and tourism. In S. Clift, & P. Gabowski
empirical examination. J Retail, 85, 321–335. https://doi.org/10.1016/j. (Eds.), Tourism and health: Risks, research and responses (pp. 151–164). London:
jretai.2009.05.005 Printer.
Schaefers, T., Lawson, S. J., & Kukar-Kinney, M. (2016a). How the burdens of ownership Wirtz, J., & Kum, D. (2004). Consumer cheating on service guarantees. J Acad Mark Sci,
promote consumer usage of access-based services. Mark Lett, 27(3), 567–577. 32(2), 159–175. https://doi.org/10.1177/0092070303261416
https://doi.org/10.1007/s11002-015-9366-x Wirtz, J., & McColl-Kennedy, J. R. (2010). Opportunistic customer claiming during
Schaefers, T., Wittkowski, K., Benoit, S., & Ferraro, R. (2016b). Contagious effects of service recovery. J Acad Mark Sci, 98, 654–675. https://doi.org/10.1007/s11747-
customer misbehavior in access-based services. J Serv Res, 19(1), 3–21. https://doi. 009-0177-6
org/10.1177/1094670515595047 Wirtz, J., So, K. K. F., Mody, M. A., Liu, S. Q., & Chun, H. H. (2019). Platforms in the
Shalvi, S., Handgraaf, M. J. J., & De Dreu, C. K. W. (2011). People avoid situations that peer-to-peer sharing economy. J Serv Manag, 30(4), 452–483. https://doi.org/
enable them to deceive others. J Exp Soc Psychol, 47(6), 1096–1106. https://doi.org/ 10.1108/JOSM-11-2018-0369
10.1016/j.jesp.2011.04.015 Wu, J.-J., Chen, Y.-H., & Chung, Y.-S. (2010). Trust factors influencing virtual
Sproull, L., & Kiesler, S. (1986). Reducing social context cues: Electronic mail in community members: A study of transaction communities. J Bus Res, 63(9–10),
organizational communication. Manag Sci, 32(11), 1942. https://doi.org/10.1287/ 1025–1032. https://doi.org/10.1016/j.jbusres.2009.03.022
mnsc.32.11.1492 Yang, S.-B., Lee, K., Lee, H., & Koo, C. (2019). In Airbnb we trust: Understanding
Srivastava, H. S., Jayasimha, K. R., & Sivakumar, K. (2022). Addressing customer consumers’ trust-attachment building mechanisms in the sharing economy. Int J
misbehavior contagion in access-based services. J Serv Mark, 36(6), 849–861. Hosp Manag, 83, 198–209. https://doi.org/10.1016/j.ijhm.2018.10.016
https://doi.org/10.1108/JSM-04-2021-0121 Ye, S., Ying, T., Zhou, L., & Wang, T. (2019). Enhancing customer trust in peer-to-peer
Steinhoff, L., Arli, D., Weaven, S., & Kozlenkova, I. V. (2019). Online relationship accommodation: A “soft” strategy via social presence. Int J Hosp Manag, 79, 1–10.
marketing. J Acad Mark Sci, 47, 369–393. https://doi.org/10.1007/s11747-018- https://doi.org/10.1016/j.ijhm.2018.11.017
0621-6 Zhao, X., Lynch, J. G., & Chen, Q. (2010). Reconsidering Barron and Kenny: Myths and
Suess, C., Woosnam, K., Mody, M., Dogru, T., & Turk, E. S. (2021). Understanding how truths about mediation analysis. J Consum Res, 37(2), 197–206. https://doi.org/
residents’ emotional solidarity with Airbnb visitors influences perceptions of their 10.1086/651257
impact on a community: The moderating role of prior experience staying at an
Airbnb. J Travel Res, 60(5), 1039–1060. https://doi.org/10.1177/
Edna Ozuna is Research Assistant at Paderborn University, Germany and Doctoral Student
0047287520921234
at the University of Rostock, Germany. She holds a bachelor’s degree in Business
Sundaram, D. S., & Webster, C. (2000). The role of nonverbal communication in service
Administration from the Piloto University, Colombia and a master’s degree in Tourism
encounters. J Serv Mark, 14(5), 378–391.
Development Strategies from the University of Applied Sciences Stralsund, Germany. Her
Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. In
research interests include customer behavior in the sharing economy, triadic relationships
S. Worchel, & W. G. Austin (Eds.), Psychology of intergroup relation (pp. 7–24).
and interactions in sharing platforms, and sustainability in tourism services.
Chicago: Hall Publishers.
Ting, D. (2018, April 23). Skift Forum Europe Preview: Airbnb’s Jeroen Merchiers on the
battle with booking. Skift. https://skift.com/2018/04/23/skift-forum-europe-previe Lena Steinhoff is Professor of Marketing and Digital Transformation at Paderborn Uni­
w-airbnbs-jeroen-merchiers-on-the-battle-with-booking/. versity, Germany. Her research focuses on relationship marketing and the implications of
Turner, L., & Ash, J. (1975). The Golden Hordes: International tourism and the pleasure digital transformation for customer relationships. Lena Steinhoff’s work has appeared in
periphery. London: Constable. the Journal of Marketing, Journal of the Academy of Marketing Science, Journal of Retailing,
Uriely, N., & Belhassen, Y. (2006). Drugs and risk-taking in tourism. Ann Tour Res, 33(2), Journal of Service Research, Journal of Business Research, Journal of International Marketing,
339–359. https://doi.org/10.1016/j.annals.2005.10.009 Journal of Service Management, and Industrial Marketing Management, among others. She has
authored the book Relationship Marketing in the Digital Age.

22