Symbolic management in the sharing economy:
the key to trust building
Josien Vos - 10356193 Draft version master thesis
Master Business Administration: Marketing Nicole Stofberg
Statement of originality
This document is written by Josien Vos who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Table of Contents
1. Introduction ... 7
2. Literature review and hypotheses ... 12
2.1 Peer-to-peer sharing ... 12
2.1.1 Defining the sharing economy ... 12
2.1.2 Barriers to sharing ... 14 2.2 Relational model theory ... 15 2.2.1 Economic exchange ... 16 2.2.2 Social exchange ... 17 2.2.3 Positioning ... 20 2.3 Perceived Connectedness ... 24 2.4 Trust ... 25 2.4.1 Peer-to-peer trust ... 26 2.4.2 Platform trust ... 29
2.4.3 Interaction between peer-to-peer and platform trust ... 31
2.5 Conceptual framework and hypotheses ... 32 3. Methodology ... 35 3.1 Platform Choice: SnappCar ... 35 3.2 Research Design ... 35 3.3 Stimuli Development (pre-test) ... 36 3.3.1 Pre-test 1... 37 3.3.2 Pre-test 2... 39 3.4 Survey Questionnaire and Measure Development ... 45 3.5 Data Collection Procedure ... 48 3.6 Sample ... 48 3.7 Data analysis ... 49 4. Results ... 51 4.1 Distribution... 51 4.2 Descriptive statistics ... 51 4.2.1 Manipulation check ... 51 4.2.2 Correlations ... 53 4.2.3 Control variables ... 55 4.3 Hypotheses testing... 55 4.3.1 Hypothesis 1 ... 57 4.3.2 Hypothesis 3 ... 57 4.3.3 Hypothesis 5 ... 58 4.3.4 Hypothesis 7 ... 59 5. Discussion ... 62 5.1 Summary of the results... 62 5.2 Discussion of the results ... 63
5.2.1 Direct relationship between platform positioning and intention to participate ... 63
5.2.2 Mediation effects ... 65
5.2.3 Sequential mediation effect of peer-to-peer trust and platform trust ... 67
5.2.4 Effects of Facebook ... 68
5.3 Theoretical implications ... 70
5.4 Managerial implications ... 72
5.5 Limitations and future research suggestions... 73
References ... 79 Appendices ... 92 Appendix A: Vignettes pre-test 1... 92 Appendix B: Survey Questionnaire Pre-tests ... 96 Appendix C: Copy of the survey questionnaire ... 98 Appendix D: Measurement Constructs... 105 Appendix E: Overview of dummies ... 106 Appendix F: Skewness and kurtosis ... 107
List of Tables
Table 1. Overview of the hypotheses………34
Table 2. Overview independent variables pre-test 1 ………..…38
Table 3. Correlations pre-test 1………..………39
Table 4. Correlations pre-test 2………..………40
Table 5. Overview final independent variables………...…40
Table 6. Correlation matrix manipulation check………...…..53
Table 7. Correlations matrix………...54
Table 8. Coefficients and significance levels mediation effect……….60
Table 9. Effect size and significance levels mediation effect………60
Table 10. Outcomes of the hypotheses………..………...61
List of Figures
Figure 1. Sharing economy and related forms of platform economies………....14Figure 2. Overview of theories sorted by monetary and relational value………19
Figure 3. Conceptual framework……….………..33
Figure 4. Vignette economic exchange without Facebook ………...41
Figure 5. Vignette economic exchange with Facebook………..…42
Figure 6. Vignette social exchange without Facebook………...43
Figure 7. Vignette social exchange with Facebook……….44
Abstract
Carsharing can lead to extensive environmental gains if practiced on a large scale. In order to convince car owners to share their cars, sharing platforms communicate utilitarian benefits. This is strengthened by academic literature emphasizing economic motives of participants in the sharing economy. However, building on relational model theory, this research argues that the positioning of a platform influences how potential participants frame peer-to-peer
relationships. Specifically, this study predicts that a utilitarian focus is argued to signal a ‘market pricing’ framing in which rational, self-serving behavior guided by opportunism is the norm. This will in turn negatively impact peer-to-peer trust, which is a principal
determinant of participation in sharing. In contrast, a social exchange positioning signals norms of ‘communality’ and in-group sharing fostering peer-to-peer trust and thus platform attractiveness. Moreover, whilst scholars to date have made a distinction between trust on the peer and platform level, little research is conducted on the links between the two. This
research argues that trust in peers can be transferred to trust in the platform, which is an essential condition for prospective participants to start sharing. An online vignette experiment (N=351) tests two ways in which peer-to-peer trust as well as platform trust can be constituted on a sharing platform: social versus economic positioning and the presence versus absence of mutual Facebook connections to strengthen social ties. The findings suggest that a positioning based on social exchange is associated with higher levels of peer-to-peer trust than based on economic exchange. This peer-to-peer trust positively impacts platform trust, which in turn results in a higher intention to participate. The manipulation regarding the mutual Facebook connections did not work correctly and is therefore eliminated.
Keywords: sharing economy, carsharing, relational model theory, symbolic management,
1. Introduction
New technologies enable strangers to connect in networks through online platforms and to share goods, services, transportation solutions, space, or money (Möhlmann 2015). In an age of overconsumption, individualism, and environmental issues, the establishment of such new sustainable ways of consumption is essential (Beck 2002; Sheth, Sethia & Srinivas 2011). Whilst the sustainability benefits of the sharing economy at large have been questioned, there is no doubt about the environmental gains of carsharing (Frenken & Schor 2017). However, for these benefits to materialize, the practice of carsharing should become mainstream, making this the focal topic of this research.
Whilst currently carsharing is still a niche activity, it is forecasted that this is changing due to urbanization, changed consumer perceptions about car ownership, environmental regulations (Rugman & Verbeeke 2000) and the increasing demand from consumers,
corporates and governments for sustainable solutions (Hart 1997). On a large scale, carsharing can have a substantial impact on the livability of cities and the environment. If personal cars are used 30% of the day instead of 4%, this would represent a reduction of 75% to 80% in the number of new cars that need to be produced (Bates & Leibling 2012; Hyek 2016). In
addition, people who participate in carsharing reduce their CO2 emissions with 240 to 390
kilograms per year (Nijland & Meerkerk 2017). With the introduction of the autonomous car, the sharing economy is even argued to result in a disruption of the car industry by 2030 (Weston 2017). It is forecasted that 95% of the miles travelled will take place in a shared car, a trend that is only expected to increase by 2050 (Gao, Kaas, Mohr and Wee 2016). To promote this trend, the Dutch government set, together with thirty commercial and non-commercial parties, a target of 100.000 shared cars on the road in 2017: the Green Deal (Rijksoverheid 2015). However, early 2017 there were 30.697 shared cars on the road, which suggests that the target has most likely not been met (CROW-KpVV 2017). The promise of
carsharing disrupting the automotive industry is big, however, the actual numbers tell a different story. For instance, whilst Cornet et al. (2012) found that 31% of the respondents want to start sharing within the next ten years, only 2.5% is participating already. The willingness to share exists, but carsharing platforms seem to have difficulty convincing car owners to actually share their car.
In an attempt to attract car owners and to make sharing appealing to them, most platforms stress the utilitarian benefits of sharing. For instance, carsharing platform SnappCar promotes cost savings that can easily reach € 1,000 a year. The social and environmental benefits are nowhere to be found in their marketing. This focus on utilitarian gains is
applauded by many academics (e.g. Bardhi & Eckhardt 2012; Hamari et al. 2016; Lamberton & Rose 2012). However, the research on carsharing carried out by these academics focuses on business-to-consumer (B2C) models. These are essentially ‘smart rental formulas’ in which convenience is the primary motivation to participate (Frenken and Schor 2017). Examples of such B2C practices are Zipcar in the US or Greenwheels and car2go in the Netherlands. Since users are renting cars owned by the platform, and in practice never meet other users, it makes sense that convenience and utility are key. Clearly, issues such as peer-to-peer trust and interpersonal experiences do not impede users’ willingness to participate. In contrast, peer-to-peer carsharing practices such as SnappCar, in which users rent cars directly from car owners, are between two peers and therefore highly interpersonal (Hamari et al. 2016). This means participants are not only using, but also providing cars. Whilst in theory anyone could be both a provider and a user, in practice people usually are in one of both groups (Schor 2014). Especially the car owners, providing strangers access to their cars, face the risk of
misbehavior and opportunistic behavior, which impedes their intention to participate (Hartl, Hofman & Kirchler 2015). Therefore, participation of car owners depends on their
well as of the platform itself (Kim, Yoon and Zo 2015). Whereas promoting functionality and cost-reduction might be effective in B2C sharing, in peer-to-peer sharing this strategy fails to overcome car owners’ fear of misbehavior and opportunism since it decreases trust in peers and in the platform rather than enlarging these. Therefore, it is argued that current marketing techniques, namely giving functional incentives, are ineffective in encouraging car owners to share. This study attempts to mitigate the barrier to peer-to-peer sharing – fear of opportunism and misbehavior – by investigating two ways to build peer-to-peer and platform trust, which are essential for car owners in order to share their cars.
Firstly, the positioning of the platform impacts car owners’ decisions to participate in sharing (Gorenflo 2012). Since participants are strangers, prospective participants can only deduce the other’s intentions from “the platform’s purpose and features, incentives and user experiences” (Stofberg 2016a). Drawing on relational model theory (Fiske 1991), this study asserts that the platform’s positioning affects how participants frame peer-to-peer
relationships. A positioning based on functionality and cost-reduction, as is often the case in B2C sharing, frames a market pricing (MP) relationship, with a focus on rational,
self-interested behavior. Such a relationship is argued to lower the peer-to-peer trust. In contrast, a positioning based on social and communal cues frames social exchange relationships, with a focus on communality and equality. Such relationships, in turn, heighten the peer-to-peer trust. To date, few studies have looked into the mechanisms in which perceptions of a social exchange relationship can be evoked. Bridoux and Stoelhorst highlight the need for empirical research on “symbolic management practices that help make a common identity salient” (2016, p. 246). This study responds to this need by emphasizing social cues in an attempt to trigger social exchange relationships to subsequently build peer-to-peer trust. In addition, this thesis asserts that emphasizing such social cues does not only build trust in the peer user, but also in the platform. Elements that build platform trust are social presence of peers and the
emphasis on the benevolent purposes of sharing on the platform (Kim, Yoon and Zo 2015). A social exchange positioning communicates these elements strongly.
A second way this study researches how to establish trust between peers and trust in the platform is through the presence of mutual Facebook friends on the platform. Whilst integrating social networking tools has not been tried in relation to the sharing economy, research in other fields provides evidence that the presence of mutual connections is an effective way to foster feelings of connectedness and a common identity, and to build trust (Moran 2005). Moreover, the presence of mutual Facebook friends is expected to highlight the perceived social presence and benevolent purposes of sharing, which increases the platform trust.
The focus on B2C sharing and utilitarian incentives in the sharing economy literature and the ineffectiveness of current marketing techniques to encourage car owners to share, puts forward the need for research that looks into symbolic management practices that foster both peer-to-peer and platform trust. This research jumps into this gap, attempting to overcome car owners’ barriers to sharing and thus increase their intention to participate. The research question that follows out of this is:
To what extent does increasing the focus on social aspects of a carsharing platform lead to higher car owners’ intention to participate?
This research question contains four underlying sub-questions:
1. What is the effect of positioning on car owners’ intention to participate?
2. What is the effect of the presence of mutual Facebook friends on car owners’ intention to participate?
3. What is the mediating role of peer-to-peer trust and platform trust;
a. on the relationship between positioning and car owners’ intention to participate?
b. on the relationship between the presence of mutual Facebook friends and car owners’ intention to participate?
4. What is the sequential mediating role of peer-to-peer trust and platform trust; a. on the relationship between positioning and car owners’ intention to
participate?
b. on the relationship between the presence of mutual Facebook friends and car owners’ intention to participate?
These questions are answered using the largest Dutch peer-to-peer carsharing platform: SnappCar. Four vignettes of the landing page for car owners are created and tested in a 2 (social vs. economic exchange) x 2 (presence vs. absence of Facebook) design. Their effects on peer-to-peer trust, platform trust and the intention to participate are analyzed (N=351).
The findings contribute to academic literature in several ways. Relational model theory (Fiske 1991) is successfully translated into vignettes in a sharing economy context, providing insights for research on brand positioning, the sharing economy and relational model theory itself. Furthermore, this research demonstrates the need for both peer-to-peer and platform trust in the sharing economy, adding to literature on online reputation and trust. Managers of sharing platforms can find insights on positioning strategies and designing for both types of trust. On a large scale, if platforms succeed in convincing car owners to participate in carsharing, large environmental gains can be realized.
This thesis will be structured as follows: the literature review provides a theoretical background, resulting in the conceptual model and corresponding hypotheses (Chapter 2). This is followed by the research methods (chapter 3) and findings (chapter 4). The discussion debates the findings and discusses the academic and managerial contributions, limitations and recommendations for future research (chapter 5). This thesis ends with the conclusion
2. Literature review and hypotheses
The literature review commences with defining different types of platforms within the sharing economy, followed by the barriers to sharing (paragraph 2.1). Two ways to overcome these barriers are introduced. Firstly, drawing on Fiske’s relational model theory (1991), symbolic management practices that frame specific relationships are discussed (paragraph 2.2). Secondly, the concept of perceived connectedness is outlined (paragraph 2.3). Lastly, the effects of peer-to-peer and platform trust are reviewed (paragraph 2.4).
2.1 Peer-to-peer sharing
2.1.1 Defining the sharing economy
Many misconceptions exist about what services and products are exactly considered to be part of the sharing economy. To clarify, Frenken and Schor (2017, p. 2-3) describe four
characteristics of the sharing economy. First of all, temporary access is given and sharing is therefore not a transfer of ownership. Subsequently, sharing takes place between consumers; peer-to-peer. Thirdly, it involves physical goods. Lastly, shared objects should be
underutilized. These characteristics combined lead to the following definition of sharing: “consumers granting each other temporary access to under-utilized physical assets (‘idle capacity’), possibly for money” (Frenken & Schor 2017, p. 2-3). Along the lines of these characteristics, a clear distinction can be made between the sharing economy and closely related so-called platform economies. Figure 1 clarifies how these platforms differ from each other.
Figure 1.
Sharing economy and related forms of platform economies
Source: Frenken & Schor 2017
These distinctions are important to avoid misconceptions. Researchers have used different definitions of the sharing economy, which has led to ambiguity in the sharing economy literature. By way of illustration, Bardhi and Eckhardt found in their study on B2C carsharing company Zipcar that consumers are primarily guided by a “self-serving and utilitarian
motivation and negative reciprocity toward the accessed object, firm and other consumers” (2012, p. 895). A large body of literature builds on this assumption that people behave out of self-interest and are looking for cost reduction (e.g. Moeller and Wittkowski 2010; Lamberton and Rose 2012). As a consequence, many carsharing platforms such as MyWheels and
SnappCar emphasize cost and convenience benefits. However, following the definition of Frenken and Schor (2017), Zipcar is not part of the sharing but of the product-service economy since it is not a peer-to-peer platform. Therefore, Bardhi & Eckhardt’s (2012) statement that users behave out of self-interest and have solely utilitarian motives is not automatically generalizable and applicable to sharing platforms. The difference between a
B2C and a peer-to-peer sharing platform (C2C in Figure 1) is that in the latter users do not only deal with the platform, but also with unknown others users. Moreover, users provide their own car, which entails risks. The main barrier that platforms need to overcome is the fear that peers will misbehave (Kim, Yoon & Zo 2015; Schaefers et al. 2016). The next paragraph elaborates on this.
2.1.2 Barriers to sharing
Generally, buyer-seller relationships are characterized by information asymmetry (Mishra et al. 1988). The uncertainty creates risks and fears of opportunistic behavior (Akerlof 1970), defined as ‘self-interest seeking with guile’ (Williamson 1975). This leads to mistrust (Jarvenpaa et al. 2000) and fear of misbehavior (Schaefers et al. 2016). Fullerton and Punj (2004) define misbehavior as deliberately violating generally accepted norms of conduct in consumption situations. When consumers have access to goods that are not their own property, the risk of misbehavior is heightened (Durgee and O’Connor 1995). This explains why platforms have specifically problems with attracting car providers. Online environments impose additional risks, since neither the identity of the seller nor the characteristics of the product can be fully assessed (Lee 1998). Since we have all been raised with the so-called ‘stranger danger bias’, meaning that stranger equals danger, sharing with strangers is often perceived as a large risk (Gebbia 2016). These risk perceptions eventually lead to a reduction in car owners’ intentions to share (Kim, Yoon & Zo 2015).
In order to overcome the fear of opportunistic behavior and misbehavior, platforms should create trust between peers and trust in the platform (Botsman 2016; Mayer, Davis & Schoorman 1995). Firstly, users have to trust other users and be confident that they will not misbehave (see section 2.4.1). This peer-to-peer trust is “generated and expressed in relations with peers in the community” (Grimsley, Meehan & Tan 2004). Secondly, users have to trust
that the platform will help in case a peer misbehaves (Botsman 2016) (see section 2.4.2). This platform trust is “generated and expressed in relations with community institutions”
(Grimsley, Meehan & Tan 2004). The higher the initial perception of the risk is, the higher the both types of trust need to be for a participant to enter into a transaction (Jarvenpaa et al. 2000; Sako & Helper 1998). Trust in both the platform and the peers is therefore essential for people in order to participate in sharing (Botsman 2016).
Given the peer-to-peer nature of the sharing economy, building interpersonal trust is key since sharing is riskiest for car owners, whereas this is less relevant in related forms of platform economies (Figure 1) such as Zipcar. Within platform economies, individual users do not grant temporary access to their personal belongings, meaning the risk of misbehavior and opportunism of a peer does not play a role and does not need to be overcome. Therefore, the approach that works very well to foster consumer adoption in a B2C context (such as when growing Zipcar or Greenwheels) will hinder growth in a peer-to-peer carsharing setting. We argue that the utilitarian and economic incentives Bardhi & Eckhardt (2012) emphasize do not build peer-to-peer trust, but reduce it. This assumption is based on Fiske’s (1991) relational model theory, which describes that relationships between people can take four forms. This study asserts that the positioning of a platform influences with which relational model participants frame peer-to-peer relationships. The following section examines the types of relational models, after which the way positioning influences the adoption of specific relational models is discussed.
2.2 Relational model theory
Relational model theory provides a framework enabling to look beyond solely economic motives in human behavior. Fiske (1991) acknowledges the relevance and existence of self-interested behavior (MP), however, he adds three other relational models. These models are
ways in which humans mentally frame their relationships with others and “internalize
relationships as part of their cognitive functioning and translate them into behavior” (Bridoux and Stoelhorst 2016, p. 230). In this research, three of the four models are used: communal sharing, equality matching and market pricing.1
Communal sharing is a relation of unity, community, undifferentiated collective identity, and kindness, typically enacted among close kin. (…) Equality matching is a one-to-one correspondence relationship in which people are distinct but equal, as manifested in balanced reciprocity (or tit-for- tat; revenge), equal share distributions or identical contributions, in-kind replacement compensation, and turn taking. Market pricing is based on an (intermodal) metric of value by which people compare different commodities and calculate exchange and cost/benefit ratios. (Fiske 1991, p. ix)
Based on Heyman & Ariely (2004), the relational models are divided into two general categories based on social exchange and economic exchange. The social exchange category includes CS and EM relationships and economic exchange the MP relationship. In the next sections, economic exchange and social exchange are elaborated on in the given order.
2.2.1 Economic exchange
A MP or economic exchange relationship is based on people who act out of self-interest and is rather impersonal. In transactions based on an economic exchange relationship, each strives to maximize utility at the other’s expense (Fiske 1991). Much well-established research on human behavior considers self-interest to be a key motivator of behavior in general, even when collaborating with others would be more beneficial than pursuing personal interests. Studies such as “The Logic of Collective Action” (Olson, 1965), the “prisoner’s dilemma” (Rapoport and Chammah, 1970), and the “tragedy of the commons” (Hardin, 1968), underline
1 Authority ranking is the fourth model, which is excluded in this research since it focuses on
asymmetry within relationships. This is not particularly relevant in a sharing context, since hierarchies and statuses do not play a substantial role in peer-to-peer sharing.
this point. According to these models, actions from human beings are based on rational reasoning, focusing on maximization of utility and cost reduction, or minimization of transaction costs (Möhlman 2015, p. 194). Such relationships are characterized by negative reciprocity and are referred to as the unsociable extreme (Sahlins 2017).
Within the sharing economy context, Belk (2010) calls this self-interest-based sharing ‘sharing out’. Sharing out focuses on pragmatic and economic interests and “preserves the self/other boundary” (Belk 2010, p. 726). Therefore, an economic exchange relationship only exists to support the transaction, not for the sake of the relationship itself, as is the case in CS relationships.
Two peers in an economic exchange relationship know the other person is trying to maximize his value and minimize his costs i.e. acting out of self-interest. The other is not only expected to act out of self-interest, this is even considered the “appropriate behavioral
response” in an economic exchange relationship (Stofberg 2017, p. 3). Such an expectation of the other person’s underlying motivations does not help in overcoming the fear of
opportunism and misbehavior. To contrary, an economic exchange relationship reduces the peer-to-peer trust (Bridoux & Stoelhorst 2016; Stofberg 2017). Since peer-to-peer trust is essential within a sharing transaction (e.g. Botsman & Rogers 2010), such a relationship is not desirable.
2.2.2 Social exchange
The social exchange category consists of both CS and EM relationships. Firstly, the CS relationship is looked into in more detail, after which the midpoint EM is discussed.
A CS relationship is based on generalized reciprocity. This is a one-way flow in which none of the actors keep up with what is received and what is given (Stofberg 2016b). In contrast to negative reciprocity in economic exchange relationships, in which actors have
opposed interests, a social exchange transaction is altruistic. Somebody gives without the expectation of necessarily getting anything in return (Sahlins 2017). Sahlins (2017) calls this the ‘solidary extreme’.
Translated to the sharing economy, Belk (2010, p. 726) refers to this type of
relationship as ‘sharing in’. This is an expression of community and “expands the sphere of extended self by expanding the domain of common property” (Belk 2010, p. 726). Although sharing in typically takes place between people who have a close relationship, Belk stresses “it can also include unseen online others who are part of a consumption community” (2014, p. 16). Instead of monetary value in market pricing, relational value is the central concept in a CS relationship. Relational value is what distinguishes the sharing economy from ‘traditional’ economies (Stofberg 2017) or as co-founder of AirBnB Gebbia states: “The connection behind the transaction is exactly what the sharing economy is aiming for. (…) The sharing economy is commerce with the promise of human connection” (2016, n.p.). Peers in a CS relationship put common interests before their own and “trust that others act in a similarly selfless manner” (Stofberg 2017, p. 3). Therefore, when sharing is considered to be a
communal act with a high focus on the interests of the group rather than one’s own, sharing is associated with high peer-to-peer trust levels (Belk 2010).
Between the two extremes MP and CS lies the midpoint EM, which falls in the social exchange category. A transaction based on an EM relationship is perceived as a precise balance in what is given and what is received, without delay. Sahlins (2017, p. 176) refers to this as ‘tit-for-tat’ or balanced reciprocity. Although the connection between two peers in an EM relationship is not as deep as in a CS relationship, it is a strong interdependent
relationship (Stofberg 2017). As a consequence, people expect the other to respect them and their property.
trust between peers is low, since both parties behave out of self-interest. In CS as well as EM relationships participants expect the others’ motivations to transcend self-interest (Stofberg 2017), which leads to high levels of peer-to-peer trust. What matters is the intention peers ascribe to others. Hence, “decisions and practices that would normally trigger stakeholders to frame their relationships as CS will not have this effect if stakeholders interpret them as being driven by a motive associated with another model” (Bridoux & Stoelhorst 2016, p. 246).
Figure 2 illustrates how the relational models relate to each other in terms of relational (Stofberg 2016b) and monetary value. Sahlins (2017) stresses that the different forms of reciprocity are a continuum and the mentioned concepts are the extremes of this spectrum.
Figure 2.
Overview of theories sorted by monetary and relational value
Source: Josien Vos, 2018.
Given the impact of the relational models on peer-to-peer trust and users’ behavior towards each other, platforms would be likely to prefer a social exchange relationship between their
users. The next paragraph examines how platforms can influence through their positioning which relational model users adopt.
2.2.3 Positioning
Various scholars agree that platforms can create an online environment in which peer-to-peer trust can thrive by means of the positioning and design of the platform (e.g. Botsman and Rogers 2010; Keymolen 2013). According to Gebbia (2016, n.p.), design is able to “overcome our most deeply rooted stranger danger bias”. Since this is essential in peer-to-peer sharing, platforms could attempt to prime social exchange relationships rather than economic exchange through symbolic management. One form of symbolic management is framing, which is a cognitive sensemaking process with the goal “to affect interpretations of events among various audiences” (Fiss & Zajac 2006, p. 1174). A frame is a “schemata of
interpretation” enabling people “to locate, perceive, identify, and label” something within their life and the world (Goffman 1974, p. 21). This study uses this principle of framing in order to shape respondents’ mental representations regarding their relationship with the other by manipulating the landing page of SnappCar (Bridoux & Stoelhorst 2016). Fiske (1991) argues that all consumers can adopt any of the four models as long as situational cues are strong enough. A platform’s positioning can be such a situational cue (Bridoux and Stoelhorst 2016). Since users do not know the other users, they can only deduct the other’s intentions and motivations from the platform itself. “The context a service sets influences how users feel about each other, the assets shared, and the service itself. The purpose, culture, user
experience, and other features shape user attitudes and behavior.” (Gorenflo 2012, n.p.) The next two sections elaborate on which situational cues frame either social or economic exchange relationships, drawing on recommendations on symbolic management practices by Habibi et al. (2017).
2.2.3.1 Framing an economic exchange relationship
A sharing platform characterized by commercial exchanges is likely to trigger an economic exchange relationship between peers. Habibi et al. (2017) argue such platforms focus on balanced transactions, efficiency, utilitarian benefits and money, for instance by emphasizing the value participants receive. Such monetary incentives cause prospective participants to not concentrate on the social and communal part of sharing, but on the economic part (Tenbrunsel & Messick 1999). Therefore, sharing in an MP relationship is not considered to be about social connections or helping other people, but rather about “dividing a resource among separate entities” (Belk 2010, p. 727). This turns the sharing exchange into a ‘cost benefit’ analysis instead of an act based on a feeling of we-ness. In addition to communicating
monetary incentives, minimization of social incentives and peer-to-peer social interaction are furthermore argued to trigger economic exchange relationships (Habibi et al. 2017). Such an anonymized platform furthermore stimulates misbehavior (Schaefers et al. 2016). An
economic exchange mindset is therefore believed to discourage responsible citizenship (Stofberg 2016b).
Economic exchange framing can be illustrated by the example of peer-to-peer sharing platform Couchsurfing (Tan 2010). The new CEO, Tony Espinoza, monetized a part of the platform, causing a shift in the focus from experience and community to costs. Instead of a community based on social values, the platform became a way to book a cheap holiday, attracting a different group of people with unsatisfied ‘traditional users’ as a result (Mettavant 2014). This shows how monetization “effectively moves the transaction out of the realm of the social and into the realm of the business” (Belk 2014, p. 12), i.e. moving from a social exchange to an economic exchange positioning. This change led to “self-selection of
participants because it sends out situational cues about the nature of the relationships among participants on the platform” (Stofberg 2016b, p. 8). Hence, a change in the situational cues
on the platform is able of changing the type of relationship that is adopted by users.
2.2.3.2 Framing a social exchange relationship
Although most sharing platforms offer a financial compensation, such as SnappCar, these platforms can still trigger a social exchange relationship by emphasizing the social and communal elements of sharing (Butcher et al. 2016). This could be accomplished by using words that emphasize peers’ common identity like “we” and “us” instead of “you” and “I” (Bridoux & Stoelhorst 2016). Additionally, phrases such as “we are a family” are argued to trigger a social exchange relationship (Ashforth & Johnson 2001). By doing so, platforms can decrease the level of anonymity and build a strong community, which prevents misbehavior and enlarges peer-to-peer trust and loyalty (Habibi et al. 2017; Schaefers et al. 2016). Moreover, symbolic management revolves around stressing the experiential benefits in contrast to the utilitarian ones (Habibi et al. 2017). Experiential consumption is researched to be “inherently more social than material consumption, resulting in greater affective value” (Habibi et al. 2017, p. 118). Sharing is associated with positive emotions, since people feel like they are doing a kind deed (Hellwig, Morhart, Girardin & Hauser 2015). Therefore, peer-to-peer sharing should be framed as an experience with large social factor (Habibi et al. 2017). Furthermore, participants of sharing practices are known for attaching value to
sustainable consumption (Bardhi & Eckhardt 2012) so communicating environmental benefits is argued to prime a CS relationship (Habibi et al. 2017). Lastly, references to money and calculations on what is given and received should be avoided (Habibi et al. 2017).
The second relational model within the social exchange category – EM – can be evoked by a sharing platform that aims at the development and maintenance of reciprocity-based relationships. This could be fostered by stressing equality between participants, for instance by giving them an equal vote in decisions (Bridoux & Stoelhorst 2016). In framing a
EM relationship, “feelings of social obligation are emphasized and terms like ‘partners’ or ‘friends’ are used to mark reciprocity and equality among the participants” (Stofberg 2016b, p. 14). CS and EM are combined into a social exchange relationship, which can be framed by focusing on interpersonal relationships, equality, and experiential and environmental benefits.
An example of a platform fostering social exchange relationships is peer-to-peer accommodation sharing platform AirBnB. From the beginning, AirBnB designed the platform with the objective to create interpersonal trust by focusing on social interactions between users (Gebbia 2016). Within nine years, AirBnB grew into a company that outperforms the hotel industry on all experience dimensions, including serendipity, localness, community and personalization (Mody et al. 2017). Whilst money can be earned through AirBnB, the
communication of the platform does not emphasize the financial gains, resulting in a successful platform on which users trust each other enough to share their homes.
In conclusion, via symbolic management platforms can frame social exchange relationships between people rather than economic exchange relationships (Bridoux & Stoelhorst 2016). This creates peer-to-peer trust (Bridoux & Stoelhorst 2016; Stofberg 2017) and subsequently helps to overcome the main barrier to sharing, which is the fear of
opportunism and misbehavior (Durgee & O’Connor 1995; Schaefers et al. 2016). Therefore, we argue that fostering a social exchange relationship between peers stimulates sharing rather than an economic exchange relationship. Thus:
H1. Platform positioning based on social exchange is associated with higher scores of intention to participate than platform positioning based on economic exchange.
The next section discusses a second way to build peer-to-peer trust and to stimulate car owners to participate in sharing.
2.3 Perceived Connectedness
The second independent variable tested in this study is the effect of perceived connectedness between peers on car owners’ intentions to participate. Perceived connectedness is
manipulated by the presence or absence of four mutual Facebook connections, called ‘friends’ on Facebook. These connections are part of an individual’s social capital, which consists of both online and offline connections. The concept originally referred to social ties in urban communities (Jacobs 1961), which was the basis for developing trust, collective actions and cooperation between members of these communities. With the rise of social media networks, the concept of social capital is often linked to those networks. Putnam (2000) distinguishes two forms of social capital. The first type is bonding social capital, referring to strong ties binding close friends and family. The second is bridging social capital, which are the weak ties, including acquaintances, friends of friends and work contacts. Social media are proved to increase and maintain social capital in general, but specifically to increase the weak ties in one’s network (Donath & Boyd 2004). Social networks of individuals often overlap,
especially within a certain geographical area as is the case with carsharing, since connections are primarily formed on the basis of offline social ties (Ellison, Steinfield & Lampe 2007). The power of social media networks is exposing the nodes where two individual social networks overlap and stressing the mutual connections two individuals have. Considering people have mostly weak ties in their social media network, two people would not have known about these mutual connections without social media (Donath & Boyd 2004). This is beneficial in the sharing economy context, since the exposure of mutual connections stresses similarities between two strangers. Since people tend to feel closer to someone similar to them than to someone who is socially distant, this leads to higher levels of peer-to-peer trust (Lewis & Weigert 1985; Fiedler, Haruvy & Li 2011). Paragraph 2.4.1 examines further how mutual Facebook friends can result in a higher trust among strangers.
Building on the principle of social proof (Cialdini 1987), this study argues that the perceived connectedness between two strangers directly effects respondents’ intentions to share. The principle of social proof asserts that someone is more likely to participate in
something if he sees that other people do it (Cialdini 1987). The more this person can relate to the other, the higher the probability he is influenced by the actions of the other. This peer power is so strong, since the probability that someone responds to influence tactics that are applied horizontally is higher than when these are applied vertically (Cialdini 1987). Two strangers with mutual friends are therefore expected to relate more to each other and as a result reinforce each other’s sharing behavior. One may hypothesize:
H2. Presence of mutual Facebook friends is associated with higher scores of intention to
participate than absence of mutual Facebook friends.
Previous paragraphs already touched upon the role of both peer-to-peer and platform trust. The next section dives deeper into these mediators.
2.4 Trust
Ensuring trust is essential for the sharing economy to function as sharing goods with strangers involves risk, even when no monetary transactions are involved (Ert, Fleischer and Magen 2016). Generally, trust can be defined as “committing to an exchange before you know how the other person will reciprocate” (Coleman 1990, chapter 5). As previously mentioned, in virtual communities two types of trust can be categorized: peer-to-peer and platform trust (Grimsley, Meehan & Tan 2004). As one might expect, trust is higher among socially close people than among strangers (Binzel & Fehr 2013). However, experiments on the
reputations (Berg, Dickhaut, & McCabe, 1995). This is essential since peers on virtual platforms do not know each other when making the decision to share.
The next paragraph addresses the importance of peer-to-peer trust on a car owner’s intention to participate in sharing, after which is discussed how respectively platform positioning and the presence of mutual Facebook friends influence this.
2.4.1 Peer-to-peer trust
Without mutual trust, the probability that two strangers engage in any monetary transaction is low (Bonsón Ponte, Carvajal-Trujillo & Escobar-Rodríguez 2015; Kim, Chung & Lee 2011). As a consequence of the stranger-danger bias, human beings are programmed to not trust strangers, which impedes the intention to share with them (Gebbia 2016). Platforms need to overcome the fear of opportunism and misbehavior by creating peer-to-peer trust, which is a key determinant for participation in the sharing economy (e.g. Botsman & Rogers 2010; Hamari et al. 2016; Möhlmann 2015; Slee 2013). The level of peer-to-peer trust furthermore predicts future sharing behaviors, which is essential for a sustainable development of the sharing economy at large (Stofberg 2017).
Logically, platforms attempt to augment the trust between users. However, the absence of face-to-face contact makes the evaluation of trustworthiness of the other actor difficult (Ishaya & Mundy 2004). Social cues such as body language that normally serve as an indication for the trustworthiness of a business partner, are now eliminated by the online environment (Pavlou & Gefen 2004). Platforms respond to this by creating reputation mechanisms. Usually these are user profiles that include all sorts of information, such as reviews and ratings from peers, pictures and information about themselves in order to minimize the social distance and maximize the degree of identification (Ert, Fleischer and Magen 2016). However, Frenken and Schor (2017) find that in general ratings are “inflated and not very accurate”. For instance, on ride sharing platform Blablacar, more than 98% of
the ratings are five stars ratings, which means the reputation mechanism cannot discriminate between trustworthy and non-trustworthy drivers and therefore fails as a reputation
mechanism (Slee 2013, p. 6). Accordingly, Ert, Fleischer and Magen (2016) find that online review scores do not have any effect on the price or the likelihood of participation. It is even argued review systems decrease the chance interpersonal relationships are formed (Schor 2014). Concluding, other ways that constitute peer-to-peer trust have to be found.
As discussed in section 2.2, peer-to-peer relationships based on either economic or social exchange come with different expectations of the other’s underlying motivations, which impacts the level of interpersonal trust (Bridoux & Stoelhorst 2016). In a relationship based on economic exchange acting out of self-interest is considered the appropriate behavior, creating “interpersonal reserve amongst participants, which limits relational depth and trust” (Bridoux & Stoelhorst 2016, p. 237). In contrast, in a social exchange relationship,
participants expect the other’s motivations to transcend merely self-interest, in which the common interest can even be put before one’s own (Stofberg 2017). Therefore, participants expect others to respect them, their interpersonal relationship and their property (Bridoux & Stoelhorst 2016), which builds peer-to-peer trust rather than diminishing it (Stofberg 2017). This leads to the following hypothesis:
H3. The relationship between positioning and intention to participate is mediated by
peer-to-peer trust, such that positioning based on social (economic) exchange is associated with higher (lower) levels of peer-to-peer trust, which positively (negatively) influences intention to participate.
A second way to build peer-to-peer trust tested in this study is the presence of mutual Facebook connections on the platform. Mutual connections are extremely powerful when it comes to trust. Trust between two strongly connected people, e.g. family members or friends,
is increasing when this relationship is embedded in a network of mutual connections (Burt & Knez 1995). Not only between people who already have a connection, this effect also holds for people without any connection as is the case in a carsharing context. Granovetter (1985) argues that trust emerges from ‘structural embeddedness’, which means trust is more likely between two individuals with mutual friends. “The degree of social connection (…) - the number of friends they have in common and the duration of their acquaintanceship - generally predicts the levels of trust and trustworthiness” (Glaeser et al. 2000, p. 814). Burt and Knez (1995) outline the reason behind the strength of mutual connections, to which they refer as ‘third parties’. There are two contexts in which trust relations take place. When two peers interact disconnected from others, the mutual trust relationship is a “cumulative result of their exchanges, or interaction games, with one another” (Burt & Knez 1995, p. 256), which is called an isolated dyad. Trust is always interpersonal, however, it is barely private. Usually the context for trust relations is an embedded dyad, in which two individuals are surrounded by their networks of interconnected friends and acquaintances. The trust game is no longer private, but surrounded by third parties (Burt & Knez 1995). Kreps (1990) explains this phenomenon with the reputation effect. The presence of indirect connections makes the behavior of both actors more public, increasing the salience of their reputation. This results in a higher probability of mutual cooperation and trust (Burt & Knez 1995), even when both actors did not have a connection beforehand (Glaeser et al. 2000). It is for this reason we expect that four mutual friends on Facebook will act in a similar manner and create an embedded dyad (Burt & Knez 1995), which enhances the trust between peers and, in turn, a car owner’s intention to participate. Thus, this study hypothesizes:
H4. The relationship between mutual Facebook friends and intention to participate is
mediated by peer-to-peer trust, such that the presence (absence) of mutual Facebook friends is associated with higher (lower) levels of peer-to-peer trust, which positively (negatively) influences intention to participate.
2.4.2 Platform trust
As previously mentioned, this study does not only address the peer-to-peer trust, but also the ‘classic’ type of trust every company encounters: platform trust. Trust in a company increases the intention to buy products or services (Stewart 2003). Many scholars from various
disciplines agree that the trust in a vendor is an important condition for the willingness of customers to engage in transactions (McKnight, Choudhury & Kacmar 2002). In online environments, platform trust is found to be even more important than in physical ones (Bailey & Bakos 1997). Given the fear of opportunism and misbehavior, users need to believe that the platform will help when undesirable behavior occurs (Botsman 2016). Hence, platform trust is important to overcome the fear of opportunism and misbehavior and argued to be a key
facilitator of electronic marketplaces (Pavlou, Tan & Gefen 2003). Therefore, platform trust is expected to be a predictor of the intention to participate.
Given the importance of platform trust in stimulating people to share, this study attempts to increase the platform trust through positioning and showing mutual Facebook friends. Building on the three antecedents of platform trust, the independent variables are argued to have a positive impact on platform trust. The three antecedents of platform trust are reputation, social presence and benevolence (Kim, Yoon and Zo 2015). Firstly, the perceived reputation of the platform to consumers is a discriminating signal promoting trust (Slee 2013). The consumer perceives this reputation prior to a transaction. Secondly, online platforms provide pictures and sound on their websites to deliver social presence. Social presence is defined as the extent to which a user perceives other users to be psychologically present (Kim,
Yoon and Zo 2015). Amongst others, Teubner et al. (2014) found social presence to be fostering trust building in online platforms. The perceived social presence “is built upon signals of user interactions that are provided via website functionality” (Kim, Yoon and Zo 2015, n.p). The last antecedent of trust is assumed to be benevolence; the greater the
benevolence is, the greater the trust. Sharing goods with strangers with benevolent purposes, in contrast to commercial purposes, leads instinctively to building platform trust (Kim, Yoon and Zo 2015). Benevolence is seen as the belief that a commercial sharing platform is
genuinely interested in the welfare of the consumer (Ba & Pavlou 2002).
Following these three antecedents of platform trust – reputation, social presence and benevolence – it appears platforms can be designed not only for peer-to-peer (Gebbia 2016; Keymolen 2013), but also for platform trust. This thesis asserts that a platform positioning based on social exchange has a higher probability to be associated with high levels of platform trust. The perceived social presence is assumed to be higher in the social exchange condition due to the information about and pictures of peer users of the platform, in contrast to functional information and businesslike pictures in the economic exchange condition (Teubner et al 2014). Furthermore, we build on Kim, Yoon and Zo’s (2015) statement that if peers share with benevolent purposes (not commercial), platform trust is built. To increase the platform trust, the platform should emphasize these benevolent intentions and additionally increase the social presence in the design of the website. We argue that a social exchange positioning is better capable of communicating social presence and benevolence than economic exchange positioning2. Following this reasoning, we hypothesize:
2 The third antecedent of platform trust, reputation, evolves from patterns of previous behavior (Mayer, Davis & Schoorman 1995) and is peer-to-peer, informal, decentralized, community driven and non-commercial (Slee 2013). Therefore, reputation cannot be manipulated in a webpage that is shown once to respondents, thus, the effect of reputation as an antecedent of platform trust is assumed to be neutral.
H5. The relationship between positioning and intention to participate is mediated by platform trust, such that positioning based on social (economic) exchange is
associated with higher (lower) levels of platform trust, which positively (negatively) influences the intention to participate.
Furthermore, this thesis asserts that showing mutual Facebook friends on the platform is a signal of user interactions on the website (Chen 2013). Facebook profiles, and especially the mutual friends, are argued to make other users feel psychologically present, which is building the perceived social presence. Additionally, the Facebook element helps consumers in judging the trustworthiness of peers, arguing that the platform invests in the consumers’ welfare and by doing this increase their perceived benevolence. As a consequence, by influencing the antecedents of platform trust (Kim, Yoon and Zo 2015), the Facebook manipulation is expected to have a positive effect on platform trust, which has in turn a positive effect on a car owner’s intention to participate.
H6. The relationship between mutual Facebook friends and intention to participate is mediated by platform trust, such that the presence (absence) of mutual Facebook friends is associated with higher (lower) levels of platform trust, which positively (negatively) influences the intention to participate.
2.4.3 Interaction between peer-to-peer and platform trust
Platform trust and peer-to-peer trust should not be considered as two independent variables. A large body of literature suggests that platform trust can be transferred to peer-to-peer trust (e.g. Zucker 1986). “Considering trust transfer as a cognitive process, transfer occurs when a
person (…) bases initial trust in an entity (…) on trust in some other related entity” (Stewart 2003, p. 4). Platform trust can thus be transferred to trust in the members of the platform, but also vice versa. Chen, Zhang and Xu (2009) give three arguments for this. First of all, when strong trust relations are present within the community, users are more likely to together generate a feeling of trust in ‘their’ institution. Secondly, a platform can cultivate users’ trust by showing it cares about peer-to-peer social relationships. Efforts of the platform to foster relationships are seen by users as a signal that the platform would not act opportunistically towards them. Lastly, the platform proves it is able to “offer effective management of the environment as well as the benevolence to further nurture a healthy environment” by gathering trustworthy people (Chen, Zhang & Xu 2009, p. 153). Because of this close interdependent relationship, the following is hypothesized:
H7. The relationship between positioning and intention to participate is sequentially mediated by platform trust and peer-to-peer trust, such that social (economic)
exchange positioning is associated with higher (lower) levels of peer-to-peer-trust and platform trust, which leads to a higher (lower) intention to participate.
H8. The relationship between the presence of mutual Facebook friends and intention to participate is sequentially mediated by platform trust and peer-to-peer trust, such that the presence (absence) of mutual Facebook friends is associated with higher (lower) levels of peer-to-peer-trust and platform trust, which leads to a higher (lower) intention to participate.
This study looks into the effects of positioning and mutual Facebook friends on car owners’ intentions to participate in carsharing, independently of each other. The positioning of the platform is based on either a social or an economic exchange relationship. The social
exchange positioning triggers CS and EM relationships between peers whereas the economic exchange positioning primes an MP relationship. The former is hypothesized to be positively related to intention to participate, whilst the latter is expected to be negatively related to car owners’ intentions to participate. The presence of the mutual Facebook friends manipulation should lead to a feeling of perceived connectedness between peers, resulting in a higher intention to participate in sharing. Both of the relationships between the independent variables and the intention to participate are expected to be mediated by peer-to-peer trust and platform trust. Given the interdependent relationship between these two mediators, they are
furthermore expected to mediate the relationships between the independent and dependent variables in serial. Figure 3 illustrates the conceptual framework with the corresponding hypotheses. Table 1 presents an overview of the hypotheses.
Figure 3.
Table 1.
3. Methodology
This chapter discusses the methodology used in this study. First of all, the choice for
carsharing platform SnappCar is explained (paragraph 3.1). This is followed by the research design (paragraph 3.2), which consists of a combination of a vignette experiment and a survey questionnaire. Two pre-tests are performed prior to the research to ensure the vignettes
communicate what they intend to communicate and to check the credibility of the hypothetical situation (paragraph 3.3). Thereafter, the constructs used in this study are presented (paragraph 3.4), followed by the data collection procedure (paragraph 3.5). After a description of our sample (paragraph 3.6), the data analysis process is outlined (paragraph 3.7).
3.1 Platform Choice: SnappCar
SnappCar is chosen for this research since it is a peer-to-peer sharing platform on which people share their own goods with other users. The company was willing to participate in order to find out how to stimulate car owners to share their cars. Like many carsharing platforms, SnappCar positions itself mainly in a functional and financial manner on their website. The current slogan of SnappCar for car owners is as follows: “Compensate your car costs and waste less at the same time? Share your car easily and carefree through SnappCar and earn up to €1,000 per year.” The fact that SnappCar is highlighting utilitarian benefits provides the opportunity to research whether, in contrast, highlighting social cues would lead to a higher intention to participate among car owners.
3.2 Research Design
This research combines a vignette study and a closed-ended question survey questionnaire to collect the data. An experimental vignette methodology enables us to manipulate and control
the independent variables in order to test the hypotheses in a realistic manner (Aguinis & Bradley 2014). A vignette, also known as a factorial survey (Rossi & Anderson 1982), uses “short descriptions of situations or persons (vignettes) that are usually shown to respondents within surveys in order to elicit their judgments about these scenarios” (Atzmüller & Steiner 2010, p. 128). This technique enables researchers to study issues “in a way which
approximates to the complexities with which such issues are surrounded in reality” (Finch 1987, p. 111). In short, vignette studies have the advantages of qualitative methods combined with the benefits of the survey method (Finch 1987). In order to enhance the realism of the situation, the vignettes are designed as SnappCar’s landing page for car owners. In these vignettes, the independent variables positioning and the mutual Facebook friends are
manipulated. The respondents are placed in the hypothetical situation in which they own a car and encounter the SnappCar landing page online. Those who do not own a car are asked to imagine this is the case. The respondents are randomly assigned to conditions in a 2 (social vs. economic exchange) x 2 (presence vs. absence of Facebook) between-subjects design.
3.3 Stimuli Development (pre-test)
In order to check whether the manipulations of the landing page are perceived as intended and to check the credibility, an exploratory pre-test is conducted. Firstly, the four vignettes were created. The content is based on the existing content on SnappCar’s website. All vignettes consist of five components: an introduction; ‘Why SnappCar’; reviews and ratings; ‘How does it work?’; and a contact section (Figure 4-7). Since the majority of the ratings on sharing platforms have five stars (Slee 2013) and ratings are found to have no effect on the intention to participate (Ert, Fleischer & Magen 2016), five-star ratings are inserted for all conditions.
The social exchange condition consists of cues stressing social benefits of sharing. The respondents are informed about SnappCar as a carsharing platform on which neighbors can
share their cars in order to get to know each other and to create an environmentally friendly and social neighborhood. Calculations on what is given and received are avoided (Habibi et al. 2017). The slogan of this condition is: “Share your car via SnappCar and help your neighbors move forward! Make your neighborhood yours again”. The terms ‘neighbors’ and ‘sharing’ are used to evoke a social exchange relationship between users.
The respondents in the economic exchange condition learn that SnappCar is a convenient way to earn money by renting out their cars at a moment they are not using it themselves. Financial and utilitarian benefits are communicated to trigger a MP relationship, such as a calculation of the money a car owner can earn (Habibi et al. 2017). The slogan used in this manipulation is as follows: “Earning money with your car when you’re not using it? Rent out your car easily and safely via SnappCar and earn up to €1000 per year”.
The second independent variable is tested by visually showing the Facebook
connections between a car owner and a potential renter. This social media network is chosen as it is the biggest social media network within the Netherlands and people from various ages are familiar with it (Van der Veer, Boekee & Peters 2017).
We ran two pre-tests to check whether our manipulations worked before finalizing our vignettes (N=73 and N=72). These results are outlined prior to the description of our final research.
3.3.1 Pre-test 1
After finalizing the content and design of the vignettes (see Appendix A), 73 respondents were randomly assigned to one of the four conditions. The vignettes were followed by a description of each of the three relational models. The respondents were asked to indicate how applicable the description was to the condition they had seen on a 7-point Likert scale, ranging from ‘not at all’ to ‘a very high degree’. To judge the match with CS theory, we used
the following definition: “A platform on which transactions are characterized by a high degree of generosity. On this platform people have the feeling they belong to the same group and people have a lot in common.” To assess the match with EM, respondents were shown: “An online sharing platform that arranges transactions on the basis of equality and reciprocity. People on this platform try to keep a healthy balance in terms of the benefits that everyone gets from the platform (money or things).” Lastly, the extent to which the website matched MP was assessed: “An online sharing platform that arranges transactions based on 'you get out of it what you put into it.' People on this platform believe that you are entitled to a good return on what you put into the platform (money or things). A bit like a business relationship.”
For the second independent variable mutual Facebook friends is tested to what extent the manipulation communicates a feeling of connectedness with the Inclusion of Other in the Self (IOS) Scale, which is a single-item, pictorial measure of closeness (Aron, Aron & Smollan 1992). The full survey questionnaire of the pre-test can be found in Appendix B. Table 2 presents an overview of the independent variables of the first pre-test.
Table 2.
Overview independent variables pre-test 1
Independent variables Levels
Positioning Social exchange
Economic exchange
Mutual Facebook friends High perceived connectedness (4 mutual friends) Low perceived connectedness (0 mutual friends)
Table 3 presents the results of pre-test 1 that shows both the social and economic exchange conditions are significantly correlated to the descriptions of CS and MP relationships (r=.283,
p<.05 and r=-.453, p<.01). The correlation between perceived connectedness and CS is also
significant (r=.351, p<.01), however, we do not find any significant results between mutual
Facebook friends and perceived connectedness (r=.137, p >.05). Even though the positioning
differences between the conditions (F(1, 70)=.013, p=.909), it is decided to do a second pre-test attempting to get significant results for both independent variables.
Table 3. Correlations pre-test 1 Mean SD 1 2 3 4 5 6 1. Positioning .562 .500 2. Facebook .589 .495 .104 3. CS 4.16 1.500 .283* -.057 4. EM 4.95 1.153 .102 .130 .094 5. MP 4.41 1.153 -.453** -.043 -.339** .226 6. Connectedness 5.96 1.409 .215 .137 .351** .258* .008
*. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed). ***. Correlation is significant at the 0.001 level (2-tailed).
N=73
3.3.2 Pre-test 2
For the second pre-test, the design and content of the vignettes were adapted to ensure the Facebook element would be more present on the page (Figure 4-7). Instead of having a high/low construction (Table 2), the new vignettes were turned into a presence/absence construction as illustrated in Table 5. Furthermore, the perspective of the Facebook element is changed so that the mutual friends would be hypothetical friends of the respondent himself, instead of visualizing connections between third parties. The Facebook element is integrated in the ‘How does it work?’ section to enhance the clarity of the page. To ensure the Facebook manipulation could not be missed by respondents, we also mentioned the four common friends in the ‘review’ section and showed the Facebook logo in the ‘Why SnappCar’ section. This attempt to make the homepage more personal and to increase the presence of the
Table 4. Correlations pre-test 2 Mean SD 1 2 3 4 5 6 1. Positioning .5 .504 2. Facebook .458 .502 .028 3. CS 4.15 1.725 .251* .374 4. EM 4.86 1.225 -.137 .082 .170 5. MP 4.26 1.768 -.293* -.186 -.397** .173 6. Connectedness 5.94 1.537 .091 .344** .449** .018 -.362**
*. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed). ***. Correlation is significant at the 0.001 level (2-tailed).
N=72
Mutual Facebook friends now correlates significantly with perceived connectedness (r=.344, p<.01). The social and economic exchange conditions are still significantly correlated with the
CS and MP descriptions (r=.251, p<.05 and r=-.293, p<.05) and the CS description with
perceived connectedness (r=.449, p<.01). Lastly, the credibility has increased to M=5.29 with
no significant differences between the groups (F(1, 70)=1.040, p=.311). We regarded these outcomes sufficient to proceed the research with the vignettes from the second pre-test (Figure 4-7).
Table 5.
Overview final independent variables
Independent variables Levels
Positioning Social exchange
Economic exchange
Mutual Facebook friends Perceived connectedness (presence of Facebook) No perceived connectedness (absence of Facebook)
Figure 4.
Figure 5.
Figure 6.
Figure 7.
3.4 Survey Questionnaire and Measure Development
After finalizing the vignettes, the survey questionnaire is created. This is used to enhance the understanding of the dependent variable intention to participate and the mediators
peer-to-peer trust and platform trust. Further, it enables the collection of information about the
demographics of the respondents and multiple control variables. An online survey
questionnaire allows to acquire a large amount of data, which are relatively easy to interpret given the standardized closed-ended questions (Saunders, Lewis & Thornhill 2009).
Additionally, it is time and cost efficient compared to face-to-face data collection (Wright 2005). A drawback could be misinterpretation or misunderstanding of the questions (Fricker & Schonlau 2002). To overcome this problem the survey questionnaire is fully written in Dutch so that the respondents were able to participate in their first language. To ensure the content of the Dutch items within the constructs are similar to the English versions, they were translated and translated by a third person. The discrepancies between the
back-translations and original questions were corrected in the final version. The original Dutch survey questionnaire can be found in Appendix C.
The measures used in this research are existing constructs that are adopted and adapted to the sharing economy context and in particular to SnappCar. By using existing constructs, we attempt to ensure the reliability and internal validity of the research. Next, the constructs are outlined for each variable. An overview of the full constructs is presented in Appendix D.
Perceived relational models
To check whether the positioning manipulations communicate the corresponding relational models, respondents were asked to indicate on a 7-point scale to what extent they expected members of the presented platform to match descriptions of the three relational models: “SnappCar is a carsharing platform one only participates in for financial gains” (MP);
