BROADENING THE UNDERSTANDING OF
ACCESSING INTENTIONS
A research on the effect of the type of provider and product value on
the consumer’s behavioral intention towards adopting access-based services
Grace Ooijen (10520066) Universiteit van Amsterdam
Master Thesis on Business Administration: Digital Business June 21, 2017
Supervisor: Merve Güvendik Second reader: …
Statement or originality
This document is written by Grace Ooijen 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
Abstract ...7
Introduction ...9
Literature review ...10
Definitions of access-based consumption ...11
Related work and theoretical framework ...12
The mediating effect of consumer trust ...13
The mediating effect of perceived risk ...14
Product value and the consumer’s price consciousness ...16
Methodology ...18
Experimental design ...18
Sample and procedure ...18
Stimuli ...19
Measures ...20
Perceived trustworthiness ...20
Perceived risk of failure ...20
Perceived financial risk ...20
Behavioral intention ...20 Price consciousness ...21 Control variables ...21 Survey pre-test ...21 Statistical procedure ...22 Results ...23 Factor analysis ...23 Correlations ...25 Hypothesis testing ...26
Direct effects of the type of provider and product value on behavioral intention ...26
Mediating effects of the perceived trustworthiness, risk of failure, and financial risk ...27
Moderating effect of price consciousness ...31
Additional exploration in the consumer’s perceived financial risk ...31
Discussion ...32
General discussion ...32
Limitations ...35
Conclusion ...37
References ...39
Appendix A: Survey questions ...42
Appendix B: Stimuli ...46
Abstract
In the course of recent years, access-based consumption has become a concept that has been widely applied in many industries and used by people from all age categories. This rapid emergence of sharing services triggered various researchers to study this concept and its drivers and deterrents to adoption. Broadening the understanding on this topic would help entrepreneurs in the field shape their business. This study aimed to examine the effect of the type of provider and the product value on the consumer’s behavioral intention of adopting access-based services. The hypotheses were tested with data obtained from an online experimental survey. Resulting from various statistical tests such as correlations, a principal factor analysis, an ANOVA, and mediation and moderation analyses, this study had demonstrated that the type of provider and the product value do not affect the consumer’s behavioral intention towards adopting access-based services. Nor did the consumer’s perceived trustworthiness in the company or their perceived risk of failure fulfill a mediating role in either of these expected relationships. Furthermore, the consumer’s price consciousness did not moderate the relationship between the product value and its behavioral intention. Lastly, evidence was found that the consumers’ perceived financial risk partially mediated the relationship between product value and their behavioral intention to adopt access-based services, such that consumers that perceive high financial risks are estimated to have less intention to adopt the service. However, limitations of the stimuli design and the characteristics of the obtained sample may have contaminated these outcomes.
Introduction
The past few years was a time of remarkable, rapid growth for companies like Uber and Airbnb. The latter was responsible for connecting 47,000 renters to hosts in the summer of 2010 (McAlone, 2015). Outstandingly, this increased to 17 million renters in the summer of 2015 (McAlone, 2015). Uber has experienced a similar staggering growth worldwide (Price, 2015). What these fast-growing companies have in common is their economic model based on sharing rather than ownership. This shift in consumption mode induced the phenomenon of access-based consumption, whereby underutilized goods are offered for renting or trading via a mediating commercial organization or social platform (Botsman & Rogers, 2010). A corollary effect of this was the disruption of various industries, such as hospitality, retail, and transportation (Owyang, Tran & Silva, 2013). For instance, households with excess capacity of accommodation can list their spare rooms on Airbnb, thereby eating into incumbent hospitality businesses. Various companies from different industries recognize this increasing threat of collaborative consumption, creating an incentive to turn it into a fruitful opportunity (Lamberton & Rose, 2012; Belk, 2014). For instance, BMW introduced DriveNow, a service lending premium cars fulfilling the need of personal mobility among millennials, whereas Citi successfully launched Citi Bike in order to be associated with an environmentally friendly transportation model (PwC, 2015). Leveraging this economic model of sharing has turned businesses to flourish (Tussyadiah, 2015) and enabledconsumers to utilize other means of obtaining product benefits (Lamberton & Rose, 2012).
Typically, an increased behavioral intention towards access-based consumption among prospective consumers would generate more revenues for the companies facilitating this sharing system. Extant literature fixated on identifying the variables that affect this behavioral intention. These variables include cost benefits associated with sharing (Lamberton & Rose, 2012; Botsman & Rogers, 2010; Tussyadiah, 2015; Hamari et al., 2015), perceived risk of sharing the good (Lamberton & Rose, 2012, Baumeister & Wangenheim, 2015), the societal aspect of sustainability (Heinrichs, 2013; Tussyadiah, 2015; Hamari, Sjöklint & Ukkonen, 2015), and the degree of added convenience (Moeller & Wittkowski, 2012). However, some variables have yet to be examined. As many companies to date have leveraged the idea of access-based consumption, Heinrichs (2013) acknowledged that academic literature regarding this matter lagged behind the fast-moving practice. To the author’s best knowledge, research specifically regarding the potential effects of the value of the accessible good (i.e., the general price level of ownership, whether it is low or high) on the behavioral intention is non-existent, as well as the effect of the type of provider of the accessible good. The provider (i.e., whether a company in business-to-consumer services or individual in peer-to-peer services) has full ownership of the good they made accessible. Hence, the research question of this study is as follows:
How do the type of provider and value of a good affect the consumer’s behavioral intention towards adopting access-based services?
In prior research on the consumer’s behavioral intention in access-based consumption, no attention has been devoted to the good’s type of provider or its value. In addressing the proposed
research question, the limited understanding of this matter is advanced. Lamberton and Rose (2012) found that the consumer’s perceived risk is associated with its sharing propensity—the likelihood of engaging in a sharing system. Similarly, this study examines the consumer’s behavioral intention towards adopting access-based services and the mediating role of the perceived risk of consuming a shared good. In doing so, the validity of the relationship between the consumer’s perceived risk and its behavioral intentions is tested in a different context and thereby substantiates the extant findings. As this study is explorative, its findings may also introduce topics for future research identifying the determinants of possible relations that have been brought to light. Moreover, exploring which condition manifests to be most auspicious is insightful for helping new entrants shape their business. The negative effects on the behavioral intention of consumers disclosed in this study can be minimized and the positive effects can be cultivated. Ultimately, thereby increasing the market share of access-based companies may provide environmental benefits (Botsman & Rogers, 2010; Heinrichs, 2013).
The context of this study will be narrowed down to transportation in the sharing economy, as most studies to date have been carried out solely in the context of tourism and hospitality. In addition, the transportation industry is inherently more suitable for the experimental survey method used in this study, considering business-to-consumer products in the travel sector generally involve legitimate hotel rooms, which commonly are less subject to the perceived risks associated with access-based consumption. Furthermore, the transportation sector has currently gained more ground than the retail industry, as 9 percent of the US adults have engaged in a sharing economy transaction involving automotive and transportation, against 2 percent involving retail (PwC, 2015).
This paper is organized as follows: the next chapter comprises the literature review including a more thorough discussion on access-based consumption, as well as elaborating on theories and constructs relevant to this study. Additionally, the hypotheses are proposed in this section. Then, chapter three outlines the methodology, in which the experimental design, research method, and initial findings on the collected data are discussed. Subsequently, the results of the data analysis are denoted and discussed in chapter four, whereafter the general discussion, the limitations of this study, and the conclusion are presented in the last chapter.
Literature review
In this section, some light has been shed on the lack of consensus on the concept of sharing systems. Then, an understanding of the constructs relevant to this study’s context is raised and hypotheses are developed. The conceptual model showcasing these hypotheses is then presented.
Definitions of access-based consumption
Although various sources fail to explicitly define the term of the phenomenon they examined in their study, the table below summarizes the many terms and the corresponding definitions of studies in the existing literature that did define their version of access-based consumption. This table stresses the lack of one general definition.
Table 1. Several terms that describe the phenomenon of sharing systems.
Term Definition Source
Sharing economy Collaborative consumption made by the activities of sharing, exchanging, and rental of resources without owning the goods.
(Lessig, 2008)
Sharing economies allow individuals and groups to make money from underused assets. In this way, physical assets are shared as services.
(PwC, 2015)
Collaborative consumption
An economic model based on sharing, swapping, trading, or renting products and services, enabling access over ownership.
(Botsman & Rogers, 2010)
Collaborative consumption is people coordinating the acquisition and distribution of a resource for a fee or other compensation.
(Belk, 2014)
The peer-to-peer-based activity of obtaining, giving, or sharing the access to goods and services, coordinated through community-based online services.
(Hamari et al., 2015)
Peer-to-peer sharing The concept involves individuals exchanging, redistributing, renting, sharing, and donating information, goods, and talent, either organizing themselves or via commercial organization by social media platforms.
(Heinrichs, 2013)
Collaborative economy The collaborative economy is an economic model where ownership and access are shared between corporations, startups, and people.
(Owyang, Tran & Silva, 2013) Prosumption Production or co-production of some good as a means
to consumption of that good.
(Toffler, 1980) The interrelated process of production and
consumption.
(Ritzer, 2014) Access-based
consumption
Transactions that can be market mediated but where no transfer of ownership takes place.
(Bardhi & Eckhardt, 2012)
The mesh The mesh describes a type of network that offers
people to connect to one another and offer sharable physical goods.
There is overlap with similar terms like ‘co-creation’ (Prahalad & Ramaswamy, 2004) and ‘commons-based peer production’ (Benkler, 2002). However, to a greater extent, these terms are associated with knowledge sharing and the creation of a new product. In this study the focus lies at sharing existing physical goods and services, to which the definition of Bardhi and Eckhardt (2012) is best suited as it describes the sharing platforms of today in a clear fashion and is not restricted by the word ‘sharing.’ Namely, Belk (2010) acknowledges that, although both ‘access’ and ‘sharing’ do not imply the transfer of ownership, they differ on several aspects: (1) sharing may be originated from altruism, which does not apply to accessing, and (2) sharing indicates collaborative consumption of personal goods, meaning it would exclusively constitute peer-to-peer activities. This study includes the case in which companies provide the goods being accessed and consumers, hence, engage in business-to-consumer activities (e.g., Zipcar). This entails companies having full ownership of the accessible goods and providing its customers a service enabling them to access their goods for a fraction of the costs of ownership. Companies facilitating peer-to-peer services (e.g., RelayRides) commonly do not own as much assets as traditional companies. Instead, the providing customers have full ownership of the goods that are exchanged with consuming customers—whether by means of a fee or non-monetary compensation (Belk, 2014). These peer-to-peer markets allow local individuals and small suppliers to compete with traditional providers of goods or services (Einav, Farronato & Levin, 2016). Remarkable about sharing systems is that the benefits enjoyed by the participating consumers commonly grow as more people participate in sharing (Belk, 2010), unlatching increased diversity in goods and locations.
Related work and theoretical framework
The key publication on access-based consumption to date is the book of Botsman and Rogers (2010), in which they extensively explore the significance of the emerging sharing economy and its effects on consumerism and sustainability. This widely cited book reported that sharing goods increases resource efficiency and reduces waste, contributing to the environment’s well-being. This was already noticed by Durgee and O’Connor (1995), who highlighted the extended lifecycle of used, but intact, products by renting as consumption. However, Schor (2014) stresses the lack of comprehensive studies proving the environmental impact of the sharing economy. She gives the example of sharing cars, namely, providing access to cars is likely to produce more footprint if this was inaccessible formerly, forcing consumers to have opted for the greener alternative of walking or bicycling. She adds that the ripple effects are often overlooked, e.g., whether renters will spend their added income on low-impact goods or not. According to Boesler (2013), Botsman and Rogers also fail to acknowledge the hypothesized future economic downside of collaborative consumption. He suggests that access-based consumption may have catastrophic economic consequences as businesses typically rely on a society where ownership is the default mode of consumption. Conclusively, Schor (2014) argues in her critical essay that the excitement for the sharing economy is nevertheless rational, as long as a movement ensuring social solidarity, democracy, and sustainability in the sharing economy is created.
Belk (1988; 2007; 2010; 2014) spent much research on ownership, sharing behavior, and the extended self—one’s possessions being the extensions of the self. The finding that “we are what we have” (Belk, 1988, p. 160) impacted academia and practitioners substantially. It suggested that ownership is a fundamental construct of consumer behavior (Belk, 1988). As, in the case of this study, the ownership of the good involves the providing party (i.e., either a company or individual), rather than the consuming customer, it would be interesting to investigate whether ownership would be a factor in a consumer’s behavioral intention. The first hypotheses are therefore developed with the type of provider of the product in mind. Then, the potential effects of the product value and the consumer’s price consciousness on the consumer’s behavioral intention towards adopting access-based services are discussed.
The mediating effect of consumer trust
Consumer trust is defined as “the willingness to rely on an exchange partner in whom one has confidence” (Moorman, Deshpande & Zaltman, 1993, p. 82). Comer, Plank, Reid and Pullins (1999) distinguish between three components of consumer trust: trust in the company, salesperson, and the product. The product trust will be addressed later on. In the case of this study, the consumer’s perceived trustworthiness relies both on how the company and the salesperson (i.e., the provider of the good) come across. However, the salesperson directly reflects on the company, therefore, the focus is on examining the effects on the latter. This would also make for interesting and relevant practical implications by addressing the question ‘how does the type of provider affect the trustworthiness of the company?’
Although the literature is sparse, Schor (2014) provides anecdotal evidence that the shared good’s type of provider contributes to the consumer’s perceived trustworthiness in access-based services. Schor (2014) argues that peer-to-peer sharing would entail a higher degree of risk as it involves trusting unfamiliar individuals with rather intimate goods. On the contrary, it has been observed that there is an extant distrust of big global brands across different industries (Gansky, 2011), indicating that consumers may prefer a peer-to-peer sharing system, rather than business-to-consumer. Also, although PwC (2015) did not make a distinction between business-to-consumer and peer-to-peer companies, they similarly reported that 69 percent of the US adults will not trust sharing economy companies until they are recommended by someone they trust. These contrasting views serve an interesting impetus to test if and how the type of provider influences consumer trust on consuming a shared good and thereby the consumer’s behavioral intentions. The expectance is that the relationship between the type of provider of a good and the consumer’s behavioral intention towards adopting access-based services is mediated by the consumer’s perceived trustworthiness. Accordingly, the following hypotheses are formulated:
H1a: The type of provider of a good affects the consumer’s perceived trustworthiness in the company, such that peer-to-peer services are trusted more than business-to-consumer services.
H1b: The consumer’s perceived trustworthiness has a positive effect on the consumer’s behavioral intention towards adopting access-based services.
The latter hypothesis is based on Vlachos, Tsamakos, Vrechopoulos and Avradmidis’ (2009) conclusion that consumers are more inclined to show desirable behavior when the trust in the service is high. Vlachos et al. (2009) examined the mediating role of consumer trust in the context of corporate social responsibility and thus found that consumers are more likely to be attracted and retained when the consumer trust is high. Their conclusion builds on the findings of Morgan and Hunt (1994), who theorized that consumer trust and commitment are central to successful relationship marketing. Thus, arguably, there is evidence that the consumer’s perceived trustworthiness may be positively correlated with its behavioral intention towards adopting access-based services. Consumer trust commonly fulfills a mediating role in various disciplines (Vlachos et al., 2009). Likewise, this study examines whether consumer trust is a mediating variable between the good’s type of provider and the consumer’s behavioral intention.
There is no literature available on the potential direct relationship between the shared good’s type of provider and the behavioral intention of the consumer towards adopting an access-based service. Nevertheless, it would be interesting to investigate whether this research gap is rightly neglected or that a direct correlation between the two variables can be discovered. Hence, the following hypothesis is proposed:
H2: The type of provider of a good affects the consumer’s behavioral intention towards adopting access-based services, such that consumers are more inclined to adopt access-based services when they involve peer-to-peer activities, rather than business-to-consumer activities.
The mediating effect of perceived risk
A determinant previously brought into relation with behavioral intentions towards adopting access-based services, is the perceived risk of consuming the shared good. The perceived risk of the consumer has numerous variations in different contexts. Stone and Grønhaug (1993) present six dimensions explaining 88.8 percent of the perceived risk in purchasing electronics, of which two risks particularly apply to access-based consumption. The first being performance risk and the second being financial risk. This is in accordance with Baumeister and Wangenheim’s (2014) comprehensive empirical study, who found the functional and monetary perceptions of consumers to be most important in determining their attitude towards various consumption modes. Regarding functional risks, they suggest that the risk of failure in an accessed product is perceived to be highly present. This risk is defined as “the perceived probability of a technical failure or partial damage of the product during use” (Baumeister & Wangenheim, 2014, p. 8) and relates to Comer et al.’s (1999) identification of product trust in successful buyer-seller relationships. After all, a shared good may be perceived to be subject of over-usage in comparison to owned goods. This perception is confirmed to emerge among 5 percent of the renters (Durgee & O’Connor, 1995), likely preventing consumers to engage in sharing systems. Featherman and Pavlou (2003) draw similar conclusions in their study on e-services adoption. Their results indicated that performance-based risk perceptions primarily negatively affect e-services adoption (Featherman & Pavlou, 2003).
The value of the accessible good—referring to the general price level of ownership—may influence this perceived risk of product failure. The line of thought here is that lower valued products, such as bicycles, have a higher risk of failure in a sharing system than higher valued products, such as cars. Typically, bicycles are considered to be more prone to deficiencies. This is because the quality standards of cars are much higher, taking into account the consequences in the case of failure. The mechanics of cars ought to be much more robust than bicycles. Baumeister and Wangenheim (2014) presented results possibly in conformity with this thought. Namely, the perceived risk of failure among consumers was found to be 37.9 percent higher in accessing bicycles than in accessing cars, a significant difference in product categories of which merely its purchase price differs (Baumeister & Wangenheim, 2014). In summary, the expectancy is the relationship between the product value and the consumer’s behavioral intention towards adopting access-based services is mediated by the consumer’s perceived risk of failure. Therefore, the following hypotheses are proposed:
H3a: The product value has a negative effect on the consumer’s perceived risk of failure.
H3b: The consumer’s perceived risk of failure has a negative effect on the consumer’s behavioral intention towards adopting access-based services.
In addition to this, Baumeister and Wangenheim (2014) suggest that the perceived risk also comprises the risk of non-availability as Lamberton & Rose (2012) found that the perceived risk of scarcity is a key determinant of the consumer’s sharing propensity. In their hypothesis development they argued that the demand of shared goods may be higher at certain times or locations, decreasing the chance of getting a hold of the good, thereby simultaneously decreasing the consumer’s willingness to participate in sharing systems. Although this is an important construct, a logical link to the good’s type of provider or its value—central in this study—is absent, making examination into this relation uncalled-for.
In contrast, the monetary risks associated with access-based consumption are considered in this study. Chen and Dubinsky (2003) extensively researched the perceived customer value in e-commerce, including a supported hypothesis proposing the positive effect of a product’s price on perceived risk. They acknowledged the diverse reflections of perceived risk, but opted for combining four types of perceived risk in one measure. Therefore, the significance of the role of financial risk in the effect of product price is not deducible, rather, a mere suspicion can be derived. Nevertheless, results indicate that the product price affects the consumer’s overall perceived risk. Baumeister and Wangenheim (2014) may have found a more direct indication of a possible relationship between the value of a good and the perceived financial risk. Their conclusions included that the predictability of costs is much less commonplace in access-based consumption than in ownership (Baumeister & Wangenheim, 2014). A possible explanation for this is that the total costs of access-based consumption are ultimately largely dependent on the individual usage intensity and potential future price fluctuations (Berry & Maricle, 1973). In addition, the predictability of costs in ownership differed between cars and bicycles, such that consumers found the cost predictability of bicycles 130.3 percent more predictable than those of cars (Baumeister & Wangenheim, 2014). This may
suggest that these consumers are uncertain in what to expect of the good financially, thereby possibly running a higher financial risk. This leads to the following hypothesis:
H4a: The product value has a positive effect on the consumer’s perceived financial risk.
Ultimately, this perceived financial risk is expected to play a mediating role in the relationship between the product value and the consumer’s behavioral intention towards adopting access-based services as well, as common sense suggests that consumers may have less incentive to engage in a transaction that carries financial risks. In fact, the concept perceived risk is by definition associated with negative effects, e.g., danger and loss (Chen & He, 2003). Chen and He (2003) define perceived financial risk as the probability that a purchase results in loss of money or other resources. Perceived financial risk has already been found to have a significant negative effect on adoption intention in adjacent topics (Chen & He, 2003; Featherman & Pavlou’s, 2003). As such, the following hypothesis is proposed:
H4b: The consumer’s perceived financial risk has a negative effect on the consumer’s behavioral intention towards adopting access-based services.
Product value and the consumer’s price consciousness
Besides the potential presence of indirect effects of product value on the consumer’s behavioral intentions towards adopting access-based services, there may exist direct effects as well. Lamberton & Rose (2012) speculated in their studies that the sharing propensity is likely to vary depending on the product being shared, but have not tested this hypothesis. This expectation seems solid as it relates to findings of Baumeister and Wangenheim (2014). They found that the monetary perception was the most important determinant for the consumer’s overall attitude towards using access-based offerings, meaning that the value of the product being shared greatly impacts the choice of access or ownership. Additionally, when the perceived total costs of a particular consumption mode are high, consumers tend to favor this consumption mode less (Baumeister & Wangenheim, 2014). Numerous studies argue that consumers are attracted to access-based consumption because it allows them to enjoy particular products without having the funds to own them privately (Bardhi & Eckhardt, 2012; Botsman & Rogers, 2010; Moeller & Wittkowski, 2012). From this, it may be reasonable to propose that one may be more likely to opt for accessing when a product is more expensive then when it is less expensive. This relates to the need of some consumers preferring premium and up-to-date products (Schor, 2014), inherently assorted in a higher price segment. The novelty of such products drives both temporary ownership and non-ownership, i.e., access-based consumption (Moeller & Wittkowski, 2012). Accordingly, the following hypothesis is proposed:
H5: The product value has a positive effect on the consumer’s behavioral intentions towards adopting access-based services.
Lamberton and Rose (2012), Tussyadiah (2015), and Botsman and Rogers (2010) found that the cost benefit associated with sharing is a primary driver of adopting access-based services among prospective and existing consumers. Hamari et al. (2015) classified these economic benefits as an extrinsic motivator and found it to be a significant predictor of continuous use intentions. This implies
that consumers that intrinsically value economic benefits more, may have more incentive to engage in sharing systems, given that the economic benefits are specified. Moeller and Wittkowski (2010) note that the extent to which this incentive is provided is dependent on diverse factors, such as income and perceptions and knowledge on price acceptability across product categories. This price consciousness therefore is a trait that varies per individual, but also per situation and product (Lichtenstein, Block & Black, 1988). Price consciousness imparts a sensitivity to the price one pays for a good or service, with a tendency to be less willing to pay a higher price for a product (Sinha & Batra, 1999). Lichtenstein et al. (1988) define price consciousness as “the degree to which the consumer uses price in its negative role as a decision-making criterion” (p. 245). In a study in the context of access-based consumption, hypotheses involving a direct relationship between price consciousness and behavioral intention were rejected after analysis (Moeller & Wittkowski, 2010). However, the current study hypothesizes price consciousness to merely fulfill a moderating role in the relationship between the value of an accessible good and the consumer’s behavioral intentions towards adopting access-based services. This builds on the previously mentioned observation that price consciousness may vary per consumer and product, demonstrating the plausibility of this hypothesis. The final hypothesis is formulated as follows:
H6: The relationship between the product value and the consumer’s behavioral intention towards adopting access-based services is moderated by the price consciousness of the consumer, such that this relationship is stronger for high values of price consciousness.
The proposed hypotheses are visualized in the conceptual model below.
!
Methodology Experimental design
In order to test the hypotheses, an online experimental survey was conducted with four conditions as the independent variables (1) type of provider and (2) product value were manipulated into two levels. Hence, participants were randomly assigned to one of the conditions of a 2 (product value: low vs. high) × 2 (type of provider: business-to-consumer vs. peer-to-peer) between-participants design as shown in Table 2.
Excluding public transport—bicycles, scooters, bikes, vans, and cars are the most common means of transportation in the Netherlands (Centraal Bureau voor de Statistiek [CBS], 2015). Overall, the most widely adopted means are bicycles and cars. Also, these are the transportation means that vary the most with regard to the general price level of ownership. Therefore, when the independent variable was manipulated, a bicycle represented the low product value condition, whereas a car represented the high product value condition. In addition, the independent variable type of provider was manipulated. Either the company had full ownership of the good, implying that the participants are to adopt a business-to-consumer service (e.g., Citi Bike), or the participants were exposed to the peer-to-peer service condition, in which they were offered to access goods owned by other customers of the company (e.g., RelayRides).
Table 2. Experimental design and participant distribution.
Sample and procedure
In total, 130 participants started the online survey. However, 22 participants did not fully complete the survey, yielding a total of 108 usable responses. The average age in the obtained sample was 24.15 years (SD = 4.24). There was a substantial skewness with regard to the participants’ gender as 71.8 percent were female. Also, the large majority of the participants, 84.5 percent, were highly educated at a university.
The target population of this study is Dutch consumers. The sample was collected by convenience sampling among students and acquaintances. As the participants were primarily approached through social media and Facebook pools, it is not possible to determine the response rate. E.g., it is impossible to determine how many people have been exposed to the post in which it is promoted to fill in the survey. Ultimately, 108 useful responses have been recorded, resulting in circa 27 responses per treatment.
Product value Low High Type of provider Business-to-consumer N=26 N=29 Peer-to-peer N=29 N=24
As participants commenced the survey, they were asked to grant permission to use their data anonymously for research purposes. Next, they were introduced to the topic of access-based consumption and were asked to assume to be searching for and interested in adopting an access-based service on transportation. Subsequently, the stimulus was shown and the questions related to their feelings on this stimulus were asked. Then the questions regarding the control variables such as demographic information were posed. Lastly, the participants were sincerely thanked for their participation and a contact address was provided in case they had any questions or remarks. The survey in its entirety that was used in this study is presented in Appendix A. The average response time (excluding extreme outliers) amounted to 4.9 minutes.
Stimuli
In developing the stimuli for this study, the situation described in the survey was visualized: the participant was to assume being interested in adopting an access-based service on transportation. Hence, the stimulus comprised a fictive advertisement of fictive company SmartRide offering access to a transportation product. Appendix B can be consulted to view the stimuli developed for this study. In accordance with the experimental design proposed in the previous sections, the participants were exposed to a condition in which the product value was low or high, and the type of provider was either the company or a peer. These manipulations manifested themselves in the use of images (e.g., photos of a bicycle or car), text (e.g., “company-owned” or “[…] from your peers”) and a statement of the good’s value (e.g., “valued at €375”) to maximize understanding of the scenario and create a realistic environment. The latter benefits the external validity of the experiment as the stimulus ought to be similar to the stimulus that would occur in the real world. For instance, the lay-out of the stimuli mimics a web page. Also, in condition 3 and 4, quotation marks were added to depict the advertisement as an actual advertisement of a peer. Overall, these manipulations operate controversially. On the one hand, the stimuli have to appear realistic, while, on the other hand, the stimuli ought to eliminate disturbances (Lynch, 1982). The lay-out, fictive company logo and name, use of color, and additional information have been kept constant to minimize further variety. In doing so, it is attempted to incorporate the two requirements for a proper experiment in a balanced manner.
With regard to the statements of the product value of the products chosen in this study, the average current purchase price of both a Dutch bicycle and car were considered. According to Rabobank (2016), consumers pay an average price of €375 on a bicycle. This number excludes the purchases of e-bikes, which are a significantly larger investment than regular bikes (Rabobank, 2016). Stichting BOVAG-RAI Mobiliteit (2016) presented in an annual article that Dutch fuel cars have an average purchase price of €25,000. Accordingly, these numbers were used in the stimuli.
Measures
Perceived trustworthiness
Ohanian’s (1990) five items on trustworthiness were used to measure the participants’ perceived trustworthiness in SmartRide. It is a very reliable measure (⍺ = .825). An example item is framed as follows: “SmartRide is…,” to which the participants are to indicate how undependable or dependable they perceive SmartRide to be on a seven-point Likert scale.
Perceived risk of failure
Stone and Grønhaug’s (1993) measure on performance risk was used to assess the participants’ perceived risk of failure. Stone and Grønhaug’s (1993) measure has three items and was validated in this study with Chronbach’s ⍺ = .884. Responses to all items were recorded on a seven-point Likert scale (1 = “strongly disagree”; 7 = “strongly agree”). Only slight adaptations to the items have been made to fit the context of this study. An example item is: “As I consider using this service, I worry about whether the product that I use will really perform as it is supposed to.”
Perceived financial risk
Perceived financial risk was assessed using Featherman and Pavlou's (2003) four items on financial risk. Substantial adaptations to the items have been made. The nominal scales that the items were recorded on, originally differed per item. This has been changed to be consistent with the other scales used in the survey. Also, the framing of the original scale was too specific and has been framed more generally. An example item is: “Use of this service is financially risky.” The first item contains a negation: “Use of this service will not lead to unforeseen costs.” The exact adaptations can be found in Appendix C, in which all original and adapted measures are presented. The adapted scale is a sufficiently reliable measure (⍺ = .730). The corrected item-total correlations indicate that all the items have a good correlation with the total score of the scale (all above .30), however, the first item only just meets this requirement (.371). Additionally, deleting this item would substantially affect the scale’s reliability.
Behavioral intention
To measure the participants’s behavioral intention towards adopting access-based consumption, Bhattacherjee’s (2002) measure on willingness to transact was used. This measure has three items and was highly reliable (⍺ = .925). It uses a seven-point Likert scale ranging from “strongly disagree” to “strongly agree.” An example item used in this study is “I am likely to utilize the services provided by SmartRide.”
Price consciousness
The participants’ price consciousness was assessed using Sinha and Batra’s (1999) measure on price consciousness. Their measure was used to measure the respondents’ price consciousness across eight different product categories. The measure has four items and was validated with Chronbach’s ⍺ = .866 in this study. All items were on a seven-point Likert scale anchored by 1 = “strongly disagree” and 7 = “strongly agree.” An example item is: “When it comes to adopting an access-based service, I rely heavily on price.”
On all measures, small adaptations to their framing have been made, e.g., Sinha and Batra’s (1999) measure on price consciousness framed the recordings on “disagree strongly” and Stone and Grønhaug’s (1993) measure on perceived risk of failure was framed as “extremely disagree.” This has both been changed to the more common “strongly disagree” format. On the contrary, more substantial changes have been made as well. Featherman and Pavlou’s (2003) measure on perceived financial risk had to be thoroughly adapted to fit the context of this particular study. For example, “What are the chances that you stand to lose money if you use the service?” from the original scale was changed to “Use of this service will not lead to unforeseen costs (reverse-coded).” This was to ensure that the response formats were consistent (i.e., Likert scale) across the measure’s items. Also, as it was expected that the majority of the participants were Dutch, some items were framed in a simplified manner to refrain people from abandoning the survey due to long and difficult sentences. Appendix C can be consulted to view all adaptations to the items.
Control variables
To rule out potentially deceptive relations, the age, gender, and education level was controlled for. These control variables are commonly used and were asked right before the end of the survey. In addition, the participants were asked if they had any prior experience with access-based services such as Citi Bike, Airbnb, Werkspot, or Car2Go, and whether they owned a car or bike themselves. Ownership of the transportation mean used in this study potentially influences the participant’s behavioral intention towards adopting access-based services. As does any (positive or negative) prior experience with adopting based services. Logically, participants that have experienced access-based services negatively in the past, are likely to hold an adverse attitude towards adopting them. Survey pre-test
A small pre-test among four Dutch participants (µ = 31.52 years, SD = 17.14), of which its demographics was attempted to be representative of the final survey sample, indicated some room for improvement. First, the framing of the items regarding the perceived risk of failure were suggested to be simplified as the participant had to read the sentence several times to understand it. Also, the stimulus description formerly did not contain the request to study the advertisement carefully, which resulted in several participants to give it a mere glance and having to go back to view the advertisement again. Lastly, prior to this pre-test, the term access-based consumption was first mentioned on the first page, yet explained on the second page. Participants felt that it was necessary
to explain the term on the first page instead. Accordingly, the explanation was relocated to the first page.
Statistical procedure
Prior to the analysis, the data had been checked for and eliminated from incomplete responses or responses from which their reliability was questioned due to, for instance, an unusually short response duration. However, there is not opted for complete listwise deletion as this reduces the effective sample size and power, thereby introducing biases in the estimations. Rather, for analyses that allowed pairwise deletion (e.g., correlations and regressions), this option was employed to maximize the usability of the data that was obtained during collection. The disadvantage of pairwise deletion is that it may introduce the same biases as listwise deletion, as well as resulting in mathematically inconsistent results or manipulated variation. However, because this only applies to 4 responses, it was determined that the pros outweighed the cons.
The survey was developed in online survey tool Qualtrics and analyzed with SPSS Statistics. Dummy variables were created for the two independent variables (i.e., type of provider and product value). The one counter-indicative item was recoded and all measures were checked for their reliability. Due to one curious outcome in the reliability analyses, a principal axis factor analysis was performed to assess the goodness of the scale. One requirement for performing such analyses is normality of the variables. Hence, normality checks were conducted for all dependent, mediator, and moderator variables. Its results can be viewed in Table 3. The rule of thumb with regard to the kurtosis and skewness tests indicates that all values between -1 and 1 are acceptable. Results of the variables complied with this rule of thumb.
Table 3. Skewness and kurtosis of dependent, mediator, and moderator variables (N = 104).
Although the kurtosis and skewness tests computed acceptable outcomes, Kolmogorov-Smirnov and Shapiro-Wilk tests were performed for an additional examination of the distributions. As almost all p-values are < .05, the variables are not normally distributed. The distribution of price consciousness is the only exception (see Table 4). Closer examination of the distributions by means of Q-Q plots revealed some outliers and it was considered to delete these from the data. However, some extreme values in a data set can be also expected under normal conditions. Also, the small sample size (N =
Skewness SE Kurtosis SE
Behavioral intention -.992 .237 .737 .469
Perceived trustworthiness -.049 .237 .223 .469
Perceived risk of failure -.094 .237 -.995 .469
Perceived financial risk .283 .237 -.604 .469
108) does not allow for much more manipulation or deletion of the data as this would result in an even smaller sample size, potentially harming the statistical design and accuracy of the results even more than performing tests on non-normally distributed variables.
Table 4. Kolmogorov-Smirnov and Shapiro-Wilk tests of dependent, mediator, and moderator variables (N = 104).
*. This is a lower bound of the true significance.
Then, correlation tests were performed to get a feel of the data at hand. In order to test the effect of the manipulated independent variables on the dependent variable, thus, whether a direct effect between the type of provider and product value, and the consumer’s behavioral intention can be ascertained (H2 and H5), a factorial ANOVA was conducted. The mediating roles of perceived trustworthiness, risk of failure, and financial risk (H1, H3, and H4) were tested by means of Hayes’ PROCESS model 4. Hayes’ (2013) approach offers more statistical power and makes fewer assumptions in comparison to Baron and Kenny’s (1986) more classic causal steps approach. Lastly, the moderating effect of price consciousness (H6) was also tested by means of Hayes’ PROCESS model 1.
Results Factor analysis
Although the reliability of all scales were satisfactory, the scale measuring the consumer’s perceived financial risk needed further investigation as the assumption (> .300) for corrected item-total correlation (.371) was barely met. A small value provides evidence that the item might not measure the same construct measured by the other items included. It may also suggest that there is some similarity between the variables. Hence, a principal axis factor analysis (PAF) was performed to evaluate the goodness of the scale. The items of the perceived financial risk and risk of failure constructs were selected as variables. This was due to the fact that either of these refer to the consumer’s perceived risk, which is the consumer’s perception of the uncertainty and collateral adverse consequences of using a service (Chen & Dubinsky, 2003). This may indicate that it is highly
Kolmogorov-Smirnov Shapiro-Wilk Statistic df p Statistic df p Behavioral intention .159 104 .000 .913 104 .000 Perceived trustworthiness .103 104 .008 .981 104 .146 Perceived risk of failure .089 104 .041 .962 104 .005 Perceived financial risk .107 104 .005 .975 104 .043 Price consciousness .073 104 .200* .976 104 .058
likely that the variable similarity can be found between these two variables. The Kaiser-Meyer-Olkin measure verified the sampling adequacy for the analysis, KMO = .802. Also, Bartlett’s test of sphericity, 𝝌2 (21) = 321.471, p = .000, indicated that correlations between items were sufficiently
large for conducting a PAF. Two factors were retained and rotated with an Oblimin with Kaiser normalization rotation. The two factors cumulatively explained 57.85% of the variance. Table 5 shows the factor loadings after rotation. Results indeed identified a problematic item. Namely, the principal loading of the second item of the perceived financial risk (“The provider of the product may be a fraud, jeopardizing my monetary investment”) shows high cross-loadings on the factor of perceived risk of failure as well. This could be due to the content of the item. The original item also referred to fraudulent practices that could impact one’s perceived financial risk: “Using an Internet bill-payment service subjects your checking account to potential fraud.” The original scale including this item was validated satisfactorily and although the existence of fraudulent parties would indeed indicate a higher perceived financial risk, a high perceived financial risk does not necessarily mean that there is a risk that a fraudulent party is involved. Accordingly, the adapted scale used in this study also shows some inconsistency among its items. However, as the aim is to retain as much data as possible and taking into account that this problematic item is an imperative risk commonly found in access-based consumption (e.g., PwC, 2015; Ufford, 2015), it is determined that the item should not be omitted for further analyses. Nonetheless, it is advised for future research that the perceived financial risk is segregated from the perceived risk of financial fraud.
Table 5. Pattern matrix of the PAF.
Note. Factor loadings over .300 appear in bold.
Rotated Factor Loadings
Item PFR PRoF
As I consider using this service, I worry about whether the product
will really perform as it is supposed to -.003 .769
If I were to use this service, I become concerned that the product
will not provide the level of benefits that I would expect .181 .770
The thought of using this service causes me to be concerned for how
really dependable and reliable the product will be -.088 .963
The use of this service will not lead to unforeseen costs .355 .097
The provider of the product may be a fraud, jeopardizing my
monetary investment .415 .314
Use of this service would lead to a financial loss for me .844 -.074
Use of this service is financially risky .801 -.048
Eigenvalues 1.16 3.60
Correlations
Table 6. Means, standard deviations, and correlations of variables (N = 108).
**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).
Prior to testing the hypotheses, the correlations between constructs were examined to get an initial idea of the data and potential outcomes of this study (see Table 6). In summary, apart from a significant correlation between the product value and one’s prior experience with access-based consumption, neither of the independent variables had any correlations with any of the mediating, moderating or dependent variables (respectively comprising the perceived trustworthiness, risk of failure, financial risk, price consciousness, and behavioral intention). This might suggest that there merely are weak mediating, moderating, and direct effects, or they are non-existent. Nonetheless, in agreement with the extant literature, the mediating and dependent variables were all significantly correlated with each other. In addition, price consciousness, hypothesized to moderate the relationship between the product value and the consumer’s behavioral intention, was significantly correlated with the latter, but not the former.
Hypothesis testing
Direct effects of the type of provider and product value on behavioral intention
It was expected that the type of provider in an access-based service would affect the consumer’s behavioral intention such that consumers tend to engage more in peer-to-peer activities than business-to-consumer activities (H2). Also, the postulation was made that the product value has a positive effect on the consumer’s behavioral intention towards adopting access-based services (H5). To test H2 and H5, a univariate analysis of variance (ANOVA) was conducted. The Levene test was used to confirm the assumption that variances are equal across the conditions, p =.660 (denoted in Table 7).
The results of the ANOVA are shown in Table 8. There was no significant direct effect of the type of provider on the consumer’s behavioral intention, F(1, 104) = .058, p = .810, 𝜂2 = 001. Results
also did not reveal any direct significant effect of the product value on the consumer’s behavioral intention, F(1, 104) = .018, p = .895, 𝜂2 = .000. If interaction effects were present, then further
investigation in H4 and H5 could be done. The lines in the profile plots described that the lines were not parallel (e.g., see Figure 2), suggesting the possible existence of interaction. However, the tests did not confirm this suspicion as there was a non-significant interaction effect between the type of provider and the product value on the consumer’s behavioral intention F(1, 104) = 1.470, p = .228, 𝜂2 = .014. Therefore, H2 and H5 were rejected.
Table 7. Levene’s test of equality of error in variance on behavioral intention (N = 104).
F df1 df2 p
Table 8. Factorial ANOVA (N = 108).
Note. Significant at the p < .05 level.
Figure 2. Profile plot of the estimated marginal means of behavioral intention.
Mediating effects of the perceived trustworthiness, risk of failure, and financial risk
To test whether the consumer’s perceived trustworthiness in the company mediated the relationship between the type of provider and the consumer’s behavioral intention (H1), Hayes’ PROCESS model 4 for mediation analyses was performed. The proposed mediation model is visualized in Figure 3. Results (see Table 9 and 10) revealed that the a1 effect was non-significant, a1 = .114, p = .461 (H1a),
meaning that it had no effect on the consumer’s perceived trustworthiness whether the product in access-based services is provided by the business or a peer. Nonetheless, the consumers that perceive the company as trustworthy report higher intentions of adopting the service (b1 = .574). This effect
was statistically significantly different from zero, t = 3.745, p = .000 (H1b). The direct c1’ effect
(c1’ = -.004, p = .985) was smaller than the total c1 effect (c1 = .061, p = .804) and was not significant,
meaning that, in agreement with the hypothesis, the consumer’s perceived trustworthiness could have acted as a mediator. However, there is no support for this due to the non-significance of the a1 effect.
Because not all requirements for a mediation have been met, H1 was rejected. Nonetheless, the model explained 21.57% of the variance of the consumer’s behavioral intention, which was statistically significant (R2 = .216, p = .003).
Sum of
squares df squareMean F 𝜂2 p
Type of provider .094 1 .094 .058 .001 .810
Product value .029 1 .029 .018 .000 .895
Type of provider * Product
value 2.388 1 2.388 1.470 .014 .228 Error 168.911 104 1.624 Total 2346.444 108 4,30 4,40 4,50 4,60 4,70 Low IV High IV Be h av io ra l i n te n ti o n Type of provider Low car High car Peer-to-peer Business-to-peer Car Bike
Figure 3. Proposed mediation model with perceived trustworthiness as mediator (*. p < .05). Table 9. Model summary of direct effects.
Table 10. Model summary of total effect.
The preceding procedure was repeated for the hypothesized mediating effect of the perceived risk of failure on the relationship between the product value and the consumer’s behavioral intention (H3), as well as on the mediating effect of the perceived financial risk on this relationship (H4). The proposed mediation model for H3 is visualized in Figure 4. As this figure reports, none of the coefficients are accompanied by an asterisk, indicating the lack of any significant effects in this model. Namely, the outcomes of the mediation analysis revealed that the a2 effect (H3a) was non-significant (a2 = .280,
p = .306), suggesting that the product value had no effect on the consumer’s perceived risk of failure of the accessed product (see Table 11 and 12). Neither did the product value and consumer’s perceived risk of failure (H3b) lead to a difference in one’s behavioral intention to use the access-based service (b2 = -.170, p = .067 and c2’ = .096, p = .696). Also, no total effect could be found
(c2 = .048, p = .845), meaning that the results do not provide evidence for the hypothesis that the
relationship between the value of the accessible good and the consumer’s behavioral intention towards adopting access-based services is mediated by its perceived risk of failure. Hence, H3 was rejected.
Consequent
Perceived trustworthiness (M) Behavioral intention (Y)
Antecedent Coeff. SE p Coeff. SE p
Type of provider (X) a1 .114 .153 .461 c1’ -.004 .230 .985 Perceived trustworthiness (M) — — — b1 .574 .153 .000 Constant i1 4.481 .680 R2 = .047 .000 i2 .908 1.226 R2 = .216 .461 F(7, 95) = .672, p = .695 F(8, 94) = 3.234, p = .003
Effect SE p LLCI ULCI
Direct effect c1’ -.004 .230 .985 — —
Total effect c1 .061 .244 .804 — —
Boot SE Boot LLCI Boot ULCI
Figure 4. Proposed mediation model with perceived risk of failure as mediator (*. p < .05). Table 11. Model summary of direct effects.
Table 12. Model summary of total effect.
Lastly, the mediating effect of the perceived financial risk on the relationship between the product value and the consumer’s behavioral intention (H4) was also tested my means of Hayes’ PROCESS model 4. The proposed mediation model for this hypothesis is visualized in Figure 5. From the results denoted in Table 13 and 14, it can be concluded that there is no significant effect of the product value on the consumer’s perceived financial risk (a3 = .325, p = .105). Nor could a direct c3’ effect be found
(c3’ = .162, p = .503). In addition, there was no indication of a total effect (c3) due to its
non-significance (c3 = .048, p = .845). This finding is in agreement with the total c2 effect in the previous
hypothesis test (c2 = .048). The direct effect (c3’ = .162) was not smaller than the total effect
(c3 = .048), meaning that the consumer’s perceived financial risk, in contrast to the initial
expectations, does not fulfill a mediating role in the relationship between the value of the accessible good and the consumer’s behavioral intention. Accordingly, H4 was rejected. Nevertheless, the
Consequent
Perceived risk of failure (M) Behavioral intention (Y)
Antecedent Coeff. SE p Coeff. SE p
Product value (X) a2 .280 .272 .306 c2’ .096 .245 .696 Perceived risk of failure (M) — — — b2 -.170 .092 .067 Constant i3 5.405 1.195 R2 = .070 .000 i4 4.396 1.180 R2 = .130 .000 F(7, 95) = 1.015, p = .426 F(8, 94) = 1.761, p = .095
Effect SE p LLCI ULCI
Direct effect c2’ .096 .245 .696 — —
Total effect c2 .048 .247 .845 — —
Boot SE Boot LLCI Boot ULCI
statistically significant effect b3 = -.350 (p = .006) indicates that consumers with a high perception of
financial risk involved with using an access-based service are less inclined to adopt this service (H4b). However, the indirect effect of a2b2 = -.114 means that two consumers of which one considers
accessing a bike and the second considers accessing a car, are estimated to differ .114 units in their reported behavioral intention as a result of an increased perception of being at financial risk when using the service offered. The indirect effect is negative, meaning that those consumers that consider accessing a product with a higher general price level of ownership (i.e., the car) are estimated to be lower in their intention to adopt the service. This effect is significant as the 95% bootstrap confidence interval does not contain zero (-.348 to -.002).
Figure 5. Proposed mediation model with perceived financial risk as mediator (*. p < .05). Table 13. Model summary of direct effects.
Table 14. Model summary of total effect.
Consequent
Perceived financial risk (M) Behavioral intention (Y)
Antecedent Coeff. SE p Coeff. SE p
Product value (X) a3 .325 .198 .105 c3’ .162 .241 .503 Perceived financial risk (M) — — — b3 -.350 .123 .006 Constant i5 4.334 .871 R2 = .051 .000 i6 4.994 1.174 R2 = .170 .000 F(7, 95) = .735, p = .643 F(8, 94) = 2.407, p = .021
Effect SE p LLCI ULCI
Direct effect c3’ .162 .241 .503 — —
Total effect c3 .048 .247 .845 — —
Boot SE Boot LLCI Boot ULCI
Moderating effect of price consciousness
The last hypothesis comprised the expectation that the relationship between the value of a good and the consumer’s behavioral intention towards adopting access-based services is moderated by the price consciousness of the consumer, such that this relationship is stronger for high values of price consciousness (H6). Hayes’ PROCESS model 1 was performed to test this hypothesis, its results are denoted in Table 15. Prior to running this test, the dependent and moderator variable were standardized. R-squared quantifies the proportion of the total variance of the consumer’s behavioral intention (Y) explained by the overall model. This model explained 23.98% of the variance of the consumer’s behavioral intention, which was statistically significant (R2 = .240, p = .012). The output
of the moderation analysis indicated that no moderation effect is taking place (interaction coefficient = .095, p = .052). Hence, no further analysis can be done. From the results it can be concluded that there is no indication that the relationship between product value and the consumer’s behavioral intention towards adopting access-based services is stronger for high values of price consciousness, confirming that H5 ought to be rejected, as well as revealing that H6 ought to be rejected too.
Table 15. Model summary of interaction.
Additional exploration in the consumer’s perceived financial risk
As initial results indicated, the item that explored the participants’ view on fraudulent activities on peer-to-peer platforms revealed its inconsistency with the averaged behavior of the other items. It was suggested that this may have been due to the content of the item. This also lead to the suspicion that the distribution of the item in question was not normal and potentially included many outliers. To examine whether the low corrected item-total correlation can also be attributed to a deviating distribution, the means and standard deviations of the variable items were computed (see Table 16), and a simple boxplot of the variable was composed (see Figure 6). However, neither the means nor the boxplot suggest that the item (i.e., Item 2) is distributed differently than the other items. The distribution may be slightly skewed, but not notably more than the other items. The box and whiskers also do not indicate that the item’s distribution has a relatively high variability. Lastly, no unusual outliers could have been identified either.
Effect SE t p Intercept i1 3.054 .788 3.878 .000 Product value (X) c1 .299 .152 1.966 .052 Price consciousness (M) c2 -.254 1.012 -.251 .802 Product value * Price consciousness (XM) c3 .095 .194 .489 .626 R2 = .240, p = .012 F = (3, 100) = 5.341
Table 16. Means and standard deviations of the perceived financial risk items (N = 105).
Figure 6. Boxplot of the perceived financial risk items (N = 105). Discussion
General discussion
The emergence of a myriad of peer-to-peer platforms boosts the modes of consumption other than ownership. Sharing products may save people time and money, and allegedly contributes to sustainable consumerism. If access-based consumption is the future, then it is imperative that these businesses are aware of the consumer intentions with regard to accessing. This study aimed to broaden this understanding by examining the effect of the type of provider and product value on the consumer’s behavioral intention towards adopting access-based services. Data was obtained from an online experimental survey in which the participants were exposed to one fictive website screenshot of a company—SmartRide—offering transportation services. Either an advertisement involving a bicycle (i.e., low product value) or a car (i.e., high product value) was shown. In addition, the
Item Mean SD
1 Use of this service will not lead to unforeseen costs (counter-indicative) 3.895 1.330 2 The provider of the product may be a fraid, jeopardizing my monetary investment 3.870 1.323
3 Use of this service would lead to a financial loss for me 3.540 1.308
