Trust in Sharing Economy

0

TRUST IN SHARING ECONOMY

LINDSTRÖM, JIM

MORSHED, MONJUR AKM

School of Business, Society & Engineering

Course: Master Thesis in Business

Administration

Course code: FOA403

15 cr

Supervisor: Konstantin Lampou Date: 2020-06-08

Abstract

Date: 2020-06-08

Level: Master’s thesis in Business Administration, 15 cr

Institution: School of Business, Society and Engineering, Mälardalen University

Authors: Jim Lindström AKM Monjur Morshed (97/08/31) (93/12/23)

Title: Trust in Sharing Economy

Tutor: Konstantin Lampou

Keywords: Trust, sharing economy, e-commerce, service

Research questions: What different factors impact trust in sharing economy platforms and

sharing economy service providers?

Purpose: The study aims to explore which factors create trust towards the sharing

economy platforms and actors providing the service, and mainly the potential relationships between the trust factors

Method: The study was conducted with a quantitative approach. A survey was used for

data collection. The data gathered from the survey was analyzed using regression to test 7 predefined hypotheses

Conclusion: Trust towards the platform is influenced by two main factors, perceived security, and perceived risk. There are four factors of trust influencing the trust towards actors providing the service. These factors are personality-based trust, experience-based trust, cognitive-based trust, and lastly the trust towards the platform where the service provider operates.

Introduction 1

1.1 Research problem 2

1.2 Research question and purpose 3

2. Theory and concepts 4

2.1 Sharing economy 4 2.2 Trust 5 2.2.1 Personality-based trust 5 2.2.2 Experience-based trust 6 2.2.3 Cognitive-based trust 6 2.2.4 Perceived security 7 2.2.5 Perceived risk 7 2.2.6 Perceived quality 8 3. Hypotheses development 9

3.1 Trust between actors 9

3.2 Trust towards platform 10

3.3 Conceptual model 12

4. Methodology 13

4.1 Research approach 13

4.2 Data collection and measurements 15

4.3 Sample 16

4.4 Operationalization 17

4.5 Reliability and validity 18

4.6 Data analysis method 19

4.7 Research ethics 19

5. Analysis 19

6. Discussion 23

7. Conclusion 26

7.1 Limitations and further research 26

7.2 Theoretical implications 27

7.3 Practical implications 27

References 28

Appendix 1 - Correlation matrix 36

Appendix 2 - Regression H1, H2 and H3 37

Appendix 3 - Regression H4, H5 and H6 38

Acknowledgment

We would like to thank our supervisor Professor Konstantin Lampou for his invaluable advice and support throughout our work process. We would also like to extend our thanks to

our seminar group where we were helped developing our thesis with great suggestions from improvement.

Finally, we would like to thank everyone who participated in the survey. Without your responses the study would not have been possible.

1

1. Introduction

Traditionally speaking, a market is confined by multiple actors that all serve a certain purpose to their common environment. In the traditional view of markets, the two main actors would be businesses and consumers (Wieland, Koskela-Huotari & Vargo, 2016). However, in more recent years, this traditional view of a market and its actors has changed. Now markets are interpreted as environments where all social and economic actors are part of the value-creation process through service ecosystems (Brodie, Löbler & Feher, 2019). With this change in the way that market actors are perceived, several services are now based on taking advantage of the new market actors (Breidbach & Brodie, 2017). These new markets also distance themselves from a traditional goods-centric view of a market and are not based purely on services. A goods-centric view has at its core that value is included in tangible, business created, goods that are later consumed by consumers. (Brodie, Löbler & Feher, 2019) Through the new perspective on actors on a market, the actors are part of the value creation process, rather than the value being purely business manufactured (Breidbach & Brodie, 2017). The actor-to-actor approach to a market can be seen in practice in many different ways (Wieland et al., 2016) and one of the more recent examples is through terms like collaborative consumption (Hamiri, Sjöklint & Ukkonen, 2016) or sharing economy (Cusumano, 2014).

The phenomenon of the sharing economy is relatively new to markets across the world. The concept behind the sharing economy is that consumers, rather than only interacting with businesses when purchasing service, consumers can interact with peers that provide the service. The sharing economy can be defined as a way to decrease the consumption of resources through effective capacity utilization mainly through sharing access to goods and services. (Nationalencyklopedin, 2020) The key aspect of the concept behind the sharing economy is that firms have a way of connecting other interdependent economic actors in different market contexts (Breidbach & Brodie, 2017). Airbnb is the largest sharing economy platform in the world with five million hosts in 191 countries and providing almost three hundred million travelers all over the countries in the hospitality business field. The platform works in a way that people can rent out their property to travelers as a substitute for staying at a hotel. Airbnb serves the function of being a network for connecting people that rent out their property with people that are looking for accommodation. (Airbnb, 2020) The concept behind Airbnb is relatively new and the fact that different actors of the Airbnb network are strangers to each other could create issues regarding trust (Mao et al., 2020) However, researchers oppose the statement that the sharing economy is a completely new phenomenon. Frenken and Schor (2019) claim that sharing resources has been part of human culture and has historically been common in many parts of the world. What is new about the sharing economy, not the business and market-oriented perspective, is that consumers now share resources with strangers (Frenken & Schor, 2019; Schor, 2016). The concept of sharing economy can be seen today in several different market contexts, where businesses are based around the interdependencies amongst market actors. This is seen in the context of accommodation (Airbnb), transportation

(Uber), finance (LendingClub), and more (Breidbach & Brodie, 2017). These businesses act as a platform that acts as a way of communication and connection between the economic actors on the market, where the actors can either purchase or provide the services (Frenken & Schor, 2019).

1.1 Research problem

What makes the sharing economy an interesting topic of research is that it provides changes as to how a market and its economic actors are perceived (Frenken & Schor, 2019; Schor, 2016) and also has a great impact on traditional business contexts (Breidbach & Brodie, 2019). There is also a diverse perception, in the research field, on the advantages and disadvantages of a sharing economy approach might bring to markets and business (Frenken & Schor, 2019) The emergence of business models based on the concept of sharing economy has a disruptive effect on certain industries, particularly in accommodation and transportation industries, where firms like Airbnb and Uber are becoming more prominent (Breidbach & Brodie, 2019). The rapid development of the concept of sharing economy in different contexts is made possible by the fact that market actors today are more engaged in technological service ecosystems and used to the e-commerce side of business (Breidbach & Brodie, 2019; Cheng, Fu, Sun, Bilgihan & Okumus, 2019). However, because sharing economic business is mostly supported by online platforms (ter Huurne, Ronteltap, Corten & Buskens, 2017) and the fact that the actors share services with strangers, there is often an issue regarding trust (Freken & Schor, 2019; Cheng et al., 2019)

Trust has always been important in traditional business and is considered a competitive advantage and it is therefore a critical point for the companies (Barney & Hansen, 1994). However, it is also crucial in a sharing economy setting (ter Huurne et al., 2017). According to PwC (2015), which is a market analysis consulting firm, 72% of customers in the U.S. are concerned about the trust factors connected to sharing economy platforms and service providers. Consumer trust issues affect the purchase intention of the consumers. The problem is finding a solution to how trust can be managed in a sharing economy setting (Kim, Ferrin & Rao, 2009). Ter Huurne et al. (2017) mention that there are several potential dangers when sharing goods and services between actors, such as the risk of damages to property or issues regarding privacy. The uncertainty is increased by the fact that the actors often are strangers to each other and only rely on the online sharing economy platform (Frenken & Schor, 2019). This makes trust in the context of the sharing economy a crucial point of interest for business development in the field (Cheng et al., 2019; ter Huurne et al., 2017).

In the sharing economy phenomenon, trust has previously been examined to understand how consumer trust evolves the users, providers, and suppliers in the real-life industry Laurell & Sandström, 2017). Research has identified several different aspects of sharing economy businesses that play part in the development of trust between the market actors (Cheng et al., 2019) Some studies have had a focus on how the online platform can create security and trust (Hawlitschek, Teubner, Weinhardt, 2016; Mao, Jones, Li, Wei & Lyu, 2020). Some studies have focused on the perceived trust in the market actors providing the service (Ert, Fleischer

& Magen 2016; Mao et al., 2020; Saksanian, Martínez-Fiestas & Timana, 2020) However, this is an emerging field of research, experts and academic researchers consider this phenomenon as a creative aspect of the transaction (Osztovits et al., 2015; Laurell & Sandström, 2017). The user perception about the sharing economy is still underdeveloped and researchers are still discussing the challenges across the trust-building process (Gonzalez-Padron, 2017). The main issue with existing research in this field is understanding how the different aspects of trust, in the context of sharing economy, are dependent on each other, since trust is thought to be more complex than previously examined (ter Huurne et al., 2017) The existing literature has focused on particular sections of trust in sharing economy, for example at trust towards the platform (Hawlitschek et al., 2016) but very few studies have analyzed if there is an interdependent connection between the different aspects of trust that are connected to the business in a sharing economy market. (ter Huurne et al., 2017) Factors that are shown to influence trust in a sharing economy setting is personality, previous experience with the service (Mao, et al., 2020) cognitive reasoning (ter Huurne, et al., 2017; Mao, et al., 2020), perceived risk, quality and security (ter Huurne et al., 2017). These factors have been previously tested and indicated an impact on trust in a sharing economy setting (Mao et al., 2020). However, the impact of the trust in the sharing economy platform on the overall trust in the service provider (host) is relatively unexplored (ter Huurne et al., 2017). This relation between the trust in a sharing economy platform and the trust towards the service provider (host) was recently tested, but the results were not significant and were therefore suggested for further research (Mao et al., 2020)

1.2 Research question and purpose

With the basis in the research problem, the question that is to be answered in this study is: What

different factors impact trust in sharing economy platforms and sharing economy service providers?

The purpose of this study is to further develop and test existing theories and research results regarding trust factors in a sharing economy market setting. The study aims to explore which factors create trust towards the sharing economy platforms and actors providing the service, and mainly the potential relationships between the trust factors. The study will be conducted quantitatively, and the data will be collected using a questionnaire. The question items used in the study will be inspired by previously tested question items of previous research. This allows for a comparison between current and previous studies.

2. Theory and concepts

The main concept of the study that needs to be defined and described using theory is trust. This chapter will give insight into how researchers describe trust and different influencing factors that might develop trust. The concept is quite complex and is based on psychology, therefore the factors found in the literature review will be the main points of interest. Several previous studies have explored the research area of trust in a sharing economical environment. A literature review was conducted to develop a common ground for the construct of trust concerning the sharing economy. The literature is divided into two parts, one that is focused on trust between the market actors and one that is focused on trust in the sharing economy platform. Previous studies have found that trust can be influenced by multiple different factors such as personality, experience, cognition (Mao et al., 2020), perceived security, perceived risk and perceived quality (ter Huurne et al., 2017).

2.1 Sharing economy

The sharing economy firms have rapid global expansion (Cannon and Summers, 2014), multiple researchers define sharing economy with basic characteristics but they did not come up with a common definition (Anwar, 2018; Key, 2017; Laurell and Sandström, 2017; Osztovits et al., 2015). According to Laurell and Sandström (2017), the sharing economy can be defined as: “Information and communication technology-enabled platforms for exchanges of goods and services drawing on non-market logics such as sharing, lending, gifting and swapping as well as market logics such as renting and selling” (p. 63).

The sharing economy is rooted in the old concept 'collaborative consumption' (Felson & Spaeth, 1978) of engaging in joint activities like sharing goods with others, for example, friends, family, or strangers (Henten and Windekilde, 2016; Laurell and Sandström, 2017). Even though the concept has existed historically it has changed and is now recognized as a new idea in the modern business world. Belk (2014) considers the sharing economy concept as 'phenomenon born of the internet age' and a digital footprint towards the post ownership economy. The digital platform creates interactions between service providers and users, where they can share information and communicate with each other. This digital era gave the sharing economy a momentum, where a two-sided market was created with profit-sharing (Laurell and Sandström, 2017). The sharing economy firms need the involvement of at least three parties like one party is offering a resource for sell, rent or co-use they act like private provider, another party is seeking the offered product or service and finally, the digital platform who manage the supply and demand by facilitates two-sided marketplaces (Hawlitschek et al., 2018; Anwar, 2018). The digital platform is owned and operated by a business that works as a mediator of matching people looking to purchase with other people providing a service (Frenken & Schor, 2019). In the case of Airbnb, the platforms act as a way where people can rent out their homes to people looking for accommodation (Breidbach & Brodie, 2019). In the sharing economy platform providers are sharing the underutilized assets with strangers and they have to trust the strangers, but they also have to assess the potential risk. There is a possibility of theft or different kind of damages that might occur when interacting with strangers. (Bellin, 2017) So

many researchers also discussed mutual trust between customers and the providers (Bellin, 2017; ter Huurne et al. 2019)

2.2 Trust

Trust has been defined by different disciplines in the area of research. There is still uncertainty on how trust should be defined, but there are parts of the concept that researchers can agree upon. Trust always has a subject (a person that creates trust) and an object (the person that manages and interprets trust) and they are interdependent. (PytlikZilling & Kimbrough, 2016). In a traditional business model, customers' trust considers as a competitive advantage for the companies, and its a critical point for the companies also because most of the transactions were based on physical activities (Barney and Hansen, 1994). In the sharing economy, it's almost the same as the traditional business model. In the US market, 72% of customers are concerned about trust in the sharing economy (PwC, 2015). Here consumers need to create their trust on online platforms and with strangers, so consumer purchase intention can be affected by the trust issues of the customers. The problem is how to manage the trust in the sharing economy and many researchers say that perceived risks, perceived benefits, and trust influencers or actors may drive the customers' purchase intention (Kim et al., 2009). Customer loyalty is directly connected with trust, company reputation, and customer retention (Delgado-Ballester et al., 2003; Erdem and Swait 2004; Matzler et al., 2008; Willmott 2003). A high level of trust in online platforms means that consumers have a lower risk towards the service provider (Kim et al., 2009). According to Mayer, Davis and Schoorman (1995) trust can be defined as: “The willingness of a party to be vulnerable to the actions of another party based on the expectation that the other will perform a particular action important to the trustor.” (Mayer, et al., 1995) This definition by Mayer et al. (1995) is also part of what PytlikZilling and Kimbrough (2016) describe. They describe that trust is always based on risk for the subject to rely on the object to perform an act that is of benefit for the subject. Putting trust in a business environment, this can be explained as the consumer’s belief in that the selling party will fulfill the transaction and their obligations toward the consumer. (Kim et al., 2009) However, trust can vary from different situations and be developed from different assessments of risk. (PytlikZilling & Kimbrough, 2016)

2.2.1 Personality-based trust

Personality-based trust can be described as a trust created from general assumptions or tendencies based on personality psychology (McKnight & Chervany, 2001). Personality-based trust is rooted in the human mind's belief system and is developed through a lifetime (Lewicki & Wiethoff, 2000). This type of trust is developed from previous social experience (Kim, Ferrin & Rao, 2008) Personality-based trust is often perceived as a crucial factor in the development of relationships. It is personality traits that portray a personal image that might create trustworthiness. (McKnight & Chervany, 2001) Azam, Qiang, & Sharif (2013) describe that personality-based trust consists of analyzing several personality traits. These are agreeableness (Azam, et al., 2013), which is connected to a person's relation to others (Lewicki & Wiethoff,

2000). People with high agreeableness have positive beliefs towards others, and people with low agreeableness have more negative beliefs towards others (Azam et al., 2013). Another factor for personality-based trust is an openness to new experience (Azam et al., 2013; Walczuch, & Lundgren, 2004). More openness leads to a person being more likely to try new things and share their experiences with others (Walczuch & Lundgren, 2004). Online platforms require a high level of trust because here suppliers, providers, and consumers dealing with each other online, and they are physically separated, possibilities are hard to measure and the relationships between the providers and users are hard to measure (McKnight et al., 2002). Trust based on personality traits has been described as important in social relations, however, it is more important in e-commerce, because the actors lack physical contact and the trust is engaged on distance (McKnight & Chervany, 2001; Kim, et al., 2009).

2.2.2 Experience-based trust

Experience-based trust entails previously acquired knowledge through social exchange and interaction with other actors in society (McKnight et al., 2002). Experience-based trust is accumulated over time and entails three important factors; (1) experience over time, (2) satisfaction and (3) communication (Walczuch & Lundgren, 2004) People that are developing trust can gain firsthand experience from interacting with other people, however, people can also get second-hand experience from hearing about other people's interactions (Gefen, 2000) i.e from the person's reputation (Morgan & Hunt, 1994) The second-hand experiences is much connected to the importance of communication (Walczuch & Lundgren, 2004) and is therefore crucial in the creation of the trust (McKnight et al., 2002). Consumers perceive experience from the providers plays a vital role in the trust-building process. These experiences may increase or decrease the trust of the customers (Mao et al., 2020) According to Metzeger (2007), many studies have shown that positive experience from a provider in an online platform influences consumers positively to make purchases online. As an example travelers who use Airbnb had a good experience with it, they perceive their high trust on this platform and they prefer a similar experience in the future (Mao et al., 2020). The aspect of satisfaction is connected to if a person's expectations are met through interaction with other parties (Walczuch & Lundgren, 2004).

2.2.3 Cognitive-based trust

Cognitive-based trust is described, in other words, as a rational trust or logical trust and is developed from concrete information that removes uncertainty between parties (Ziegler & Golbeck, 2007). Cognitive trust is based around the choice of interaction with other parties depending on how much concrete information is available, to make a rational decision. (Swift & Hwang, 2013) Cognitive trust is also based on competence to deliver the promised act to a satisfactory result. When deciding to trust a party, it is often considered whether the party is capable of completing the task or not. (Downell, Morrison & Heffernan, 2015) This decision is made on rational and concrete information (Ziegler & Golbeck, 2007). Cognitive trust is most important in newly founded relationships between different parties. In online platforms, consumers face many difficulties to evaluate the credibility of the online purchase. For this reason, online reviews systems from both sides are equally effective for customers and

providers to evaluate each other's also increase trust among them. This online review system is a vital mechanism for customers to make a purchase decision from the platform and it creates a perceived trust to the providers (Ganapati & Reddick, 2018). As an example, in Airbnb platforms, positive reviews of hosts and apps ensure a high level of trust to the customers, in the long term it creates sustainable business for them (Mao et al., 2020).

This is because there is a lack of affective factors between the parties if they were recently introduced to each other. In these situations, facts and information help mediate trust between the two. (Downell et al., 2015)

2.2.4 Perceived security

Perceive security depends on the customers' perceptions of payment methods which are related to the financial information from different types of unauthorized sources (Riquelme & Román, 2014). Online purchases are directly related to consumers' personal and financial information and this is the main concern for the customers for sharing information and companies are more focused on the security of the payment method (Janda, 2008). Consumers perceived security directly related to a positive contribution to the trust factors in an online purchase (Kim, Lee, & Chung, 2013). Many researchers discussed that trust has a direct impact on the consumer's willingness to pay on the online platform (Bhattacherjee, 2002; Gefen, 2002). Security has been a known issue in e-commerce for a long time and consumers may have a lack of trust by sharing personal information in online platforms during any transaction (Hoffman et al., 1999). Mutual trust in the sharing economy is equally important for the service providers and customers, though researchers have focused on the customers. Due to service providers' perceived security, mutual trust between customers and providers are equally important (Bellin, 2017). As the providers share their car, house, or other property to strangers, there can be some risk from the stranger customers because of it there can be a possibility of theft or other damage. So, mutual trust in the sharing economy can be established by the consumers, and service provides feedback or the review system and for the ratings customers and providers for assurance trust on each other (Ganapati & Reddick, 2018). In the sharing economy platform customers have several trust issues with transactions between the suppliers and providers over the internet in the consumer to consumer (C2C) business model (Huurne et al., 2017).

2.2.5 Perceived risk

Perceived risk is one of the key factors in the creation of trust and is described as closely connected to the concept of trust (Viklund, 2003). Risk can be described as the uncertainty or probability of a negative outcome. The variance in results (mostly negative) is, therefore, risk and is a major component of developing trust. (Das & Teng, 2004) Trust and risk have a dynamic relationship where the perceived risk can influence trust, but it can also be the other way, that trust influences perceived risk. If the party is trusted already, the perceived risk is lowered. (Kim & Koo, 2016) Risk factors mainly focus on the formal and informal obligations which occur by the selling party, such as violating individual norms, quality, and commitments. For this reason, consumers who perceive the high level of trust in the selling party perceive low risk because they believe they will not violate transactional obligations. (Kim, et al., 2009)

The mutual trust between both parties increases when the perceived benefit is greater than the perceived risk (Ganapati & Reddick, 2018). According to Kim et al. (2009) assessed risk in business environments, and described that a consumer's belief about the potential negative values from the online transaction causes uncertainty, which is perceived as a risk. In the sharing economy platform companies have lots of scandals worldwide, customers feel that companies may disclose their personal information such as credit card or address, in terms of perceived risk customers feel hesitant to do the transaction in the sharing economy platform (McKnight et al., 2002; Lee et al., 2018).

2.2.6 Perceived quality

Gregg and Walczak (2010) describe website quality as the perceived usefulness of website attributes. In service, quality can be measured in different ways like offering varieties of products, on-time delivery, and being accountable to the customers (ter Huurne et al., 2017). Uncertainty of online platforms depends on the quality of information, high quality information helps to improve the consumers' uncertainty perception (Kim et al., 2008). If the online-based company website has a high quality of information, it creates trustworthiness towards the customers. Consumer trust depends on the website quality also because it helps to reduce the social and psychological distance between the online shoppers are the sellers (Luo, Ba & Zhang. 2012). On the online platform, suppliers do not have any physical appearance, so first impressions are the most vital part of an online store. The trust is depending on the supplier's quality of the online platform, if the quality of the online platform is high, consumers perceive a high level of trust. Customers consider the online platform has competence, integrity, and benevolence, and customers will depend on the supplier and platform (McKnight et al., 2002). According to Fung and Lee (1999) if the information on the online platform has high quality, and the platform has a good quality of design, customers' trust generally increases. As an example, a bank has an extraordinary, prosperous physical appearance, customers generally willing to trust the service of the bank. Here the customers don't know the bank employees who run the bank but they have trust in them because of their out appearances which implies them trustworthy (McKnight et al., 2002). Trustworthiness on the online platform works to reduce the negative effects of online sellers' invisibility and product's ambiguity (Luo et al., 2012). The major aspect of perceived quality is meeting the expectations of the customer (Cristobal, Flavián & Guinalíu, 2007).

3. Hypotheses development

The literature review was conducted to find factors of trust that are found in previous studies and therefore could create hypotheses for the current study.

3.1 Trust between actors

Mao et al. (2020) have conducted research where they found significant factors of trust in relation to sharing economy actors. The research showed that significant trust factors were based on (1) personality, (2) previous experience, or (3) cognition (Mao et al., 2020). What is important to consider is that Mao et al. (2020) found these significant factors of trust between market actors are a sharing economical environment, i.e the trust between the person purchasing the service and the person providing the service (hosts). Furthermore, this has been researched by other authors as well. According to Cheng, Su, and Yang (2020) identified four antecedents of trust in a sharing economy environment. Integrity, benevolence. reputation and motivation were the factors of trust (Cheng, Su & Yang, 2020). Parts of the findings of Cheng et al. (2020) are similar to what Mao et al. (2020) found in their study. Cheng et al. (2020) found that benevolence was constructed by personal characteristics of the service provider, such as appearance, politeness, and empathy which are similar to what Mao et al. (2020) describe as personality-based trust. The concept of trust based on personality or personal attributes is also found by Ert, Fleischer, and Magen (2016). The authors found in their study visual information, such as photos, appearance, and reputation of the service provider were key factors in developing trust (Ert et al., 2016). However, if the actor purchasing the service do not identify with the personal characteristics of the service provider, the impact on trust might be negative (Yang, Lee, Lee, & Koo, 2019) Because of the discussion above, it is hypothesized that:

H1: Personality-based trust has a positive effect on trust between actors.

The concept of experience-based trust by Mao et al. (2020) is described as the purchaser of the service already having tried the service before or have gotten information regarding the service previously. The experience-based trust can be based on other people's experience of the service provider (Ert et al., 2016) and is evaluated through the reputation of the service provider (ter Huurne et al., 2017; Wang, Zheng, Chen & Ding, 2015). Furthermore, negative reviews of the service provider would have a stronger negative impact on trust than the positive impact of positive reviews (Li, Guo, Wang, & Zhang, 2016). However, positive reviews and a positive reputation of the service provider is believed to have a positive impact on trust (Bente, Baptist, & Leuschner, 2012) Because of the discussion above, it is hypothesized that:

H2: Previous experience of the service provider has a positive effect on trust.

Furthermore, Mao et al. (2020) also found that cognitive-based trust had a positive impact on the trust between actors in a sharing economy setting. Cognitive trust entails a thought process to make a decision (Mao et al., 2020) where the purchaser evaluates the quality of information

from the service provider (ter Huurne et al., 2017; Chen, Lai, & Lin, 2014). One aspect of cognitive trust is that buyers' thought process can be based on the verification of the seller, such as access to background information, certification, and competence, which is to prove that the seller exists (Kang, Gao, Wang, & Zheng, 2016). The quality of the information available is also important in cognitive decision making (ter Huurne et al., 2017), and generally, the more information about the service provider, the more they can be trusted (Chen, Lai & Lin, 2014). Cognitive-based trust is usually stronger when the relationship between actors is new. This type of trust is often more impactful than affective or emotionally-based trust (Yang et al., 2019) Because of the discussion above, it is hypothesized that:

H3: Cognitive evaluation of the service provider has a positive effect on trust.

3.2 Trust towards the platform

Also important when examining trust in a sharing economy setting is to research trust towards the platform for connecting actors (ter Huurne et al., 2017). The main issues found when evaluating trust towards online sharing economy platforms was regarding perceived security and privacy (Mao et al., 2020; Chen et al., 2014; Kang et al., 2016). In terms of the trust, consumers perceive risk in terms of disclosing their personal information such as credit cards or address. These create hesitation to the customers to do the transaction in a sharing economy platform. (McKnight, Choudhury & Kacmar, 2002; Lee, Ahn, Song & Ahn, 2018) Many researchers discussed that trust has a direct impact on the consumer's willingness to pay on the online platform (Bhattacherjee, 2002; Gefen, 2002) This issue has been well known in the field of e-commerce for a long time, consumers may have a lack of trust by sharing personal information on an online platform during any transaction (Hoffman, Novak, & Peralta, 1999). Mao et al. (2020) found that if the buyer of the service perceives that the online platform is secure, they are more likely to trust the platform. Because of the discussion above, it is hypothesized that:

H4: High perceived security increases trust towards the platform.

In addition to this Mao et al. (2020) finds that the perceived risk of mishappenings when purchasing the service through an online platform would decrease the trust. The perceived risk of failure for delivering the service is countered by several online platforms providing a refund guarantee if the service is not delivered (San-Martín & Camarero, 2014). According to Kim et al. (2009) any purchase decision on an online platform is revealed by both perceived risk and benefit. A low level of risk in online platforms will cause consumers to have a high level of trust towards the service provider (Kim et al., 2009). In the sharing economy arena, online mutual feedback from both ends and the rating system among the service consumers and providers influence individual consumers to perceive their trust on the platform (Ganapati & Reddick, 2018). Because of the discussion above, it is hypothesized that:

Ter Huurne et al. (2017) identified through reviewing previous literature that the perceived quality of the service and the web platform factor into developing trust in a sharing economy setting. If the quality of the service, website and information on the website is perceived as good, buyers will generally have a higher level of trust towards the platform (Mao, et al., 2020). Service quality in the context of sharing economy platforms could be the responsiveness of the business behind the platform (San-Martín & Camarero, 2014) or to which level the platform meets the buyer’s expectations (Mao et al., 2020; Cheng et al., 2019). A positive perception of the website, regarding the layout, has a positive impact on the trust towards the platform (Filieri, Alguezaui, & McLeay, 2015) Because of the discussion above, it is hypothesized that:

H6: High perceived quality increases trust towards the platform

In the sharing economy platform customers have several factors of trust in connection transactions between the suppliers and providers over the internet in the consumer to consumer (C2C) business model (ter Huurne et al., 2017). Company reputation, consumer retention, purchase intention, willingness to act and overall market performance influenced by trust. (Delgado-Ballester et al., 2003; Erdem and Swait 2004; Matzler et al., 2008; Willmott 2003). To understand trust in this context, the relationship between the trust towards the service provider and the trust towards the platform needs to be analyzed. It is possible that a consumer could have feelings of trust towards the sharing economy platform, but not towards the actor providing the service. (ter Huurne et al., 2017). It is therefore interesting to examine the potential impact of trusting the sharing economy platform on the overall trust of the actor providing the service. This statement of a trusted platform leading to a trusted actor providing the service is relatively unexplored in previous literature (ter Huurne et al., 2017). The assumption that trust in the sharing economy platform positively influences the trust towards the actors providing the service has been previously tested, but the results showed no relation (Mao et al., 2020). Therefore, to reproduce the test of this predefined assumption, hypothesis 7 was developed.

3.3 Conceptual model

Based on the literature review, six factors of trust were identified as important and connected to the trust between actors and the trust towards the platform. With the basis of previous literature, seven hypotheses were developed with assumed relationships between the different trust factors.

Figure 1. Conceptual model of relationships between trust factors (own construction)

Based on previous studies, the model (as you see in Figure 1) and hypotheses 1-3 assume that personality-, experience-, and cognitive-based trust all have a positive influence on the trust between actors. These hypotheses are all based on previous study results (see Mao, et al., 2020; ter Huurne, et al., 2017) Also based on previous studies (see ter Huurne, et al., 2017), the conceptual model and hypotheses 4-6 assume that perceived security, risk and quality all influence the trust towards the sharing economy platform. Lastly, hypothesis 7 is made to add to existing literature and see if there is a relationship or influence between the trust for the platform and the trust towards the actors providing the service.

Trust between actors

Trust towards platform

Personality Experience Cognition

H1 H2 H3 Low risk H5 Security H4 Perceived quality

4. Methodology

This chapter contains a detailed overview of how the study was conducted. The methodology includes a presentation of the research design and the processes of how the measurements were made.

4.1 Research approach

The research has been conducted quantitatively. This methodological approach was motivated by the purpose to find an overview of connections between different concepts (Bryman & Bell, 2011) of trust. The fact that statistical data was used for the study helps explore and describe the phenomenon (Canela, Alegre, & Ibarra, 2019) of trust in a sharing economy setting. Quantitative methods are more suitable when the purpose is to describe the reality in connection to different concepts (Bryman & Bell, 2011) and was therefore chosen for this study. The purpose behind using a quantitative research method was to study a problem, based on previously developed theories or constructs, in order to generalize (Sogunro, 2002). Because of the need to test previously developed theories or constructs, the study was conducted with a deductive approach, meaning that the starting point of the study was reviewing previously published literature within the field of research.

When gathering previous literature for the study, Google Scholar was used as a search engine and database for previous studies published in scientific journals. To find relevant literature and journals the searches were defined with a set of specific search terms. The most used search terms in the gathering of literature were; trust, sharing economy, risk, e-commerce. To increase the reliability of the paper, the authors chose only to use literature that had been through a peer-review process. Articles that are published in journals are generally peer-reviewed by other researchers before they are published, which increases the reliability of the information (Bryman & Bell, 2011). In addition to this, a literature review (by ter Huurne, et al., 2017) was also used to create an overview of recent research with the concepts of trust and sharing economy. The literature by ter Huurne, et al. (2017) served a great function as a means of snowballing references (Bryman & Bell, 2011). The previous studies laid a foundation for the development of hypotheses.

Figure 2. Research process (own construction)

In figure 2 the process of how the research was conducted was visualized. This was made to highlight the important role of previous literature in the deductive research approach (Bryman & Bell, 2011). A major benefit of using previous literature as the foundation of a new study was that replication and testing of pre-defined concepts highly increased the credibility of the new study (Freese, 2007). A replication or reproduction of previously tested concepts increased the likelihood of the results being objective if the studies amounted to similar results, which then increased the validity of the constructs and the reliability of the studies (Bryman & Bell, 2011). When replication or reproduction of a research method was used from previous literature, it had a positive effect on showing the transparency of the current study (Freese, 2007). It was also important to consider that the research that was reproduced was of relevance in the aspect of time of publication (Bryman & Bell, 2011)

For this study, the concepts of measurement were replicated from previous research conducted by Mao, et al. (2020), which also examined several factors of trust in a sharing economy environment. This study by Mao, et al. (2020) was perceived as relevant also because of its recent date of publication. The goal of using the predefined constructs for measurement was to test the reliability of the previous literature. Also, the quantitative research approach allowed for the examination of potential differences in results. (Bryman & Bell, 2011)

4.2 Data collection and measurements

The collection of data was done by conducting a questionnaire with question items. When creating the questionnaire, it was decided that the survey should contain statements or closed questions to make it easier to respond (Bryman & Bell, 2011). Also, the questionnaire was built around an example company (Airbnb) that operates in the sharing economy market to make the statements more relatable for the respondents. It was therefore important to clearly state that respondents would have had to be familiar with the service of Airbnb. However, the fact that a specific sharing economy based company could be argued to be beneficial since it could have generated more honest responses if the respondents are interested in the service (Bryman & Bell, 2011).

The questionnaire aimed to examine 8 different concepts, (1) trust between actors, (2) trust

towards the platform, (3) personality-based trust, (4) experience-based trust, (5) cognitive-based trust, (6) perceived security, (7) perceived risk and (8) perceived quality. The

questionnaire, therefore, contained a total of 27 statements (question items) to grasp the concepts of the study. Each of the 8 concepts had 3 or 4 corresponding statements. The items were assessed with a psychometric 7-point Likert scale, ranging from strongly disagree (1) to strongly agree (7). The question items have been previously tested in similar types of studies (see Mao et al., 2020). Using a Likert scale as a way of measurement is commonly used and accepted in quantitative research, and was suitable when examining relationships between several concepts (Norman, 2010).

Using a Likert scale also improved the computerization and the quantification of the responses, which was crucial for the analysis (Bryman & Bell, 2011) Also, using a pre-set scale as a way of measuring made the responses more consistent, and therefore made the responses easier to compare to previously conducted research (Arnon & Reichel, 2009). However, some negatives of using closed questions or statements were taken into consideration. When using statements as a way of measurement, it might occur that the respondents interpreted words differently and had no way of explaining their thought process. If this was to occur it would be an issue regarding the validity of the constructs. (Bryman & Bell, 2011) It was therefore crucial to preemptively test the reliability and validity of the constructs before the data was collected (Norman, 2010).

The questionnaire was designed online and was sent out for potential respondents digitally. However, questionnaires that are distributed online bring some advantages and disadvantages (Bryman & Bell, 2011) that have been considered. The advantages of using digital questionnaires are that the responses can be processed more rapidly than other forms of surveys (Wright, 2005). Another advantage of the data being collected digitally was that respondents are easily accessible through the internet (Bryman & Bell, 2011; Wright, 2005) Since sharing economy businesses (platforms) operate in an online environment, it was perceived as a suitable way of data collection. The main disadvantage of collecting data digitally was that it was difficult to guarantee that the respondents fit the intended sample. Apart from using demographic variables, it was problematic controlling the sample. (Wright, 2005). In addition,

it was also problematic to ensure the response rate of the questionnaire. It was previously shown that questionnaires that are performed digitally, in general, have a lower response rate. (Wright & Schwager, 2008)

4.3 Sample

The sampling method used for the study was convenience sampling, meaning that respondents were chosen based on availability to the authors of the study (Bryman & Bell, 2011). The convenience sampling approach is a form of non-probability sampling, meaning that the population is selected in a specific way, diminishing the likelihood of a random selection. In this case, the specific way of selecting respondents was based on availability to the authors. Convenience sampling posed issues regarding determining the generalizability to a larger population. This was because of the problem with ensuring that the chosen respondents had thoughts regarding the research area and therefore were a representative group (Hultsch, MacDonald, Hunter, Maitland, & Dixon, 2002). Convenience samples are seen as relatively inclusive, and the representativity of the group of respondents may vary between studies (Bryman & Bell, 2011). Because of the fact that convenience sampling had issues with determining the representativity of the respondents in a larger population, it was difficult to draw accurate conclusions (Hultsch, et al., 2002) To try to counteract these issues certain specific criteria were selected to improve the credibility of the study (Bryman & Bell, 2011). Because of the focus on Airbnb in the study, it was stated that the respondents would have had previous experience to strengthen the possibility that the sample could provide a credible result. The questionnaire was sent to 2459 potential respondents. In total, the questionnaire received 275 responses (response rate 11,18%). However, 2 out of 275 responses were rejected since these responses only specified the gender demographic but did not complete the rest of the survey. The questionnaire was sent in digital form to potential respondents in the personal networks of the authors (convenience sample). In addition to this, the survey was also posted in Facebook groups based on discussions of Airbnb. The fact that the groups were based around Airbnb was a way of trying to counteract the issue regarding representativity for a larger population. The main ways of sending the questionnaire were through e-mail and personal messages on the social media platform Facebook. However, the fact that the survey was distributed using a convenience sample makes it that the potential relationships only apply to the dataset. The generalisability and the overall representativity is limited because of the sampling approach.

The international dataset contained respondents from 30 different countries. The most common countries of residences were Sweden, Bangladesh, Pakistan, and India. The majority of the respondents were aged 18-29 and the most common occupation was full-time employment.

Table 1. Demographics of data set (*other signify countries with less than 10 respondents) Gender n. % Occupation n. % Female Male Other 118 156 1 43,4 56,3 0,3 Full-time employed Student Other 144 107 22 52,7 39,2 8,1 Age Country 18-29 30-39 40-49 50-59 60+ 181 79 10 1 1 66,5 29,0 3,7 0,4 0,4 Sweden Bangladesh Pakistan India Other* 101 50 13 11 98 37,0 18,3 4,8 4,0 35,9

4.4 Operationalization

The operationalization showed the connection between the chosen constructs and the items for the measurement of the study. Also, it showed where the used items for measurement had been previously tested and therefore the source of the reproduction.

Table 2. Operationalization of question items related to trust.

Constructs Purpose Items References

Trust between actors Chronbach’s Alpha - 0,934

The purpose of the construct is to measure the existing trust towards service providers on Airbnb

1 - I am confident most Airbnb hosts are

reliable.

2 - I am confident most Airbnb hosts are

honest.

3 - I am confident most Airbnb hosts are

trustworthy.

4 - I am confident about what to expect from

Airbnb hosts. Mao, et al. (2020) Trust towards platform Chronbach’s Alpha - 0,930

The purpose of the construct is to measure the existing trust towards Airbnb as a platform.

5 -Airbnb is competent when dealing with

tenants

6 - Statements provided by Airbnb are

trustworthy

7 - Airbnb is honest in dealing with my

private data

8 - Airbnb delivers agreed service to the

tenants Hawlitschek, et al. (2016) Personality-based trust Chronbach’s Alpha - 0,855

The purpose of the construct is to see how personality impacts the overall trust towards Airbnb service providers.

9 - I see myself as someone who likes to

cooperate with others

10 - I believe that people in general keep

their promises

11 - I believe that professional people in

general do a good job at their work

12 - I trust people in general until they give

me a reason not to trust them

Azam, et al. (2013)

McKnight, et al. (2002)

Experience-based trust Chronbach’s Alpha - 0,917

The purpose of the construct is to see how previous experience impacts the overall trust towards Airbnb service providers.

13 - My past experiences with Airbnb hosts

were positive

14 -I have received enjoyable service when

staying with Airbnb hosts

15 - Airbnb hosts have done satisfactory jobs

in providing me accommodations Mao, et al. (2020) Cognitive-based trust Chronbach’s Alpha - 0,906

The purpose of the construct is to see how cognitive reasoning impacts the overall trust towards Airbnb service providers.

16 - For me, online review systems in

general provide accurate information about Airbnb host’ reputations

17 - A considerable amount of useful

information about hosts is available through Airbnb's online review systems

18 - For me, Airbnb's online review systems

are effective Mao, et al. (2020) Perceived security Chronbach’s Alpha - 0,765

The purpose of the construct is to see how perceived security impacts the overall trust towards Airbnb as a platform.

19 - I feel secure about Airbnb's electronic

payment systems

20 - I am willing to use my credit card on

Airbnb's website to make a purchase

21 - I have no problem sharing my personal

information on Airbnb’s website

Kim, et al. (2008)

Perceived risk Chronbach’s Alpha - 0,916

The purpose of the construct is to see how perceived risk impacts the overall trust towards Airbnb as a platform.

22 - For me, using Airbnb when traveling

involves considerable low risk

23 - For me, using Airbnb when traveling

involves a low potential risk for loss

24 - My decision to use Airbnb when

traveling is not risky

Mao, et al. (2020) Perceived quality Chronbach’s Alpha - 0,890

The purpose of the construct is to see how perceived quality impacts the overall trust towards Airbnb as a platform.

25 - It is easy to find appropriate

accommodation through the Airbnb website

26 - The Airbnb website has high speed

webpage loading

27 -The Airbnb website has an attractive

look and feel

Filieri, et al. (2015)

4.5 Reliability and validity

The main objective of conduction studies is to examine and present a credible and trustworthy result. Because of this, the study was reliant on the reliability and validity of the constructs used for measurement. In order to test the internal reliability of the constructs, Cronbach’s Alpha was used. This test served the function of ensuring that that the question items were related and therefore, could provide an accurate measurement of the chosen constructs. (Bryman & Bell, 2011) In order for the construct to be reliable, the alpha coefficient had to reach a value of 0,700. (Bobko, 2001). This meant that the alpha coefficient of the combined question items had to reach a minimum level of 0,700 in order to be computed into constructs. For the Cronbach’s alpha test several question items were combined to see if they were accurate measurements of the intended constructs. The Cronbach's alpha reliability test of combining variables into constructs showed that all intended constructs were internally reliable since all exceeded the 0,700 value as shown in Table 2. The construct validity was handled through

using predefined concepts form studies that had previously tested the question items. The fact that the items were tested previously highly strengthens the validity and credibility of the concepts. (Bryman & Bell, 2011) The content validity was also controlled, with the method of letting the question items be peer-reviewed. Because the study was done as part of the International Marketing Masters Programme at Mälardalen university all question items were reviewed by peers participating in the master's program. The fact that the question items were reviewed by others than the authors strengthens the likelihood of the content being valid (Bryman & Bell, 2011)

4.6 Data analysis method

To analyze the data that was collected through the survey, the statistical program SPSS was used. The data were first analyzed using Pearson's correlation. This method was chosen since the data was based on values from a Likert scale. Pearson's correlation is most widely used to interpret data in scale form (Bryman & Bell, 2011) The correlation analysis shows linear relationships between the different variables of the data set. The strength of the linear relationships also works as an indicator if the data can be further analyzed in a regression analysis model. A correlation alpha coefficient with the value above 0.500 is interpreted to be a moderate relationship, and values above 0.700 are interpreted as a strong relationship (Bryman & Bell, 2011).

For this study, a moderate relationship that was statistically significant was able to be further analyzed in a multi-linear regression model. In addition to this, the moderate correlation strengthened the validity of the constructs, since it proved that the constructs are used as measurements for the different variables (Bryman & Bell, 2011). To analyze the data with the multi-linear regression, it was important to take the beta coefficient into consideration as well as seeing if the t-value exceeded the critical value. To find the critical t-value first the degree of freedom (Df) had to be determined. The Df was calculated through Df = N - 1, which in this study resulted in Df = 272. The critical t-value for Df above 100 at the 95% significance level was 1.660 (Bryman & Bell, 2011). This meant that t-values above this level can be perceived as a reliable result. However, the data analysis only applied to the specific data set because of the fact that a convenience sampling approach was used.

4.7 Research ethics

When the research was conducted several ethical standpoints were taken into consideration. It was important to be transparent towards the respondent. Therefore, all respondents who participated in the survey were informed that the survey was voluntary. The respondents were informed that they had the right to avoid answering questions in the survey and could at any time choose to leave the survey. Therefore, none of the survey questions were programmed as obligatory to answer, so that the respondents could proceed with the survey, even if avoiding multiple answers. In addition to this, the respondents were informed regarding the purpose of the survey and how the data was used. The respondents were also informed that all data were treated confidentially.

5. Analysis

The data has been analyzed through the statistical program SPSS. This method allowed for multiple statistical tests of the data. The most important statistical test was regression analysis since it tested the predefined hypothesis. Regression analysis not only showed connections between variables but also shows the causal relationship between variables (Bryman & Bell, 2011). A correlation analysis was also conducted, however, this type of analysis only showed if the variables had an impact on each other, but did not show a causal relationship. The data that was analyzed was first viewed using the descriptive statistics of the data set. This was done to get an overview of the data that had been collected and to see if the variables would fit into upcoming methods for analysis.

Table 3. Descriptive statistics of dataset.

The Pearson correlation analysis indicated a strong positive relationship between all of the constructs that were measured (see figure 3).). In addition to the strong correlations, they were all significant (below 0.01 sig. two-tailed). All of the Pearson correlation coefficients had a value of at least 0.500, indicating relatively strong positive relations between the different constructs. Because of the fact that the correlation analysis indicated that the constructs influenced each other, it was important to examine how the different variables could be used to find causality, however, the relationship between the variables only applies to the specific dataset. Mean Std. Deviation N HOSTtrust 21,9634 5,42382 273 PLATFORMtrust 21,7985 5,33369 273 PERSONALtrust 17,2132 3,35239 272 EXPtrust 16,7519 3,93836 270 COGtrust 17,2463 3,51722 272 SECURITY 16,1882 4,13360 271 RISK 16,6889 3,86090 270 QUALITY 17,0520 3,73231 269

Figure 3. Pearson correlation matrix

In order to test hypotheses H1 (personality-based trust has a positive effect on trust between

actors), H2 (experience-based trust has a positive effect on trust between actors) and H3 (cognitive-based trust has a positive effect on trust between actors) a multi-linear regression

test was performed (see Appendix 2). The test included using three constructs, personality-based trust, experience-personality-based trust, and cognitive-personality-based trust as predictors of the dependent variable trust between factors. The model of these three predictors of the dependent variable had an R2-value of 0,607 and was tested with an ANOVA-test, showing that the model was significant (sig. 0,000). The regression analysis of testing H1, using personality-based trust as a predictor of trust between actors resulted in the standardized beta coefficient 0,338 and was statistically significant t=5,880**. This means that H1 is accepted as a predictor of trust between actors. For testing H2 where experience-based trust is a predictor of trust between actors the standardized beta coefficient was 0,267 and was statistically significant through

t=4,500**. Therefore, H2 is also accepted. The third construct tested H3 where cognitive-based

trust was the predictor which resulted in a standardized beta coefficient of 0,266 and was also significant through t=4,267**. This means that H3 also is accepted. Based on the regression analysis with an indication from the beta coefficient, personality-based trust is the strongest predictor (amongst the three measured) for trust between actors. In addition to this, experience-, and cognitive-based trust also had a positive influence on trust between actors.

The same type of test was used to test hypotheses H4 (perceived security increases trust

towards the platform), H5 (perceived risk increases trust towards the platform) and H6 (perceived quality increases trust towards the platform). The constructs included in the test

were perceived security, perceived risk and perceived quality as predictors of trust towards the platform (see Appendix 3). The model of the three predictors of the dependent variable had an R2-value of 0,508 and was tested with an ANOVA-test, showing that the model was significant (sig. 0,000). The test of H3 indicated perceived security as a predictor of trust towards platform through the standardized beta coefficient 0,262 and was statistically significant through the t-test, where t=4,120**. Therefore, H3 is accepted. The test of H4 showed perceived risk as a predictor of trust towards platform with the standardized beta coefficient 0,374 and was statistically significant through t=4,908**. Therefore, H5 is also accepted. The last of the

predictors was perceived quality on trust towards platform where the beta coefficient was 0,151, however, this was not significant t=1,871 and the significance level 0,063. Therefore,

H6 is rejected.

The last hypothesis, H7 (trust towards the platform increases the trust towards the actors) , was tested in the same way as the other hypotheses (see Appendix 4). For this test, trust towards the platform was used as a predictor of trust between actors. The model of prediction had a R2 -value of 0,599 and the ANOVA-test showed it being significant (sig. 0,000). The standardized beta coefficient was 0,775 and was statistically significant through t=20,194**. Therefore, H7 is accepted where trust towards the platform is a predictor of trust between actors.

In order to summarize the analysis of the data collection table 4 is used. The findings of the study show that 6 out of the 7 hypotheses are supported and only hypothesis 6 (perceived

quality increases trust towards the platform) was rejected.

Table 4. Summary of multi-linear regression

T-value - ** = sig < 0.05 H1 5,880** Supported H2 4,500** Supported H3 4,267** Supported H4 4,120** Supported H5 4,908** Supported H6 1,871 Rejected H7 20,194** Supported

6. Discussion

The key findings of the study indicate that the trust towards the actor providing a service in a sharing economy context is influenced by personality-, experience-, and cognitive-based trust. The findings show that a person's personality in connection to trust is the strongest influencing factor. This could be explained by the theory of McKnight and Chervany (2001) where trust is discussed. The authors mention that personality-based trust is often built in a beginning phase of a relationship, when not much information is available (McKnight & Chervany, 2001). This could be because the service provider is usually unknown to the one purchasing the service, making the one purchasing the service rely more on their intuition. This can be connected to openness and agreeableness, where people with high agreeableness have positive thoughts about others (Azam, et al., 2013). It could be that the purchaser in general has positive feelings about others, and therefore chose to trust the actor providing the service. These findings are contributing to confirming previous findings from other authors (e.g. Mao, et al., 2020) where these factors also were proven to increase trust towards the actor providing the service. Furthermore, experience-, and cognitive-based trust were almost equally important, which was interesting. Both the t-value and the beta coefficient were almost equal for both variables, indicating that these factors are of almost equal importance to the people purchasing a service in a sharing economy setting. There are many consumers who have trust issues on online platforms like they do not want to share their personal information. In the sharing economy companies can be associated with bad experiences and negative reviews. However, because customers feel comfortable to use the service and to use the platform because of perceived benefits (ter Huurne et al., 2017). Experience-based trust had a positive impact, but it was not necessarily a strong relationship. The fact that previous experience only had a small effect on the overall trust between actors in a sharing economy environment could be because the actors interact with new actors for many transactions. Furthermore, cognitive-based trust was almost equally important at previous experience. If the buyer of the service has not met the provider of the service previously, it is often so that the decision-making lies on more cognitive aspects (Yang, et al., 2019). However, the finding of this study does not indicate that cognitive-based trust increases the trust towards a new actor providing the service more than the factors types of trust. The results of this study show that trust in a sharing economy context is mostly reliant on personality-based trust.

The second part of the findings of this study indicates that trust towards a sharing economy platform is influenced by perceived security and perceived risk, but in contrast to previous literature (e.g. Mao, et al., 2020), not perceived quality. The study shows that if the online sharing economy platform is perceived as secure, people are more likely to trust in sharing their personal information. This comes as no surprise since the issues have been well-known and thoroughly discussed in previous literature for many years (Kim, Lee, & Chung, 2013; Hoffmann, et al, 1999; ter Huurne, et al., 2017). This study also measured the impact of perceived risk on the trust towards the sharing economy platform. The purchasers that perceived the risk as low were more likely to trust the sharing economy platform. This has also been previously studied (e.g. Kim et al., 2009) and these new findings contribute to strengthening these previous results. In this study, the construct of risk was connected to