Rating mechanisms among participants in sharing economy platforms
First Monday

Rating mechanisms among participants in sharing economy platforms by Lene Pettersen



Abstract
This study shares findings from 23 qualitative interviews of participants from five sharing economy platforms in Norway (2016) about how they make sense of rating scores, use rating scores when making decisions and provide ratings of others in sharing economy platforms. Online ratings or review scores in sharing economy services tend to be positive on average. Consumers need to develop a larger awareness about the social mechanisms at play when peers rate each other, how a given platform may control and change rating categories and how rank is measured and presented by algorithms. Rating scores are vital to purchasing decisions.

Contents

Introduction
Literature review
Methodology
Data analysis
Findings
Discussion and conclusions

 


 

Introduction

Social networking tools are used to identify whom to trust in decision-making processes (Masum and Tovey, 2011). Central in both e-commerce and sharing economy services or platforms, is the notion that online user evaluations or ratings nurture confidence and lubricate relations between individuals that do not know each other in person. Botsman and Rogers (2010) list critical mass and trust as two important principles in the sharing economy. Trust is a mental process consisting of expectations and interpretation, a kind of “faith” [1]. Botsman (2017) argued that a new paradigm shift driven by new technologies is emerging, characterized by ‘distributed trust’ which “is rewriting the rules of human relationships” [2]. ‘Distributed trust’ denote a kind of trust between strangers, such as renting an Airbnb listing or taking a short drive with an unknown Über driver. ‘Distributed trust’ also concerns trusting non-human actors, for example bots, algorithms, artificial intelligence or self-driving cars [3]. A key design element to nurture trust are “trust building tools” (Hausemer, et al., 2017); evaluation mechanisms that represent scores or rating of products or services (Pettersen, 2017).

The use of rating systems or online verification mechanisms is a recent development, emerging from the evolution of Web 2.0 technologies (Scott and Orlikowski, 2012). Analysts estimate that online reviews or user evaluations of products and services represent a formidable value for the peer-to-peer (P2P) economy. Ratings have a direct effect on the sales of goods and services (Chevalier and Mayzlin, 2006; Luca and Zervas, 2016). Negative reviews on eBay (Resnick and Zeckhauser, 2002) and TripAdvisor (Scott and Orlikowski, 2012) are found to have a negative impact on price.

Arguably, user ratings influence consumer purchase decisions and shopping behavior. However, researchers have found that online ratings are mainly positive and that negative reviews are rarely given (Aral, 2014; Luca and Zervas, 2016; Mellet, et al., 2014; Slee, 2013; Zervas, et al., 2015). Moreover, Hausemer, et al. (2017) found in their screening of 485 platforms, a user survey, focus groups and case studies, that evaluation or rating systems were ‘neither fully reliable nor transparent. Their effectiveness is therefore subject to serious doubt’ [4].

While there is a growing body of research into “professional” online rating systems initiated by businesses (e.g., TripAdvisor, eBay, and Yelp), less research has been conducted on how consumers and product or service providers of sharing economy platforms use rating scores in making decisions for selecting one product or service over another, and how and why they rate. One exception is Fagerstrøm, et al. (2017) that examined the role of profiles in decisions in Airbnb.

Few studies have investigated in depth how peer consumers and service providers make decisions in sharing economy platforms. Investigating these processes contributes to a better understanding of why there is a tendency to give positive scores in rating systems, and in particular in sharing economy services. Insights into (1) how rating scores are provided and interpreted by platform users; and (2) sociological aspects in both rating and decision processes are important (Zervas, et al., 2015). These two aspects lead to these research questions — (a) How do people reason when they choose one product or service over another in sharing economy platforms? (b) How do people reason in their reviews? Answers to these research questions will also provide some empirical basis to Botsman’s [5] observation that “distributed trust is rewriting the rules of human relationships”.

 

++++++++++

Literature review

Rating or review systems

Online feedback mechanisms or reputation systems stimulate large word-of-mouth networks in which individuals share opinions and experiences on a wide range of topics (e.g., companies, products and services) (Dellarocas, 2003). With digtial, networked platforms, restaurant reviews were democratized, including a wider range of restaurants and a more open review process (Mellet, et al., 2014).

Online user ratings come in many shapes and forms. Reviews can be provided locally (on an individual product, Web site or platform), distributed (on a company’s Facebook page), or in comparison services (platforms that exclusively present overall consumers’ assessments of goods and services, such as TripAdvisor or Yelp). In some sharing economy services, one party independently evaluates another, while in other platforms, a review is not published unless both parties participate in constructing a review.

Moreover, rating systems in commercial platforms (e.g., rating a product bought from Amazon), and online reviews of services (e.g., rating an individual selling a book in Amazon’s online marketplace) illustrates that rating a product from a business player (professional seller) differs from rating a service from a person, even if the deal is completed within a commercial context (e.g., Amazon’s platform) [6]. This illustrates the implications of rating when several services, markets and business models operate within the same sharing economy platform (Pettersen, 2017).

The diversity of evaluation types and the complexity and blurred borders of what a person actually evaluates makes it problematic to discuss and compare online user systems on equal terms.

Rating scores

While grades or scores in traditional grading systems (for example, Michelin Guide) were distributed diversely between low and high values, online evaluations from consumers appear to have little diverse distribution, as ratings are mainly positive (Aral, 2014; Luca and Zervas, 2016; Mellet, et al., 2014; Slee, 2013; Zervas, et al., 2015). In a study of six comparison services of French restaurants, Mellet, et al. (2014) found that a majority of restaurants received an average rating of four out of five. Similarly, a study of eBay (Resnick and Zeckhauser, 2002) found that buyers’ feedback was overwhelmingly positive; 99.1 percent of comments were positive, 0.6 percent were negative and 0.3 percent were neutral (Dellarocas, 2003).

Resnick and Zeckhauser (2002) found that neutral reviews were left by buyers and sellers when a trade had been somewhat problematic (e.g., delays and poor communication), while negative reviews were only provided when there had been a truly negative experience (e.g., never shipped as agreed, destroyed or cheated). A study of BlaBlaCars, a French sharing economy company that connects drivers with people traveling the same way throughout Europe, found that more than 98 percent of ratings were five out of five stars [7]. Zervas, et al. (2015) analyzed ratings for over 600,000 properties listed on Airbnb worldwide, and found that nearly 95 percent of Airbnb properties boasted an average user-generated rating of either 4.5 or five stars (the maximum); virtually none had less than a 3.5 star rating. Mellet, et al. (2014) found that ratings did not follow a normal distribution. In a normally distributed variable, the value lies close to the mean and rarely do the values have large deviations. Rating scores, however, typically lie outside normal distribution and thus could be interpreted as not reliable (Mellet, et al., 2014).

Composing a review is voluntary, meaning that only those who review are counted or measured. Although Resnick and Zeckhauser (2002) found that just over half of the buyers gave feedback after their purchase on eBay, only 13 percent of the users left an evaluation or review at TripAdvisor (Gretzel, et al., 2010). Clearly, quite a bit of data is missing in review samples, but little is known of why people tend to leave positive reviews.

As one explanation for the tendency of high rating scores for Airbnb, Zervas, et al. (2015) suggest that “some sociological factors are at work, whereby individuals rate other individuals differently or more tactfully, than they rate firms such as hotels, independent of the platform” [8].

Aral (2014) stated that older consumers’ ratings have an effect on new ratings, nurturing “rating bubbles”. Ratings are subjective; what one person finds important may be something completely different from another individual. Different dimensions may be considered for the same product or service. However, personal dimensions matter in making buying or renting decisions in sharing economy platforms. Fagerstrøm, et al. (2017), for example, examined the role that the profile pictures on Airbnb listings play in consumers’ rental choices. Their findings suggest that after price, facial expressions in a profile picture and customer ratings have the largest effects on how consumers approach listings.

Edelman and Luca (2014) confirmed the importance of listing photos at Airbnb in attracting higher prices. However, they also found that non-black and black hosts received strikingly different rent levels. When controlling for location, price, size and other factors, Edelman and Luca (2014) revealed that non-black hosts earned 12 percent more for a similar listing with similar ratings and photos having black hosts [9]. In a later study, Edelman, et al. (2016) found that requests from guests with distinctively African-American names were 16 percent less likely to be accepted than identical guests with distinctively white names.

Rating system design and algorithms

Because constructing and interpreting reviews and ratings are cognitive, social and subjective processes, it is problematic that they are objectively measured and presented in the form of a grading system or quantitative ranking (Scott and Orlikowski, 2012).

In addition, rating systems are designed differently. Which categories should be considered may be elements that a platform defines as important, not necessarily elements that benefit a consumer or service provider. How reviews appear depends on how assessments are designed. Blank (2006) and Scott and Orlikowski (2012) pointed out that ranking systems amount to a material-discursive practice in which the outcomes are heavily dependent on the design and specification of individual assessments.

In sharing platforms, users (consumers and service providers) rate each other, with different rating systems aligned with different roles. The different roles and different perspectives are expressed in ratings as one single objective symbol or representation presented visually as numbers or a quantity of stars, thumbs or grades. Moreover, personal and qualitative comments are weighted less than a quantitative score that the ranking system calculates (Scott and Orlikowski, 2012). Some platforms list the latest comments by default at the top, yet some comments and ratings are hidden elsewhere by the system because the system decides that some comments are not reliable or trustworthy (Aral, 2014). Yelp, for example, has been criticized for deciding which ratings are displayed.

Yelp, like other ranking systems, uses recommendation software. This software sorts ratings based on “quality,” “reliability,” and “activity,” and more active reviewers are calculated in Yelp as more trustful (Yelp, 2013). Thus, in Yelp, active participants/reviewers have a more important voice and therefore have more visibility than those who are not active. Elite users operating at Yelp and similar platforms (Airbnb prioritizes, for example, “superhosts”) gain various exclusive status symbols (e.g., icons of medals and trophies). Smaller, exclusive consumer groups thus have great influence on both the range of user evaluations as well as what is stated in them.

Another challenge the literature on user ratings reveals is how algorithms in ranking systems are designed and constantly changed (Scott and Orlikowski, 2012). The Yelp evaluation software runs continuously and makes adjustments as it secures more information about ratings and reviewers. Yelp, as with other rating systems, therefore changes how ratings are measured and presented to the user by changing their algorithms. TripAdvisor utilizes a popularity index with the intention of restricting fake reviews. The software that TripAdvisor uses seems to be based on the language and usage patterns of reviewers (Scott and Orlikowski, 2012). Rating systems (Mellet, et al., 2014), sharing economy platforms, social networking sites and search engines continually change their algorithms, often without consumers or service providers being aware of these changes (Mittelstadt, et al., 2016).

Trust and trusting others

People use social networking tools to find out whom they can trust and rely on for decisions (Masum and Tovey, 2011). In platforms, “trust-building tools” (Hausemer, et al., 2017) such as rating systems are an essential design component (Pettersen, 2017). Sharing economy platforms are concerned with trust at all levels, in particular with evaluation and payment systems. Trust can be described as a state of desired expectations of other people's actions and intentions (Möllering, 2001). Trust is the basic element that is managed to minimize risk (Coleman, 1988), increase collaboration, reduce social complexity and create order (Coleman, 1990).

The concept of trust has for decades attracted research from various disciplines. Psychologists (e.g., Rotter) have focused on the attributes of individuals and their propensity to trust, while sociologists (e.g., Simmel, Granovetter) have been concerned with relationships among people and institutions, and economists (e.g., Williamson) viewed trust as a calculative decision (Swärd, 2013).

Social capital broadly refers to the recognition and value of social networks and the role of social ties (Ling, 2008). The ability to form and maintain relationships is understood as a basic precondition for the accumulation of social capital (Lin, 1999), where trust is a key concept. Although social capital denote relationships between people that know each other, the construct ‘imagined community’ refer to a socially constructed community, imagined by those who perceive themselves as part of that group without knowing others in person (Anderson, 1991). Social exchange theory argues that social structure is a premise for human action by group members following some commonly agreed rules for interaction (Pettersen, 2012). In social exchange theory, there are two types of exchanges or motivations between individuals: social (intrinsic) and economic (tangible). Social and economic exchanges differ in terms of the types of motivation that initiates exchanges. Social exchange is based on reciprocity, feelings of diffuse obligations, while economic exchange is based on more impersonal premises (Shore, et al., 2006). Trust is important for both types of motivations.

Earlier research considered trust to be personal, while later work regarded trust as relational where constructs of ‘agency’ and ‘reciprocity’ are emphasized (Swärd, 2013). These more practice-theoretical (Bräuchler and Postill, 2010; Feldman and Orlikowski, 2011; Schatzki, et al., 2001) and process-oriented views on trust (Möllering, 2013) move our analytical lens from ‘trust’ to ‘trusting’, stressing relational aspects between actors. Trusting is “a kind of faith” — a mental process consisting of expectations, interpretations and suspension (Möllering, 2001). People will constantly interpret their relationships with others and try to find reasons to trust either in the context or in the actions of their partner (Möllering, 2001; Swärd, 2013). However, there will always be uncertainty related to trust. This is referred to as a “leap of faith” [10].

Botsman’s (2017) construct ‘distributed trust’ denotes a kind of trust between strangers, such as renting an Airbnb listing, entering a car with an unknown Über driver, or when we

are putting our faith in algorithms over humans in our daily lives, whether it’s trusting Amazon’s recommendations on what to read or Netflix’s suggestions on what to watch. (...) We will soon be riding around in self-driving cars, trusting our very lives to the unseen hands of technology. [11]

However, trust in abstract or computer systems is not a new phenomenon. In his discussion of modernity, Giddens (1991, 1984), distinguish between two types of trust: (1) personal trust (building trust in other people); and (2) trusting abstract or expert systems. Expert systems are systems of technical or professional expertise [12]. According to Giddens (1990, 1984) individuals have no other choice than trusting systems developed by experts. For example, in an automobile you enter:

(...) settings which is permeated by expert knowledge, involving the design and construction of automobiles, highways, intersections, traffic lights and many other items. Everyone knows that driving a car is a dangerous activity, entailing the risk of accident (...) I have minimal knowledge of the technicalities of modes of roadbuilding, the maintaining of the road surfaces, or the computers which help control the movement of the traffic. When I park the car at the airport and board a plane, I enter other expert systems, of which my own technical knowledge is at best rudimentary. [13]

According to Giddens (1990, 1984), individuals have no other option than to trust experts systems although they have no knowledge of how they operate.

Botsman (2017) argued that distributed trust explains why people rate “everything from restaurants to chatbots to Über drivers (and why passengers are rated, too)” [14]. Botsman noted that distributed trust may be rewriting the rules of human relationships [15], without actually saying much about how these rules are rewritten or what the implications of this might be, beyond stating that there appears to be a collapse in trust in institutions [16].

Clearly, empirical insights are highly important into the decision-making processes in sharing economy platforms, where individuals create trust and use design elements such as rating scores in this process. Given that some research has discovered that only a small number of consumers generate reviews, and that those reviews largely denote high scores, it is critical to understand how people actually use, trust and interpret ratings when making decisions.

Investigating both consumers and service providers will give us a better understanding of why there is a tendency to give positive scores in rating systems, and in particular in sharing economy services. Insights into how rating scores are provided and interpreted by platform users and the sociological aspects of rating processes are important (Zervas, et al., 2015).

 

++++++++++

Methodology

Research design

Research questions in this study require an exploratory and qualitative case study design because we want to learn more about meaningful processes, aiming for analytical generalizations (Yin, 2012). This research project involved the collection of personal and confidential information and was approved by the Privacy Issues Unit at the Norwegian Social Science Data Service (Norsk senter for forskningsdata, NSD). This approval ensured that collecting, safeguarding, storing and reusing personal data in this study complied with ethical standards and legal requirements.

Case sampling

As a first step in a larger research project in Norway (Pettersen, et al., 2016) about the collaborative or sharing economy, an online list at list.ly was launched in September 2016 [17]. The aim of this list was to secure a better overview of the diversity of these services in the Norwegian marketplace. This overview was an important first step for case sampling. After analyzing list.ly tips on sharing economy platforms (Pettersen, 2017), five sharing economy platforms and one traditional business (a contractor whose business model facilitates sharing economy platforms) was chosen for further research.

The contractor was included in this study because of the new kinds of businesses that emerge as new services in the sharing economy [18]. These businesses facilitate relationships between a service provider and consumer; for example, Easybnb takes care of all the administration when renting out apartments at Airbnb. We wanted to include one of these services in our sample to see whether the customer perceived that they were dealing with a business that was working on behalf of the host and to learn how ratings worked when someone else was facilitating the relationship between customer and host.

Three of the sharing platforms (food production, assets rental and car rental) are Norwegian (which were anonymized to protect young local startups), two (person transportation and apartment accommodation) are American actors (Airbnb and ÜberPop). The contractor represents a new kind of company that facilitates different sharing economy services (similar to Easybnb) and is Norwegian.

Interview sampling

Four of the informants were reached through the snowball method, starting from one of the researcher’s extended (non-direct) network. Four of the platform owners and the professional contractor were contacted directly because they were listed as top managers of the services in Norway. Airbnb was reached via the platform, where one of their public policy staff replied to questions via e-mail. Two informants were contacted due to their unfortunate experiences with a car rental service and accommodation services.

The remaining 11 informants were chosen because they were listed in a given platform, either as offering a product or service or as previous users of a product or service. The members of the latter category had written reviews on a platform, and a limitation in the sample is that we mainly interviewed consumers who had constructed reviews. Despite this methodological shortcoming, we feel confident that the sample renders valid and reliable insights into how users of sharing economy platforms make decisions, how they reason when writing reviews and rating their experiences, and how they built trust in sharing economy platforms.

Data collection

All the platforms and companies in the sample were tested and analyzed by three researchers involved in this project. Profiles were created, and products or services were ordered, completed and paid for, and delivered or returned. Through testing, insights into the platforms gave researchers important domain knowledge and assisted in designing an open and unstructured interview guide. Due to research ethics, none of the researchers wrote reviews or ratings after using services.

Structured and unstructured interviews

Data was collected during September–December 2016 from four structured and 19 unstructured in-depth interviews and participant observations. For each sharing economy platform in the sample, three roles (consumer, peer service provider and platform owner) could be assigned. These three roles were investigated in each platform. By examining different roles in the same platform, it was possible to follow up on insights from other roles in the same platform during interviews. That said, the majority of participants were consumers and service providers (Table 1), two roles that are not mutually exclusive in the peer-to-peer economy (Botsman and Rogers, 2010). Furthermore, although one of the service providers was an individual, he operates as a professional vendor because he has an apartment that he exclusively rents out on Airbnb, and uses a professional agency to take care of administration.

 

Table 1: Interview sample.
Note: Overview of the roles and number of informants in this study’s sample. N = 23. Two of the service providers are represented by a couple renting out a room in their apartment at Airbnb. They are counted as two service providers in the sample.
Consumers8
Service providers8
Professional service-provider (vendor)1
Platform owners5
Professional contractor1

 

More participants were interviewed in two of the platforms (Table 2). We wanted to include different dimensions and customer experiences in the same platform to include participants that had unfortunate experiences to see how they reasoned when rating those experiences. Also, because the professional contractor is a firm that facilitates sharing economy services (e.g., Easybnb) in close collaboration with — in our research — Airbnb, interviews with consumers and service providers dealing with the contractor as a middleman were listed in the Airbnb sample, as shown in Table 2. The interview sample sorted on platforms or business players:

 

Table 2: Interview sample sorted on platforms or players.
Note: Overview of the interview sample sorted by platforms or business players. N = 23. *Because the contractor is a firm that facilitates sharing economy services (e.g., Easybnb) in close collaboration with Airbnb, the interviews with consumers and service providers were listed in Airbnb’s category. Two of the seven individuals listed in the Airbnb sample were informants that either consumed or offered rentals facilitated by a professional contractor.
Airbnb*7
Über4
Food production platform3
Car rental platform5
Assets rental platform3
Professional contractor*1

 

Except for Airbnb, all informants in this study were located in Norway, with the majority in the capital city of Oslo.

Five of the participants were women and 18 men, with ages ranging from 18 to 70 years old. Seven of the interviews were conducted in homes, five in meeting rooms at the research institution, two at a cafe, five by phone, and four participants via e-mail. The four participants that replied via e-mail were characterized by a structured interview format (because e-mail messages do not easily enable follow-up questions on the fly or the ability to interpret tone or oral communicative signals despite several e-mail correspondents included follow-up questions). In addition to interviews, some informants were contacted again by telephone or e-mail with follow-up questions that emerged from data analysis.

The interview guide [19] consisted of open, unstructured questions regarding the specific informant’s role (consumer, service-provider, platform owner). For example, “Can you tell us about your experience of buying/renting/getting/selling/lending out/giving away [the given asset or service]?”, “What determines your decision of buying/renting/getting/selling/lending out/giving away [a given asset or service]?” or “Which role does ratings or reviews play for your decision of completing a deal?” There were also questions related to a specific service that the platform or player facilitates. For example, “Can you tell us a little bit about official marketing policies/food preparation regulations/consumer protection legislations/insurance rules?” or “Who do you think is in charge of the [specific] service you offer?”

The interviews lasted approximately one hour and were carried out by either one or two researchers. All the interviews were recorded and transcribed. Each participant received a gift card as an incentive for their participation.

In addition to the 23 interviews, informal conversations with consumers and peer service providers were completed when different sharing economy services were tested. Insights from these informal conversations and participant observations were collected as ethnographic field notes. These notes played a key role in the data analysis.

 

++++++++++

Data analysis

The data analysis was carried out in three phases:

(a) Researchers read all interview transcripts and coded them individually before holding a two-day workshop where the findings were systematized. The findings were systematized in a total of 44 groups (with more than 1,000 findings or nodes) and sorted by similarity, where some of the categories had one or more sublevel or subgroup (e.g., positive and negative experiences). The next step was to group the findings across roles to reveal patterns independent of roles (such as trust) and to highlight different dimensions of the key findings;

(b) Key findings noted in the field diaries were shared among researchers during the analysis workshop; and

(c) discussed in light of other relevant research and theory.

 

++++++++++

Findings

The decision-making process

As a starting point, the analysis revealed that the consumers and peer service providers followed a social negotiation process consisting of several dimensions when making decisions on whether they would complete a deal or contract with individuals outside their social network on sharing economy capital platforms. The dimensions were elements in what I label the “social trust compass” that both parties use as a navigation tool to minimize uncertainty and risk between the involved parties. The dimensions in the social trust compass are (Figure 1): (a) rating scores, (b) comments, (c) profile pictures and profile texts, (d) up-front communication, (e) honesty, (f) authenticity and (g) a social, personal, and polite tone.

 

The social trust compass
 
Figure 1: The social trust compass. Consumers and peer service providers use the social trust compass to navigate when browsing products and services in sharing economy platforms. The compass consists of dimensions that all play key roles for making decisions in sharing economy platforms.

 

In the following section, I elaborate on the dimensions in the social trust compass.

Rating scores

After selecting the item, date and location, the first thing that participants looked for when researching products or services in sharing economy platforms was previous reviews or rating scores. Scores played a formidable role for all consumers and peer service providers in this study when making decisions about whom to deal with, as one of the consumers of an asset rental platform stated; “Others’ reviews are important because they help to give an expectation if the deal will be smooth and unproblematic”.

This reasoning is in line with the participant using the car rental service;

That [the car] is in decent shape is something I trust due to the reviews. If the cars were crappy that would have been shown in the rating I assume. I have to [believe so]. If the reviews are good [I will trust the vendor]. That’s the only thing I have at hand that I can reason with. I think it’s worse the other way around, that they trust me ...

The participants steered consistently away from users who had previously received negative reviews: “Ratings matter. Especially, negative ratings matter. If a listing has mainly low rating scores and several have been dissatisfied, I will not book the apartment” (consumer of Airbnb). This tendency was also revealed amongst the service providers as the following participants explains:

One has the opportunity to read people’s reviews (...) I’m happy to lend out my assets to people unless they have received negative reviews previously (service provider of an asset rental platform).

Having someone with a negative review staying in our place is out of the question. Because it is our apartment, and we don’t remove things when we lend it out (service provider for Airbnb).

Reviews have therefore a significant influence for both consumers’ and peer service providers’ future opportunities, and were the participants’ first measure for screening options. Hence, participants think that rating scores are trustworthy. Moreover, the existence of reviews is therefore important, and quantity of reviews plays a role. These findings agree with Botsman and Rogers (2010) and Wall (2014), stressing the importance of a critical mass as evidence that a practice is safe and attractive.

Comments

After the first screening, many participants took a closer look at comments left by others having the same role as themselves (consumers or peer service providers). Rating scores and comments seemed to be treated as an account that could be filled and used. If a user had negative reviews, even from a long time ago, that negative impression could be corrected with newer, more positive reviews: “His [the car renter’s] first rating was poor, and then he had four, five ratings after that, which was good. [But this made you skeptical?] Yes it did, of course” (car rental service provider).

More recent reviews and comments are thus most important for consumer and peer service providers. This corresponds well with Dellarocas (2003) who found that the most influential factors in affecting buyer behavior are the overall number of positive and negative ratings and then the number of recently posted negative comments [20]. This is also in accord with Scott and Orlikowski (2012), who pointed out that text comments are subjective and add meaning that a calculated measurement system cannot capture.

Profile photo and profile text

A profile photograph of a person played an important role for many, but not all, participants because it helped individuals create a first impression about the person they are about to engage in a contract. With a profile picture, “One becomes a little acquainted with this person,” as one of the participants explained. One of the peer service providers at Airbnb once canceled a request because the consumer did not have a profile picture or previous reviews (the consumer was a new member of the platform and did not have many experiences).

However, the profile is much more than just a picture. The profile text also played a key role for users in making decisions. The text complemented the photo when the parties were forming a larger opinion about an individual.

Communication up front

If the dimensions described earlier leave a positive impression, an important next step follows: the consumer contacts the other party to make a deal. This process plays a bigger role than one might perhaps think and seems to be crucial in relation to decision-making. Except for Über, in all the other four sharing platforms in our sample, consumers and providers communicated directly with each other before making a deal. This conversation offered a space for interpretation for both parties, where small signs and signals were interpreted, assisting a “gut feeling” over whether to make an agreement or withdraw:

First they fill in date and time for pick-up and return, and then they send a request where they write a small text “Hi, can I rent this and that day.” And then I’ll answer whether the car is available, and “what are you going to do?”. I can have a dialogue back and forth before I approve. And after I have talked with them after a couple of dialogues, I will quickly find out whether I’ll rent it out or not. Gut feeling [makes him turn the deal down]. If they are brief, or if they are not nice sort of, in their writing — if they only say that they need to book it from then to then — that makes me skeptical. (...) Sometimes they have a picture, and I’ll look at the persons, if I feel that they are ... picture is important. You get an impression of how the person is. And what they write [about themselves]. I was so nervous the first time that I went to Facebook and checked out who the person was (car rental service provider).

How people presented themselves and the nature of communications were important indicators for trust.

Social, personal, and polite tone

In sharing economy services, peers tend to expect a more social, personal and polite tone, where one typically may not be as direct or “commanding” as when ordering goods and services in traditional businesses:

In regard of rating, I think it is the politeness we look for the most. And I rate even if they are sitting in the back seat. When you greet and they don’t greet back, that’s a bit ... well, then it’s not sure I’ll rate five. Because then you’ll see ... and when some says “Don’t follow the GPS, I’ll tell you where to drive,” a bit instructive. Then you’ll think that the contact you get is a bit like ... that the person [is not very nice] (Über service provider).

On questions of whether service providers had negative experiences with guests or customers, they explained that up-front communication assisted in minimizing it:

No, there has never been any problems. One gets a certain feeling of people when they write to us. We have said no to a few, but that is seldom. (...) we read the person’s profile description and think a bit, and the message they write to us is probably the most important [for accepting or not accepting the guest] (Airbnb service provider).

Thus, the person is an important part of the transaction, and as will be shown later in this section, this personal dimension is often the one that was evaluated and rated. Initial courtesy and humbleness in dialog signaled to the peer service provider whether a consumer would take care in a given object that they rented. However, communication is closely related to norms and rules and are not universal rules (Giddens, 1984). Thus, what is considered “polite” and “personal” will be interpreted differently by those holding different communication practices.

Honest and authentic

Through communication prior to a deal, all parties grew confident that they would make a deal corresponding to expectations. Honesty was therefore a key dimension in the social trust compass with regard to expectation management. As one of the participants noted, “In the sharing economy, it is very important to be honest about what you deliver.”

Communication was also important in regard to a consumer’s expectation that a product or service was authentic and that the parties were peers and not a business, unless clearly communicated in advance. Consumers and providers expected that they were dealing with those parties that they initially communicated with. Thus, consumers and peer service providers expected that an individual, and not a company, would hand over keys or lend a film projector and so on. If others, such as a neighbor, friend, or a professional intermediary, provided assistance, customers thought it was all right as long as it was communicated in advance.

Meanwhile, we also found that not all professional services were honest about their roles. Customers interpreted this dishonesty as unfortunate when they anticipated that they were dealing with a peer. When using sharing economy platforms, customers sought authentic experiences and exchanges. As one of the car rental platform consumers stated; “The whole point with sharing economy services is that it is a private person you are dealing with [If there was a business] it would suck big time. That would have been stupid.”

Consumers’ interpreted that sharing economy platforms were used by peers, and not by companies, and they entered into a relationship with a different attitude than when buying products or services from traditional businesses. This meant that consumers often ignored elements that they might complain about when dealing with traditional companies and not individuals. In many ways, the social dimension of the relationship between the consumer and the provider moved trade from a business relationship to a friend relationship.

Consumers consistently had lower expectations about service and appeals when using sharing services compared to traditional businesses. If something did not match an expectation in sharing platforms, the user placed further importance on the social relationship and was willing to overlook certain elements because the relationship was perceived as social and not strictly an economic exchange. This was a key element in understanding the ratings that all parties provided.

When professional services entered the sharing economy platform, consumers used the same social trust compass as a navigation tool in making decisions and followed the same reasoning in creating reviews and rating products or services. However, we found that consumers did not always understand that they were dealing with a professional service provider and therefore rating an intermediary (contractor) rather than the actual professional provider. The rating score, however, was listed on the professional service provider’s site.

The dimensions in the social trust compass all played key roles in building trust among people outside their social network. Honesty and good reviews led to increased confidence. One of the consumers explained that he once got a feeling when buying a service that the provider did not pay tax, which led him to mistrust that person. He withdrew from making the deal.

Rating people, not products or services

The analysis found that the reviews left and gained by participants in all of the platforms studied were mainly positive:

I think I have driven more than 300,400 rides. And 10 has rated three, or was it four. I have only four and five (...) No twos and no negative comments. And approximately 50 have left a positive comment (Über service provider).

We try to, and we feel that we do a good job, and we like this very much. So we have very good reviews. Most often full score, 90 percent full score I think, so that is very pleasant (Airbnb service provider).

According to Resnick and Zeckhauser (2002), people wrote reviews due to civil duty, courtesy, gratitude and reciprocity. These are all social norms related to social aspects with others. Also Zervas, et al. (2015) asked if sociological factors might explain why ratings at Airbnb were very high on average. This study found that ratings were closely related to sociological factors. For example, one of the customers of the food production platform stated the following when asked why he wrote a review after buying a grilled cheese sandwich:

Because I wanted to recommend it to others. Because it tasted very good and well, it surprised me actually how good it was. And I wanted to give a positive response to the chef, and to the platform, so they will continue the service. Because it was a good thing.

The participant pointed to reciprocity when he explained why he wrote a review. Similarity, a consumer of Über explained reasoning over ratings:

[Ratingscore is] based on the conversation, the driver’s looks, was he nice, was he clean, how was the music, did he drive correct, those kind of things. I rate. From three to five I’d say. Three if, for example the car is located in the wrong street, if the driver doesn’t talk much, then it’s a three. I’ll give five if he is fast, very nice, very funny and such, making jokes (consumer of Über).

The participant pointed primarily to personal characteristics and to dimensions in the relationship. This reasoning illustrated a consistent pattern in this study, namely that participants rated their relationships rather than a product or a service in isolation. Because payment of an asset or service is facilitated by the platform and not parties, the economic aspects in deals was also removed from the relationships between the parties. As noted earlier, social and economic exchanges differed in terms of the type of motivation that initiated an exchange. Social exchange is based on feelings of diffuse obligations and reciprocity, while economic exchange is based on more impersonal premises (Shore, et al., 2006).

Many consumers and service providers tended to withdraw from submitting ratings rather than give negative reviews. When a consumer was expected to receive a negative review, the service provider rated the person politely, simply because they were dealing with individuals, not businesses. Except from one of the participants, most people in this study did not rate another party negatively unless a major breach of expectations and the contract between the parties took place. Participants typically explained that “everyone can have a bad day or be misunderstood.” A consumer of a car rental platform explained how he was unfortunate to crash a rental car:

It was of course very awkward (laughter), and difficult to bring up [to the lessor]. (...) But it was nothing else to do but tell it the way it was. And it went very well, he was an understanding person and not a grumpy kind of person that got really pissed (...). I showed him the damage and explained what happened (...) And he brought up a story about how he once crashed his father in law’s car (laughter) (...) Then we thanked each other for the tenancy (laughter)

Later the unfortunate consumer received a note from the platform stating that the service provider had written a review of him, and that he was encouraged to write one as well. The consumer explained that the service provider’s review was

All right, he wrote he would be more than happy to lend me his car again another time, but that he had to give me a somehow poorer rating due to the accident (a little laughter). So, that was a bit ..., ok he mentioned it, but it was a bummer, but that’s how it is.

However, even when the trade has been unfortunate “it is bloody difficult to leave a poor rating”, one other participant explained. People did not want to destroy other’s reputations and credibility by writing negative elements in a review. The content in ratings typically denoted that communication was good throughout the process, “everything went fine,” and people were nice, trustworthy and honest, were typical statements written about other parties.

At the same time, many said that they were keen to write honest reviews because they wanted to help others in the same roles as themselves. Consumers saw themselves in a community with other consumers, while peer service providers identified with other peer service providers. Thus, the motivations behind ratings were done on behalf of a community that they identified with, or an imaginary social community, to use Anderson’s (1991) terminology.

Consumers and service providers stressed that it was important that reviews should not be deleted by a given platform, pointing to authenticity and censorship. It is a universal practice that ratings are never deleted or removed (Scott and Orlikowski, 2012) unless they have racist or harassing characteristics. In the interviews, the platform owners confirmed they would not delete or remove reviews unless they had harassing characteristics. Both platform owners and users of sharing services in this study did not think ratings and reviews should be removed by request, as this could be interpreted as censorship or cheating, implying that the service therefore would not be authentic and reliable. Deleting one’s profile could therefore be a strategy that users might adapt to manage bad reviews. Ratings could quickly become counterproductive if those with negative reviews deleted their profiles and instead created new ones. That scenario could also assist in creating a homogenous rating picture.

Objective measures of subjective experiences

Many of the participants said they could live with another party having received one or more negative ratings when they were balanced with more recent and positive reviews. This has implications for the design of rank in rating or evaluation systems. Rating systems however, are modelled on quantitative measures which ignores qualitative aspects found in comments. Those comments add important contextual details to a quantitative score. In this sense, rating systems shows only one side of a two-sided evaluation coin. The unfortunate driver, for example, did not get a high rating score from the vendor, yet the vendor added important details as a comment, namely that he would be happy to lend him a car again at a later date.

Both constructing and interpreting reviews and ratings are cognitive, social and subjective processes closely related to practices at play in the context in which the person is part of. What is recognized as ‘polite’, ‘nice’ or ‘good’ for example, do not hold universal rules. These are practices closely related to the context in which these practices are part of (Giddens, 1984; Pettersen, 2016). It is problematic that rating scores are objectively measured and presented in the form of a grading system or quantitative ranking (Scott and Orlikowski, 2012). How ratings and reviews are represented in sharing platforms, however, are in control of the platforms, and decided by platform algorithms.

Algorithms in rating scores

Algorithms determine how ratings influence how a product or service is presented in a platform. One of the participants who used Airbnb on a professional level explained how the platform changed review categories and how they were weighted:

Airbnb has changed the way guests give their feedback. Before, location was described in the profile. And if it was correct, the guest gave five. Four if there were some things one had forgotten to mention, and so on. But now it’s like; “Did you feel safe at this location?,” “What do you think of the apartment’s location?,” “Was it cozy?,” “How was the people around you?,” and so on. That kind of thing. And it’s completely irrelevant to me as a host. So it [the categories the users must fill out in the review] does not concern me and my apartment, but is related to the area. (...) Suddenly, I did not have five out of five anymore at all. And then I rented an apartment myself once, and then I got the review categories [mentioned above] (...) Before I had a full score. Now I have 4.6 or something like that. So it has not made any dramatic difference, but it did mean that I am no longer a super-host, which enables you to be on top of the search results and ranks.

As the informant illustrated, Airbnb changed feedback categories and how feedback was weighted. In this case, it was in the informant’s disfavor because his status as super-host disappeared. In addition, the informant showed which benefits super-hosts secure in regard to visibility. Being ranked high in search lists in sharing economy platforms was calculated based on past user ratings, possibly issued under entirely different conditions than today. Algorithmic accountability is important (Mittelstadt, et al., 2016; Raine and Anderson, 2017), because rating systems are neither fully reliable nor transparent (Hausemer, et al., 2017). Better visibility in search results, exclusive status symbols within a platform and invitations from a platform to exclusive events are some of the advantages as a super-host at Airbnb. The participants in this study, however, thought rating systems were trustworthy and used rating scores as a first screening and starting point in their searches for services.

Ratings are very important for platform owners and they interfere in varying degrees if service providers receive bad reviews. Several of the platforms in this study have very high demand on the service providers’ ratings. For example, one of the platforms requires an average of four out of five grades. Another platform requires an average of 4.8 to earn a super-provider status. Service providers at Über explained that expulsion are used when certain thresholds of good reviews were not met. Airbnb was strict over cancellations again with sanctions triggered by a certain threshold. All of the platforms in this study provided advice on how to earn good reviews, such as responding to inquiries quickly, not altering agreements, providing good service and using profile pictures. Pressure from the platforms to gain high review scores might explain why the average scores were very high.

 

++++++++++

Discussion and conclusions

To return to the first research questions discussed in this paper — How do people reason when they choose one product or service over the other in sharing economy platforms? — the findings illustrate that previous ratings play a key role in decision-making. Although several of the participants questioned the accountability of ratings, they trust rating systems devised by platforms. This accords well with Giddens’ (1990, 1984) statement that individuals have no other option than to trust expert systems. None of the informants in this study could imagine sharing economy platforms without rating mechanisms, indicating the importance of using other’s (previous) experiences as a compass. Parties entered a social negotiation process consisting of several dimensions in a social trust compass (see Figure 1).

The social trust compass is used by both consumers and peer service providers as a navigation tool in dealing with those outside their social network to minimize uncertainty and risk and facilitate decisions. Rating scores represent the first step in a screening process, followed by a close reading of comments and profiles. If the result is a positive impression, a consumer contacts a service provider. This is an essential phase for decision-making. However, this study also found that platforms control and change rating categories with subsequent consequences. Thus, evaluation systems are neither fully reliable nor transparent, in agreement with Hausemer, et al. (2017). Algorithmic transparency and functionality of rating systems are important because these systems play key roles for users in making decisions.

A participant in a given platform expects the other party is authentic, that is a private person and not a formal business. The same reasoning would likely follow if users were communicating with non-humans (bots) masking themselves as human. In this study, engaging in relationships with strangers in sharing economy platforms is not rewriting rules of human relationships, as suggested by Botsman (2017). On the contrary, the process of growing trust follows established practices and different trust-building phases, where ratings are only one of several aspects that is taken into consideration.

Components in the social trust compass play a key role for all participants in this study in making decisions. There were exceptions, for a professional service provider and a business contractor facilitating sharing economy services. In those cases, their goal was not social but solely economic.

In terms of the second research question, participants generally rated positively, providing high scores unless something unexpected occurred. A negative review was only generated when things went terribly wrong. Even in these cases, participants in this study found it difficult to write a negative review because of personal and social dimensions. Participants appeared not to rate products or services, but relationships. This was observed due to moving our analytical lens from trust to trusting (Giddens, 1984; Möllering, 2013; Schatzki, et al., 2001; Swärd, 2013). Ratings ignore more subjective and two-dimensional social relationships between parties. Thus, a rating score is a symbol of a social relationship, where only one side of the rating coin is represented in evaluation systems. Hence high rating scores in reputation systems and sharing economy platforms were closely related to altruistic behavior, social motivation and reciprocity.

The social dimension between consumers and service providers moves trade from business relationships to friendships. A a result, consumers were consistently entering sharing services with lower expectations about service, appeals and standards. These findings support assumptions by Zervas, et al. (2015) in which “individuals rate other individuals differently or more tactfully, than they rate firms such as hotels, independent of the platform” [21].

The introduction of Web 2.0 is often associated with increasing consumer participation, consumer power and review diversity (Benkler, 2006; Blank, 2006). However, research on online ratings (Aral, 2014; Luca and Zervas, 2016; Mellet, et al., 2014; Zervas, et al., 2015) has found that reviews were mainly positive. This paper found that there were personal and social dimensions at play in ratings. These dimensions points to reasons for why ratings do not follow a normal statistical distribution.

Individuals have lower expectations in peer-to-peer sharing economy services, shedding light on high ratings. When service providers receive low scores or ratings, platforms provide both advice and sanctions, putting pressure on the service providers for positive reviews. Social mechanisms were also important when service providers asssessed their customers.

Consumers need to be more aware about social mechanisms at play when peers rate each other. They also need to better understand how platforms control and change ratings and how rank is measured. The findings illustrate the importance of algorithmic transparency in evaluation systems, as noted by Scott and Orlikowski (2012) and Raine and Anderson (2017).

This study is not without limitations. Only users of sharing platforms in Norway were studied. Further research should study how these findings correspond to users in other countries with participants that hold different values and communication practices. Future studies should also examine how individuals reason when they rate businesses rather than peers. End of article

 

About the author

Lene Pettersen is Senior Researcher at Consumption Research Norway (SIFO), Oslo and Akershus University College of Applied Sciences (HiOA).
E-mail: lene [dot] pettersen [at] sifo [dot] hioa [dot] no

 

Acknowledgements

This work was supported by The Ministry of Children and Equality in Norway (BLD) [project numbers 11201609, 11201619]. I would like to thank researcher Ingrid Kjørstad and research assistant Thea Grav Rosenberg — both from SIFO, HioA — for conducting several of the interviews and that took part in the data analysis. Thanks to First Monday’s anonymous reviewers of this paper for constructive comments.

 

Notes

1. Möllering, 2001, p. 404.

2. Botsman, 2017, p. 10.

3. Botsman, 2017, p. 4.

4. Hausemer, et al., 2017, p. 14.

5. Botsman, 2017, p. 10.

6. Thanks to Assistant Professor Nils S. Borchers at the Institute of Communication and Media Studies at the University of Leipzig for pointing this distinction out.

7. Slee, 2013, p. 5.

8. Zervas, et al., 2015, p. 3.

9. Edelman and Luca, 2014, p. 10.

10. Swärd, 2013, p. 16.

11. Botsman, 2017, p. 4.

12. Giddens, 1990, p. 27.

13. Giddens, 1990, p. 28.

14. Botsman, 2017, p. 8.

15. Botsman, 2017, p. 10.

16. Botsman, 2017, p. 252. Botsman (2017) seems to generalize characteristics from an American context, with only 30 percent trusting the government. However, there are many countries in 2017 that hold high levels of trust to their country’s government; India scores 73 percent trusting their government (Hausemer, et al., 2017; McCarthy, 2017).

17. See Pettersen (2017) for details of the list.ly list at http://dx.doi.org/10.5210/fm.v22i18.7805.

18. See Pettersen (2017) for further elaboration.

19. The full interview guide — in Norwegian — can be found in Pettersen, et al. (2016, pp. 201–206), at http://www.hioa.no/Om-HiOA/Senter-for-velferds-og-arbeidslivsforskning/SIFO/Publikasjoner-fra-SIFO/Delingsoekonomi-et-kvalitativt-oeyeblikksbilde-fra-Norge.

20. Dellarocas, 2003, p. 1,412.

21. Zervas, et al., 2015, p. 3.

 

References

Benedict Anderson, 1991. Imagined communities: Reflections on the origin and spread of nationalism. Revised and extended edition. London: Verso.

Sinan Aral, 2014. “The problem with online ratings,” MIT Sloan Management Review, volume 55, number 2, at https://sloanreview.mit.edu/article/the-problem-with-online-ratings-2/, accessed 28 November 2017.

Yochai Benkler, 2006. The wealth of networks: How social production transforms markets and freedom. New Haven, Conn.: Yale University Press.

Grant Blank, 2006. Critics, ratings, and society: The sociology of reviews. Lanham, Md.: Rowman & Littlefield.

Rachel Botsman, 2017. Who can you trust? How technology brought us together and why it might drive us apart. New York: PublicAffairs.

Rachel Botsman and Roo Rogers, 2010. Whats mine is yours: The rise of collaborative consumption. New York: Harper Business.

Birgit Bräuchler and John Postill (editors), 2010. Theorising media and practice. New York: Berghahn Books.

Judith A. Chevalier and Dina Mayzlin, 2006. “The effect of word of mouth on sales: Online book reviews,” Journal of Marketing Research, volume 43, number 3, pp. 345–354.
doi: https://doi.org/10.1509/jmkr.43.3.345, accessed 28 November 2017.

James S. Coleman, 1990. Foundations of social theory. Cambridge, Mass.: Belknap Press of Harvard University Press.

James S. Coleman, 1988. “Social capital in the creation of human capital,” American Journal of Sociology, volume 94, pp. S95–S120.
doi: https://doi.org/10.1086/228943, accessed 28 November 2017.

Chrysanthos Dellarocas, 2003. “The digitization of word of mouth: Promise and challenges of online feedback mechanisms,” Management Science, volume 49, number 10, pp. 1,407–1,424.
doi: https://doi.org/10.1287/mnsc.49.10.1407.17308, accessed 28 November 2017.

Benjamin G. Edelman and Michael Luca, 2014. “Digital discrimination: The case of airbnb. com,” Harvard Business School (HBS) Working Paper, number 14–054, at https://hbswk.hbs.edu/item/digital-discrimination-the-case-of-airbnb-com, accessed 28 November 2017.

Benjamin G. Edelman, Michael Luca and Daniel Svirsky, 2016. “Racial discrimination in the sharing economy: Evidence from a field experiment,” American Economic Journal, volume 9, number 2, pp. 1–22.
doi: https://doi.org/10.1257/app.20160213, accessed 28 November 2017.

Asle Fagerstrøm, Sanchit Pawar, Valdimar Sigurdsson, Gordon R. Foxall and Mirella Yani-de-Soriano, 2017. “That personal profile image might jeopardize your rental opportunity! On the relative impact of the seller’s facial expressions upon buying behavior on Airbnb,” Computers in Human Behavior, volume 72, pp. 123–131.
doi: https://doi.org/10.1016/j.chb.2017.02.029, accessed 28 November 2017.

Martha S. Feldman and Wanda J. Orlikowski, 2011. “Theorizing practice and practicing theory,” Organization Science, volume 22, number 5, pp. 1,240–1,253.
doi: https://doi.org/10.1287/orsc.1100.0612, accessed 28 November 2017.

Anthony Giddens, 1991. Modernity and self-identity: Self and society in the late modern age. Stanford, Calif.: Stanford University Press.

Anthony Giddens, 1990. The consequences of modernity. Cambridge: Polity Press.

Anthony Giddens, 1984. The constitution of society: Outline of the theory of structuration. Berkeley: University of California Press.

Pierre Hausemer, Julia Rzepecka, Marius Dragulin, Simone Vitiello, Lison Rabuel, Madalina Nunu, Adriana Diaz Rodriguez, Emma Psaila, Sara Fiorentini, Sara Gysen, Tim Meeusen, Simon Quaschning, Allison Dunne, Vadim Grinevich, Franz Huber and Linda Baines, 2017. “Exploratory study of consumer issues in online peer-to-peer platform markets,” European Commission (EUR 2017.4058 EN), at http://ec.europa.eu/newsroom/document.cfm?doc_id=45245, accessed 28 November 2017.

Nan Lin, 1999. “Social networks and status attainment,” Annual Review of Sociology, volume 25, pp. 467–487.
doi: https://doi.org/10.1146/annurev.soc.25.1.467, accessed 28 November 2017.

Rich Ling, 2008. New tech, new ties: How mobile communication is reshaping social cohesion. Cambridge, Mass.: MIT Press.

Michael Luca and Georgios Zervas, 2016. “Fake it till you make it: Reputation, competition, and Yelp review fraud,” Management Science, volume 62, number 12, pp. 3,412–3,427.
doi: https://doi.org/10.1287/mnsc.2015.2304, accessed 28 November 2017.

Hassan Masum and Mark Tovey (editors), 2011. The reputation society: How online opinions are reshaping the offline world. Cambridge, Mass.: MIT Press.

Niall McCarthy, 2017. “Which countries have the most and least confidence in their governments?” Forbes (13 July), at https://www.forbes.com/sites/niallmccarthy/2017/07/13/which-countries-have-the-most-and-least-confidence-in-their-governments-infographic/, accessed 28 November 2017.

Kevin Mellet, Thomas Beauvisage, Jean-Samuel Beuscart and Marie Trespeuch, 2014. “A ‘democratization’ of markets? Online consumer reviews in the restaurant industry,” Valuation Studies, volume 2, number 1, pp. 5–41.
doi: https://doi.org/10.3384/vs.2001-5992.14215, accessed 28 November 2017.

Brent Mittelstadt, Patrick Allo, Mariarosaria Taddeo, Sandra Wachter and Luciano Floridi, 2016. “The ethics of algorithms: Mapping the debate,” Big Data & Society, volume 3, number 2.
doi: https://doi.org/10.1177/2053951716679679, accessed 28 November 2017.

Guido Möllering, 2013. “Process views of trusting and crisis,” In: R. Bachmann and A. Zaheer (editors) Handbook of advances in trust research. Cheltenham: Edward Elgar, pp. 285–306.
doi: http://dx.doi.org/10.4337/9780857931382, accessed 28 November 2017.

Guido Möllering, 2001. “The nature of trust: From Georg Simmel to a theory of expectation, interpretation and suspension,” Sociology, volume 35, number 2, pp. 403–420.
doi: https://doi.org/10.1177/S0038038501000190, accessed 28 November 2017.

Lene Pettersen, 2017. “Sorting things out: A typology of the digital collaborative economy,” First Monday, volume 22, number 8, at http://firstmonday.org/article/view/7805/6513, accessed 28 November 2017.
doi: http://dx.doi.org/10.5210/fm.v22i18.7805, accessed 28 November 2017.

Lene Pettersen, 2016. “The role of offline places for communication and social interaction in online and virtual spaces in the multinational workplace,” Nordicom Review, volume 37, pp. 131–146.
doi: https://doi.org/10.1515/nor-2016-0028, accessed 28 November 2017.

Lene Pettersen, 2012. “The echoing paradox in SNS,” ISM 2012 Workshop Proceedings, pp 7–15.

Lene Pettersen, Ingrid Kjørstad and Thea G. Rosenberg, 2016. “Delingsøkonomi — et kvalitativt øyeblikksbilde fra Norge (The sharing economy — A qualitative snapshot from Norway),” at http://www.hioa.no/Om-HiOA/Senter-for-velferds-og-arbeidslivsforskning/SIFO/Publikasjoner-fra-SIFO/Delingsoekonomi-et-kvalitativt-oeyeblikksbilde-fra-Norge, accessed 28 November 2017.

Lee Raine and Janna Anderson, 2017. “Code-dependent: Pros and cons in the algorithm age,” Pew Research Center (8 February), at http://www.pewinternet.org/2017/02/08/code-dependent-pros-and-cons-of-the-algorithm-age/, accessed 28 November 2017.

Paul Resnick and Richard Zeckhauser, 2002. “Trust among strangers in internet transactions: Empirical analysis of eBay’s reputation system,” In: Michael R. Baye (editor). The economics of the Internet and e-commerce. Amsterdam: JAI, pp. 127–157.

Theodore R. Schatzki, Karin Knorr Cetina, and Eike von Savigny (editors), 2001. The practice turn in contemporary theory. London: Routledge.

Susan V. Scott and Wanda Orlikowski, 2012. “Reconfiguring relations of accountability: Materialization of social media in the travel sector,” Accounting, Organizations and Society, volume 37, number 1, pp. 26–40.
doi: https://doi.org/10.1016/j.aos.2011.11.005, accessed 28 November 2017.

Lynn M. Shore, Lois E. Tetrick, Patricia Lynch and Kevin Barksdale, 2006. “Social and economic exchange: Construct development and validation,” Journal of Applied Social Psychology, volume 36, number 4, pp. 837–867.
doi: https://doi.org/10.1111/j.0021-9029.2006.00046.x, accessed 28 November 2017.

Tom Slee, 2013. “Some obvious things about Internet reputation systems” (29 September), at http://tomslee.net/2013/09/some-obvious-things-about-internet-reputation-systems.html, accessed 6 October 2015.

Anna Swärd, 2013. “Trust processes in fixed-duration alliances: A multi-level, multi-dimensional, and temporal view on trust,” at http://web.bi.no/forskning%5Cpapers.nsf/wSeriesDissertation/2DCC82BCDDDC690BC1257B59003CD6E0, accessed 28 November 2017.

Yelp, 2013. “Why does Yelp recommend reviews?” YouTube (25 January), at https://www.youtube.com/watch?v=PniMEnM89iY, accessed 28 November 2017.

Robert K. Yin, 2012. Applications of case study research. Third edition. Thousand Oaks, Calif: Sage.

Derek Wall, 2014. The sustainable economics of Elinor Ostrom: Commons, contestation and craft. London: Routledge.

Georgios Zervas, Davide Proserpio and John Byers, 2015. “A first look at online reputation on Airbnb, where every stay is above average” (12 April), at http://www-bcf.usc.edu/~proserpi/papers/airbnbreputation.pdf, accessed 28 November 2017.

 


Editorial history

Received 29 May 2017; revised 17 November 2017; accepted 26 November 2017.


Creative Commons LicenseThis paper is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Rating mechanisms among participants in sharing economy platforms
by Lene Pettersen.
First Monday, Volume 22, Number 12 - 4 December 2017
http://firstmonday.org/ojs/index.php/fm/article/view/7908/6586
doi: http://dx.doi.org/10.5210/fm.v22i112.7908





A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2017. ISSN 1396-0466.