Week 4: Couch Surfers and Hobbo Hotel

This week I looked at two different online social communities and compared the mechanisms of social capital/trust. The first community I chose was “Couch Surfing,” a network for travelers to meet and host other travelers. The second one I chose was “Habbo Hotel”, a virtual world in which a user controls an avatar that owns a room in a hotel.

Social capital has many different definitions, so for simplicity I focused mainly on Putnum’s definition of Social Capital (from Bowling Alone): “The collective value of all social networks and the inclinations that arise from these networks to do things from each other.” With respect to this definition I examined how these two communities encouraged trust, reciprocity, information and cooperation. Williams also argued that social capital is cyclical and is not the network itself, but rather social capital is generated by the network.

Before I begin any analysis I should give a little more background about both of these groups. Couch Surfing is an international network that has the goal of increasing global connections by offering free accommodation to travelers (ie. letting them sleep on your couch) with the hope that people will form friendships based on these real life hosting experiences. It’s a site where people join with the expectation of meeting other people on the site face to face. Because there is a lot of trust involved in letting people stay in your house, this site has interesting mechanisms including rating systems, evaluations, etc…. If you want to visit somewhere you can search by location for couch surfers in the nearby area who are available to host and request that they host you. Users have the choice to respond to your request and accept you or not.

I found the profile of a professor of a class in which I'm the TA. It’s kind of weird to find out she knows fencing or that she’s been to over 50 different countries through Couch Surfing.

Unlike Couch Surfing, Hobbo Hotel doesn’t have a mission statement or defined purpose. It’s a virtual environment in which users can interact with other users via an avatar. You start off with a customizable avatar that owns one room in a large hotel. You can decorate your room, visit other people’s rooms, visit other hotel rooms, play games, chat with other users,… among many other features. One of the main features in Hobbo Hotel is buying furniture and decorating your room. There are two types of currency in the game: Credits which are bought with real money or sometimes won through events, and pixels which are received by accomplishing tasks and satisfying certain conditions. Habbo Hotel encourages users to explore the different features of the world by giving badges and pixels to users for participating in events.

My first interaction in Hobbo Hotel with a guide. I did ask him questions about Hobbo Hotel but I was nervous when talking to him.

To examine social capital I think it’s important to look at the different social roles that users play within each community. Gleave et al. stated that the creation and maintenance of social spaces depends on the complex social ecology created by the interaction of several social roles. In Couch Surfing there are two major roles: the host and the hosted. Users will switch roles at times (sometimes hosting, sometimes being hosted), but much of the interaction of the event will take place offline. Initial contact might occur online, but like dating websites, the contact is meant to extend to face to face interactions. Online participation comes in the form of evaluating an experience or interaction with the person that hosted you or that you hosted. Another role in the community is that of a friend, ie. a person who can vouch for you and verify that you’re not a serial killer. All of these roles serve to increase the amount of social capital or trust within the couch surfing community. Generous hosts and gracious couch surfers make the experience of hosting someone positive, and encourage a growing network of friendships. Recommendations and positive reviews encourage civil behavior between people.

In Hobbo Hotel, I quickly noticed two types of social roles upon signing up. When I first began, I was given the option of asking for a guide to help me out since I was a first time user. I was in the role of a newbie and my guide, a more experienced Hobbo, helped me out by giving me some information about how the world works. This serves to ease the transition of a first time user and encourages helping others. I was rewarded 20 pixels for receiving help and I believe that my guide was also rewarded pixels for helping me out. This direct, point based incentive improved social capital within the community. I witnessed several people playing games, so another role could be a teammate. I noticed people also buy, sell, and trade items within the game. There was a host/visitor relationship in Hobbo Hotel as well although it was not as invested a relationship as the ones in couch surfing. Some users would open up their rooms for other people to hang out in. My room is only rated 3 points and all it has is a standard desk and chair, a lamp and a plant. I visited a few rooms with Disco Balls, intricate sets of furniture, flashy lights, etc… Although I feel that the decorating is more of an economic issue (a way for Sulake Co. to make money off of credit purchases) one could argue that the accumulation of furniture promotes social capital. People are encouraged to express themselves through decoration which leads to them exploring other rooms, meeting people, establishing relationships, etc….

Some other guy's room. It's pretty nice compared to mine

Putnam mentions two different types of social capital, “bridging” and “bonding.” Bridging is considered more inclusive and occurs when individuals of different backgrounds create tentative relationships. It’s more about breadth as opposed to depth. Bonding is considered more exclusive and people have stronger connections like those between close friends and family. When attempting to measure social capital in these two communities, one should take into account the differences between strong emotional ties between close friends and weak ties between acquaintances. Williams offers a matrix with online bonding, offline bonding, online bridging and offline bridging to separate and compare different types of social capital.

I think that when measuring social capital with respect to couch surfing, one must take into account the objective of the site. For instance, Couch Surfing encourages meeting with other users in real life so we can’t say the experience of a user is purely online. With Couch Surfing there is an online and an offline component to social capital. In comparison, I feel more comfortable arguing that Hobbo Hotel doesn’t have an offline component as there is nothing to suggest that Hobbo Hotel promotes offline social capital between members. Both sites encourage the building of bridging social capital, but I believe that only Couch Surfing offers a bonding experience. If we look at the criteria Williams used to access the degree of online bridging, I think both sites do a good job in promoting meeting a broad range of people, and meeting someone new online. Habbo Hotel is more likely to use chat room functions (I believe), but Couch Surfing has more external links to information. However, the bonding measures greatly favor Couch Surfing. I don’t think Habbo Hotel is a good means of keeping in touch with someone far away, but Couch Surfing seems to encourage it, especially after a hosting experience. With Couch Surfing, there also is more of a trust factor as people share a lot of personal information in their profiles to try and find the best matches for a host/ hosted relationship. I noticed many people use their real names and full names in their profile. Age, languages spoken, likes, dislikes, pictures, descriptions of your house, etc… are all on display. Habbo Hotel is more anonymous. Most users do not use their full name and you don’t really know what they look like, where they’re from, what they like or dislike, etc…

In his article on Trust and Modeling in E-Services, Massa summarizes several features of Couch Surfing which he categorized as a Social/Entertainment site. One of the mechanisms that he mentions that supports social capital is the users expressing level of trust in other users. In this way you not only know a user’s friends but also the depth of their relationship. Another mechanism of trust is activity history in which other people can view who the user hosted, by whom the user was hosted, and evaluation comments of those experiences by other users. You can also verify actual physical location through exchange of mail with administrators and also verification of personal identity through credit card payments.

I don’t know how Massa would classify Hobbo Hotel, but I don’t think it fits in with his other classifications. For one thing, Hobbo Hotel is not a website. Like Second Life, Hobbo Hotel is a virtual environment. I would still classify it under Social/Entertainment. One mechanism that supports social interaction is having built in activities in the environment. For example, there are games people can play with others in the hotel. This gives people a reason and opportunity to meet and interact with new people. There are also in game incentives (pixels) and badges that mark social worth. I noticed that I would periodically get pixels in the game for no reason, and I think that you’re rewarded just for being logged in and walking around the hotel. I also noticed some people doing character role playing. For instance I saw one character who said she was looking for a “boyfriend” in her avatar description. I’ve also seen advertisement that said “looking for a daughter” while browsing some events. It seems like people do role playing and create imaginary social situations just to have something to do.

Overall I feel that Couch Surfing does a good job in establishing mechanisms of trust. There are a few things I would suggest to ensure that users don’t take advantage of others. One thing I noticed in Couch Surfing is that there is no mechanism that prevents people from only coach surfing and never hosting. The Allen article on social networks of trust touched on the idea that often times what’s best for the individual is not what’s best for the social capital of the community.

The classic cooperation paradox applied to CouchSurfing: Being selfish (ie. Only couch surfing and never hosting) is the most gain (financially speaking) but the collective outcome is socially irrational (if everyone wanted to travel but no one hosted, then the system would fall apart.)

I’ve seen some other social networking sites like Book Mooch that implemented a system to prevent people from only being hosted and never hosting. Book Mooch is a site in which users can request to receive and ship out books. On that site, a user must send at least 1 book for every 2 that he or she receives. This ensures that a user doesn’t simply receive a ton of books without shipping any out. A similar system could be implemented on Couch Surfing in which a user must host at least 1 person for every 2 times he or she gets hosted by someone else. Couch Surfing explicitly stated that even people who can’t host (ie. Don’t have an available place for others to stay) can still sign up provided that they anticipate being able to host in the future. While not making hosting a requirement shows a lot of trust in the users, ultimately it may cause people to take advantage of others.

Another potential problem I saw with Couch Surfers is that it may be difficult to initially engage in interactions with others because you don’t have any starting credentials. For example, if I just joined Couch Surfers and didn’t know anyone else on the site, I believe it would be very difficult to establish relationships. One of the good features of Couch Surfer is the ability to comment on your hosting experiences with other members. However, people use this as criteria to determine someone’s trustworthiness. While this is great for people who have hosted and proven themselves to be good travelers, it could be discouraging for people who are just starting and don’t have such credentials.

For example when I joined Couch Surfers I thought about how difficult it would be for me to host someone. Honolulu has a lot of users and why would someone want to stay with me, when he or she could stay with a person with a lot of positive reviews (like my high school friend Shawn who I found on Couch Surfers). I also found a professor that I work for on Couch Surfers and her profile is a lot more interesting than mine. She has a much nicer apartment, speaks more languages, been to more places has more friends and references, etc… I’m not sure how to remedy this exactly but I think that there should be a guide or introduction sequence to help initial users establish a base of trust in the system, similar to that in Hobbo Hotel. As a first time user I felt a little intimidated by other members and their established contacts. Initial frustration for first time users may cause them to leave the community.

My friend Shawn has so many good reviews. Why would you hang out with me when you can hang out with a guy who will serenade you to sleep?

I could suggest many features to improve the experience of Hobbo Hotel, but that would require making it more like Second Life or WoW. I think it would be nice to have user generated furniture like in Second Life. This would allow a user to express even more individuality when setting up his or her room. Another thing I would like is having areas in the game to chat about specific topics of interest. I felt awkward starting conversations with someone just because my avatar happened to be in the same room. The only person I talked to was the guide who I summoned when I first started in the hotel.

From personal experience, I feel that virtual environments should offer a variety of quests, minigames and themed events in order to improve social capital. Hobbo Hotel has a few basic missions but I think if there were more in depth missions that required collaboration, people would build up a greater system of trust. Working together to solve a problem builds camaraderie and leads to stronger ties between people. I like the fact that pixels are given out for simple quests like asking for a guide and I think more features like this would encourage more interactions.

I’ve seen some themed events in virtual environments like MapleStory that I thought were very encouraging for social capital. I read somewhere that Hobbo Hotel has events but I couldn’t find them. I think themed events (for example a valentine’s day ball) would take advantage of the visual properties of Habbo Hotel. One of the advantages of a virtual environment like Habbo Hotel is that it has interactive graphical features like a decorative environment. I believe virtual worlds constantly need to add new environments and features, otherwise people get bored with the world and leave. WoW takes advantage of this by always expanding the world and adding new quests.

Another thing that Habbo Hotel lacks is a grouping element which helps to build bonding relationships. This may go along with collaborative quests since people would form small groups. I think the concept of “guilds” strengthens bonds because people become more exclusive and find a common identity within a smaller group than the greater population.

A lounge in Habbo Hotel.

Overall it was difficult to compare the two communities because they had different objectives. CouchSurfing has a defined purpose (hosting people, getting hosted), while Habbo Hotel is more recreational. In Couch Surfing users are encouraged to have offline interactions and relationships with other users to promote and establish friendships. In Hobbo Hotel users interact pretty much within the constraints of the world do so for many different reasons.

The Ellison article on college students and facebook mentions that previous research suggested that Facebook users engaged in searching for people they had an offline connection with more often that complete strangers. Somehow I think Couch Surfing is different because the “objective” of the site is to find people to host you, and typically there isn’t much contact between you and those people. Users might chat a bit and set up a meeting, but overall there’s a good possibility that these people are meeting face to face for the first time because of Couch Surfing. Also, Couch Surfing appeals to a specific niche group. It seems like most people on there are extroverts who have a lot of experience travelling. You also need to be somewhat financially stable to afford travel and have a stable housing situation in order to host.

*************************************************
One question that I’ve been thinking about deals with evaluating a user’s social value based on incentives and other visible markers. For instance a user’s level in answerbag affects the way someone might perceive him or her. Users in answerbag celebrated Firebrand reaching level 100 due to her contributions to the community. However, you don’t see parties in facebook for people who have achieved a certain number of walls posts or a certain number of friends. How do the measures of status in a community affect social capital within the community?

In Couch Surfing social worth/status is in the form of friendships, the degrees of those friendships, positive comments, number and quality of host/hosted experiences and response rate. It seems like an individual user’s friendships are an important criteria in gaining opportunities with that site. Social worth in Hobbo Hotel is in the form of badges, pixels, coins, the quality of your room, and the number of friends on your friend list. However, I don’t think more highly of a user who has a lot of friends in Hobbo Hotel whereas I put more trust in a user on Couch Surfing who has many friends and positive relationships. How does the perception of social value in these social communities affect interactions between people in those communities? Is social worth of users in one community more important to interactions than the social worth of users in another community? What makes it so?

To explore this issue I wanted to look at two sites designed so that users who interact online also have opportunities and are encouraged to meet face to face. I’ve been talking with BJ about these issues and we might study Couch Surfing and OKCupid. For Couch Surfing I plan to examine the aspects of a person's profile that encourage trust. Since positive reviews seem like they establish trust I want to examine the number of positive and negative reviews for a number of random Couch Surfers and see how that correlates with the amount of close friends and the total number of interactions. I believe BJ will do a similar analysis with OkCupid (but don't quote me on that as we're still discussing the project.)

The Answerbag Journey

Before I begin I want to share some information about myself which will help my classmates understand my mentality about this assignment and how it changed as I spent more time trying to achieve 2 of the 4 requirements of answerbag. I have never liked social networking websites and I don’t care to participate in any of them or even lurk on them. The Schrack article on social media usage mentioned that extroverts may be more likely to participates that involve interacting with others like social networking sites. Well, I’m not an extrovert and I especially don’t like sites in which you can add “friends.”

The first thing I noticed about answerbag was the usage of performance appraisal reinforcers in the form of points. Questions and Answers were rated on a point scale and all users (I think) could add or subtract points. There was also a leveling system in which users who accumulated more points could wield more power. I soon found out that meant that users on a higher level could add or subtract more points based on their level.

My first strategy was to simply think of one good question that would get me 40+ points and 8+ responses. After browsing through the questions, I settled on dating as a category that was popular and accessible. I noticed that many highly rated questions were easy to answer such as “Can you describe yourself in one word>,” or “What do you look like?” I’ll admit that my original strategy didn’t have much merit. I actually refined my strategy over time when I realized that it wasn’t working.

I noticed questions that were too specific in genre didn’t get as many responses as those that were answerable by many people. I saw dating was popular so I tried asking: “What’s the worst dating advice you’ve ever received?” I got one quick response in a few minutes, but nothing after that. Then I realized that someone else already asked that question.

I also tried to answer some questions. I spent half an hour crafting a long answer to a question: “What’s your favorite female superhero?” It was about a paragraph long and I gave a pretty sarcastic answer. I thought that maybe it would get more points because it was more descriptive than most other peoples. I soon realized that this was not true.

After experiencing relatively little success with my first few questions, I reformatted my strategy based on my initial failures and some theories I got from the reading. My first goal was to just get a question that got 8 or more responses. I used some of the theories Tedjamulia and co. article on motivating content contributions to formulate a new plan of attack.

Proposition 3 of the article states "Members with trust in the community will contribute and participate more frequently." Although this is stated on a greater macro level, my theory was that the same would be true for my particular case. If I establish trust with people, they will be more likely to respond to my questions and contribute points. I decided to write a little bit about myself in my profile and also add a picture. I hoped that other members would feel I was opening myself up to them and thus they might be more likely to contribute to my questions. I also decided that I should also start responding to other people’s questions with well thought out answers so that they could see I was an individual worth responding to.

Proposition 11 in that article said that rewards and reinforcement that are informative and not controlling will increase a person’s desire to contribute. I applied this theory by responding to people’s answers on my questions and giving them points. I noticed this phenomena applied first hand by NUNYA, one of the most popular contributors. I answered several of her questions and she always responded and gave me +6 points. It’s true that it made me want to answer more of her questions because I felt like my answers meant something to her.

In addition I wanted to build up relationships with others in order to get them to contribute. Initially your question will be given priority since it is fresh, but after a while it will get buried under a bunch of new questions unless someone else bothers to look it up and respond. Usually only a friend would bother to do that.

The Java and Co. article on “Why We Twitter:…” argued that people in friendship communities often know each other. Since we didn't have much time to establish friendships on the site, I realized that it would be highly beneficial to get some support from my classmates. I talked to BJ in real life (he works in the cubicle next to me) and added him as a friend (Richard Sorge). I also added Junie (babalu) and George Lee(yops2k). I tried to add some people who I thought were in the class as well. I saw Oracle13 was a computer programmer that was asking a lot of questions and he joined Feb 9th and so I added him. I don’t know if he’s actually in the class. I also added Rinmarie who joined around the same time and was asking a lot of questions. I’m pretty sure I made a mistake with Rinmarie because even after she got 8+ responses and 40+ points on a single question, she continued to post questions.

***Edit****
Oracle13 is TomJenni. Thanks to her too for commenting on my questions.
**********

I typed in some of the names of the people in the class and found thechickenbus. I answered some of her questions and gave her points in an attempt to establish friendly ties. I think it was a very good move for her to identify herself to everyone else in the class by using the same name as in her blog. Junie and I both commented on her "Americans are ____" question which revived her question thread and allowed her to achieve the goal rather quickly. Her question on” Americans are ______” was an excellent example of taking advantage of trend analysis as mentioned in the Java article. The article mentioned that the word “school” is more frequent in the early week, while “friends” was more present during the weekend. She posted her question late at night (around 10:00 pm in Hawaii) which would mean that most Americans would be asleep but users in other countries might be on. I assume that people from other countries will be more opinionated on Americans. After observing that I realized that I needed to take into account my audience and even things like when I post and how I construct my question.

The question that I asked that got 22 responses was:

“I’m thinking of writing a poem to my girlfriend for Valentine’s Day. Ladies, would you prefer that or an expensive gift like jewelry?”

I wrote the first part of my question to give some background information about myself and establish a sense of trust. It followed from Hypothesis 1 in the paper on “Using Social Psychology to Motivate Contributions to Online Communities” by Ling and co. Hypothesis 1 in that paper is: People will contribute more to online communities when given personalized information showing that their contributions would be unique.” By including that information about my Valentine’s Day plans I hoped that people would be more willing to answer the question. I also addressed a specific audience (Women) with content matter that was relevant to the time (Valentine's Day). Based on my observations I noticed that there were a lot of women on answerbag (at least I didn’t notice if males outnumbered females by a wide margin). This went along with the Schrack paper on gender which said that MySpace challenged the digital divide that the internet was primarily middle class, Caucasian males. I hoped that by addressing women directly, I would make women feel that I valued their specific contribution.

While I did achieve my goal of 8+ responses with that question, I was disappointed to see that even with 21 responses (the 22nd was myself), I only got 33 points. I didn’t think of it back then, but maybe it was because I didn’t comment on anyone’s responses. I also realized that in order to get points it would be much easier to achieve 40 points if I got positive ratings from higher level people. It’s much easier to try and get 7 people who can give 6 points each to an answer rather than get 40 people who can give only 1 point each. I probably should have tried to make friends with high level people, but instead I tried to friend low level people. This realization didn’t cheer me up because I had already invested several hours trying to get this to work.

The question that got me 40+ points and finished the assignment was one that I didn’t expect to get me many points. I asked a genuine question about my volunteer work at Easter Seals.

“I help people with mental/physical disabilities do activities around town and I noticed that some people are disturbed when they see us. Is it unsettling for you to see someone with down syndrome? (Honesty appreciated)”

I was impressed by the responses that I got. One person mentioned that his brother as disabled and grew to appreciate people with special needs. Two others mentioned that they worked with people with disabilities for over a decade. I was touched that some people wrote paragraphs in response. The Ridings article on Social Communities mentioned that online communities provide emotional support and I experienced emotional support first hand through this question. Social support was the second most popular reason, as cited in the article, for health/wellness. The question became a starting point for people to talk about their experiences with handicapped people. At that point I felt like I became the third type of participant in the Tedjamulia and co. article; One that responded to others, engaged in a social interaction and made intelligent, distinct contributions.

I should also note that my down syndrom thread died with only a few responses and nine points, but then I Junie (babalu) commented on it to help me out. Beyond the wording and content of the question, I also needed the support of others to keep it from being buried under the sea of new questions.

I think that one of the keys of this question was the ending part (honesty appreciated). I did get some people who did admit that they were uncomfortable dealing with people with down syndrome. I’m not sure if these people would be willing to express a more controversial view without this statement. I also think that when I responded positively to many of the comments, it also offered positive reinforcement for others.

To the classmates BJ, Junie, George and whoever else who helped me pass this assignment I owe you one. I tried to help out others in the class that I could recognize but it was hard to find them. Even though I started on Monday I didn't achieve the two conditions until Friday. I learned a lot from this experience and I have a new appreciation for online social communities.

A link to my question. Click on my username PillowEatsBear to see my profile