session 3: nature vs. nurture and weeds

Readings: nature vs. nurture?

            Are Internet users' actions and frequency of contribution  predetermined by fairly stable personality traits, or can systems/community managers/peers motivate users to take action?

            Ling et al. (2005) found that alerting users about unique information they could add to the system motivated users to contribute more to a movie rating site. One mind-boggling point was that a small percentage of users contributes a disproportionate amount of content. I wondered, who are these "super-users," and what do they have in common? What traits, skills or knowledge do they have that influence this behavior? By understanding this, perhaps we can find ways to develop these traits in other users to ensure stronger virtual communities. Tedjamulia, Olsen, Dean and Albrecht (2005) explored the traits of consistent contributors, looking at factors such as self-efficacy, high need to achieve, intrinsic motivation, and trust, as well as environmental factors and goal setting/commitment.

            This statement also intrigued me: "Surprisingly, subjects in the uniqueness condition also perceived themselves as having views more similar to others in the group than did those in the non-unique group." (Ling et al., 2005).  I wondered whether  subjects who felt valued for their uniqueness then perceived the group as similar to strengthen their already positive feelings.

            Ling et al. also pointed out that users need feedback on how their contributions improve the community. For instance, the system might generate a congratulatory e-mail if a user rates a "rarely rated" movie, or even quantify a user's impact, by saying that the system's  recommendation ability has improved x% due to the user's actions. Alternatively, reviews could have a "like" button, or, like yelp, have a useful/funny/cool button so users can see how others respond to their contributions. This connects well with Tedjamulia, Olsen, Dean and Albrecht's assertion that "Extrinsic awards can be either controlling or informative." (2005, p. 6) This distinction jumped out at me based on my work experience. Having worked as a classroom teacher for 2 years, the issue of how to motivate students without taking away from the intrinsic satisfaction of completing a task was something I thought about daily. Rewards that remind contributors of their importance to the community and acknowledge their efforts can boost intrinsic motivation.

            Another fascinating finding was that users who were reminded that rating movies helps them or helps others reduced their ratings! I interpreted this as saying people have  a holistic, pro-social view of their virtual communities - they think about the benefit to themselves and to others, and appealing to one or the other isn't effective because it doesn't jibe with users' views.

            One question I asked myself after reading Ling et al. was: how do communities support user contribution over time? While these e-mails led to bumps in ratings the week after the e-mail was sent out, what fosters long-term participation?

            In contrast, Ridings and Gefen (2004) did not attempt to manipulate or influence user contributions, but understand why people join virtual communities. They found that friendship and information sharing were predominant reasons for membership, but that people also sought social support and recreation online. This is interesting to look at in light of Ling et al. (2005), as they focused primarily on motivating users to share more information (rate movies) but did not consider in depth the effects of friendship, social support and recreation and how these factors might influence information sharing. I also made a connection to Cheshire and Antin (2010), who argue that some of the current research presents the user as a "tabula rasa - lacking distinctive attitudes and predispositions but inclined to respond to various incentives and motivational feedback." It was no doubt beyond the scope of the Ling et al.'s work, but it would have been interesting to have the e-mail recipients provide some feedback on how they perceived the reminder e-mails.

            Ridings and Gefen pointed out that not every group on the Internet is a community. They used phrases such as  "lasting relationships," "persistently interacting members," and "ongoing exchanges" to highlight how a virtual community comprises a group of people who interact online over time. Despite that, the authors reminded us that "The Internet context is unique in that it provides a way for an individual to observe and learn from a group without the group even knowing that the individual is present."

            They also noted that user motivations for joining depended on whether groups catered to a "freely chosen" interest (like a hobby) or an aspect of their life that may not be voluntary (such as work or health issue).

Experiences: confirmations and extensions 

             My experience is in line with Ling et al.'s point about uniqueness being a motivating factor. For instance, when reading blog posts, I am more motivated to post a comment if I notice, after scanning others' comments, that the aspect of the post I wanted to address has not been discussed. However, if the post has dozens of comments and multiple commenters have made the same point, I usually don't leave a comment.
            One issue I was interested in was not addressed in the readings. While Ridings and Gefen looked at what motivated people to start participating in a virtual community, and Ling et al. and Tedjamulia, Dean, Olsen and Albrecht focused on participation, I wondered about linkages between the two. Because participants are seeking different things, they no doubt use virtual community features in different ways. It would have been interesting to see what features (messaging, search functions, ability to rate messages for usefulness, ability to add friends) users with different motivations used the most and which they ranked as most important. Every community draws people with different motivations, and understanding which features support the largest base of user needs may be key to maintaining a vigorous community.

Online community: Amazon.com's Young Adult Forum           

            The online community I investigated is the one of the forums (called "Customer Discussions") at Amazon.com.  One thing that got me interested me in exploring this community is that it doesn't really seem to be promoted by Amazon. For instance, the forums are not visible from the homepage. Even clicking on "shop by department" and choosing books doesn't lead you easily to the forums. You need to scroll all the way to the bottom before you get a list of the latest posts -- but only for the book forum. I clicked on the "What Are Customer Discussions?" link at the top right and found out that there appears to be no central gateway to see all forums. Instead, discussions must be accessed from product pages.

            I decided to type in a few of my favorite authors and see what kinds of forums popped up, since there was no central page to access a master list. By typing in YA author Markus Zusak, I found that there was a forum dedicated to his novel "The Book Thief," as well as a Textbook Buyback Forum and Young Adult Forum.

            Next  I looked up Haruki Murakami. His author page included his author forum, but clicking on an individual book, such as his latest novel "1Q84," or "Hard-Boiled Wonderland and the End of the World,"  showed no discussions had been created for the  book. This was surprising, particularly as 1Q84 has 537 customer reviews. Looking at the author forum showed that 1 of the 3 discussions had to do with 1Q84 and generated 19 responses.

            I investigated the literary fiction forum, which was suggested as a related forum from the 1Q84 page. I ran a search for "Haruki Murakami" and retrieved 26 results in that forum. Searching with the "search only this forum" limiter clicked off gave me 573 results from all Customer Discussions.

            For this post, I decided to examine the Young Adult Forum.

Modes of participation

•        "add your own message to the discussion" - includes "insert a product link" option (surprise, surprise!) and option to receive an e-mail when a new post is added

•  reply
•   permalink 
•    report abuse
•     x of x people think this post adds to the discussion. Do you? yes/no. 
•     track this discussion - shows how many others are tracking
•      RSS feed
•      search customer discussions with option to search discussion currently viewing

            Clicking on a forum participant's name will lead you to his or her profile page.  Users can display a photo, location, and link to their web site, but there is no built-in messaging option. Profiles can include badges such as "real name" or "top reviewer." Profiles also display the percentage of reviews deemed useful, reviews, interests, and frequently used tags.

Encouraging participation

            One way that participation is encouraged is through the "adds to the discussion" rating. This can be seen as encouraging users by providing some feedback on how others perceive their message, which can motivate posters to do more if their comments are well-received.  As Cheshire (2007, as cited in Cheshire and Antin, 2010) notes, "online interactions do not always have a defined audience, making it difficult to assess how many individuals might benefit from one's contribution." This feature could also be seen as conveying and upholding community expectations by transmitting approval and disapproval. This feature could be effective in theory, but in scanning the posts in the Young Adult Forum it seems that many readers do not take the time to rate others' participation, which limits feedback to users. Perhaps groups could set up a code of conduct that encourages others to rate responses on topics they view or comment on, or the system itself could provide a read/rate ratio and encourage users to rate a certain amount of  what they read. A simple message such as, "You clicked on x posts in the Young Adult Forum in the last week, and you rated x responses," could be helpful. Drawing on Ling et al.'s (2005) finding that talking about both self-benefit and group benefit led to a rise in movie ratings, this message could include a reminder about how rating responses is a quick and easy way to participate and stay active in the forums, while it helps build community and drive meaningful discussion.

            A second way that participation is encouraged is through badges, which Amazon says are used to identify "our best content contributors." (Badges are not visible within Customer Discussions, but can be viewed by clicking on a participant's profile). There are 10,000 people listed as Top Reviewers. This is effective in ensuring quality reviews, in that one component of the Top Reviewer ranking draws on how many people deem a review helpful. However, given that millions of people access Amazon daily, this badge system seems to encourage too small of a group. Thinking about books, it would be useful to have specialized badges such as "cookbook junkie" to highlight participants who focus on certain genres. It could also be useful to allow users to  rate reviews not just on helpfulness, but factors such as organization, accuracy (if the reader has used the same product) and detail. Or -- and I don't even know if the software/program exists for this -- readers could highlight parts of a review that they particularly liked and the program would display these or even analyze them and assign descriptors. What I'm envisioning would be something like a tag cloud, but more sophisticated. Lastly, badges are currently focused on user activity on product pages, but it would be helpful to create badges for forum participation.

            A third feature that encourages participation is users who respond positively to others' comments and extend the conversation (see example below).  I think this is one of the most powerful ways to build virtual community participation. One way that positive connections between forum users could be fostered is through the addition of basic SNS features such as private messaging or adding people as "friends."  

          
           A fourth feature -- if we consider lurkers as participants, as  Tedjamulia, Olsen, Dean and Albrecht (2005) and Ridings and Gefen (2004) do -- is the search function. I think this is fairly effective, as it allows users to limit their search to one forum or search all customer discussions. One improvement I would like to see is an incorporation of these search results with individual book and author pages. For instance, searching all customer discussions for Haruki Murakami brought up 573 results. It would be great to display the 10 most recent or 10 most popular customer discussions from all forums on his author page or on the page for each of his books.

Exploring the Young Adult Forum

            The Young Adult Forum covers 41 pages. I pulled 50 posts from throughout its history, and sought a diversity of posts in terms of content and amount of response. These are the five categories that emerged as most common:

• Request for recommendations (example: "YA futuristic short stories, kind of like Kara Dalkey's 'Hives'"):  18
• Seeking opinion ("should sex be encouraged in YA novels?"):  7
• Stumpers, or posts by people seeking to identify a book when they don't know the title or author ("Need HelpLooking for a YA werewolf book"):  7
•  Self-promotion of books, blogs, giveaways, etc ("Free book download - 'T'Aragam'"):  5
•   Two cents, or general questions that will appeal to a wide range of readers, as they are not specific to genre or authors ("What are you currently reading? (no self promos)"):  3

     

Category	Number (out of 50)	Mean participants	Mean total posts
Request for recommendations	18	26.94	42.88
Seeking opinions	7	14	26.7
Stumpers	7	2.28	3
Self-promotion	5	1.2	4.2
Two cents	3	97.3	1,276

            Three of the five post categories -- request for recommendations, seeking opinions, and stumpers -- indicate that the forum is well-used by participants with information needs. These posts altogether account for 32 of the 50 posts, or 64%.  
           Requests for recommendations were, as expected, the most common posts in the forum. They vary greatly in content, with some people specifying desired literary elements and age of reader
 ("Can you help me find books outside the fantasy genre that have wit or rich characters and stories, without weighty depression/relationship dysfunction themes, or vacuous beach girl mentality? I have a smart, poetic, 14 1/2 year-old who wants to read something interesting and well written. Thanks for your suggestions.") or mentioning a few titles that were liked or disliked ("Looking to buy my daughter some books for Christmas, she is 15 loved hunger games and uglies, hated twilight. I've ordered some of my favorites but not really up on current books, Any ideas?"). Some were very vague, such as "I am looking for some good recommendations for YA fiction for a 13 year old who is an advanced reader. I am looking for clean fic so no sex or swearing please... "

           In my initial classification, 3 of the 7 seeking opinion posts were classified "opinion/values," as they addressed the issue of objectionable content (namely, sex) in young adult books.  Several of these posts were written by parents seeking guidance in determining whether content was appropriate for their child. Other, less volatile, topics in this category included Kindle use and what readers liked and didn't like to see in sequels.

        I was not surprised to see that stumpers, while equally as common as opinion posts, generated low mean participants and mean responses.  These questions fare better on librarian listservs, where the average participant is a skilled searcher and usually has been exposed to many more books than the average reader.    
         Another unsurprising low performer in terms of participation and response were the self promotion posts. However,  I do wonder how effective these promotion posts are. It's impossible to know how many forum users downloaded a free e-book or entered a giveaway after viewing the post. Amazon forum information does not include views of posts, so there's no way to track how many people click on these types of posts.
          While 64% of the posts qualified as information-sharing, it also appears that friendship is a powerful motivator in the Young Adult Forum. While two cents posts accounted for the least amount -- only 3, or 6%  -- of the 50 posts I examined, the number of mean participants and mean responses is amazing when compared to the other 4 descriptive categories. Avid readers want to get to know other avid readers, and posts like these generate a lot of interest and response. It seems that two cents posts may be a good way to draw new members into the forums and build relationships among existing users.    

           

Discussion and evaluation

            Amazon's Young Adult Forum reminds me of a patch of weeds flourishing in the crack of a sidewalk. I'm not saying that the forum is a pest, but choose that term to highlight the resilience and persistence of folks who want to use the forums to connect with others despite the sometimes challenging environment. These challenges include lack of basic social networking site features such as private messaging and friending features and lack of a central access point. As Ridings and Gefen (2004) said, "patrons of virtual communities are motivated primarily not only for information but also because of the friendship and social support these virtual communities can entail."

            Ridings and Gefen offered a few suggestions on how virtual community managers could tailor their communities to better suit users' needs. For instance, communities with an emphasis on information sharing can ensure that bulletin boards are clearly labeled and named and offer strong search features. Because Amazon lacks a central access point for all forums, establishing this in itself would make the forums easier to locate and use. This could also facilitate the creation of subforums; for instance, from scanning the Young Adult Forum, it seems there may be enough interest to support subforums based on YA romance and dystopian fiction. Searching could be improved by adding a breakdown of results by forum of the "all Customer Discussions" search so users could easily determine which forums have the most discussion on their topic.
            Tedjamulia, Dean, Olsen and Albrecht (2005) discussed usability (an environmental factor) as having an influence on the personality factors that encourage contribution. Another environmental factor that Amazon could positively influence is personal responsibility. The "real name" badge is a good first step, but more could be done. Right now, the forums only show the commenter's hyperlinked name. A more information-rich ID that includes how many comments the participant has made, when they joined, and any forum badges earned could encourage more participation. This would be in line with Ridings and Gefen's suggestion that communities focused on friendship and social support make it easy for users to see who's responding to a post and member profiles.
          From what I've seen, it's clear that the forums are somewhat of an afterthought for Amazon. They appear to exist only because they may help someone decide to purchase a product. While I can understand that adding SNS features such as friending abilities may not be desirable for a retail site, I do think simple things like a central forums page and a way to see the forums (maybe latest posts or most active forums in the past hour) from the homepage would support the many users who are already using Amazon to pursue personal interests and build online communities.