Monday, April 22, 2013

session 7: the wisdom of not replying



Governing documents
         YouTube has Community Guidelines and Terms of Service. The guidelines can be considered what Grimes, Jaeger and Fleischmann (2008) refer to as community standards, which address "highly contextualized problems" and are "written in an elusive and haphazard manner." YouTube details how it handles policy enforcement here. When YouTube removes content for violating community guidelines, the user receives  a strike. These strikes last for 6 months. Users who believe their content was mistakenly removed can appeal.

3 examples
        To find examples of rule-breaking and conflict, I took a look at the most-viewed videos of the 11 YouTube users who responded to my starter survey. I decided that more views would equal more comments, thus the greater chance for rule-breaking and conflict.
        On ElectricDade's most-viewed video, "FTMtransition: One year on testosterone," I found a derogatory comment from muffO3O. This goes against YouTube's community guidelines, which state: "We encourage free speech and defend everyone's right to express unpopular points of view. But we don't permit hate speech (speech which attacks or demeans a group based on race or ethnic origin, religion, disability, gender, age, veteran status, and sexual orientation/gender identity)."

        This comment was marked as spam, and drew responses from ElectricDade and user Alyssa Bailey. Electric Dade got 46 upvotes for one of his responses, while Alyssa Bailey drew 40 upvotes. The ironic thing about upvoting supportive comments is that it can push the questionable comment to prominence in "Top Comments" section. This effect is especially pronounced in videos with lots of views and comments, such as this one, which as of April 20, 2013, had 7,881 comments and 1.07 million views.

      My second example comes from the same video.  In this case, user PilotRussia takes issue with Tully Roll's positive response to the video. PilotRussia's comment was downvoted so it's initially hidden.

       
        My third example comes from RaideretteGirl24's most-viewed video "Before & After HRT 5 months progress video," which as of April 20, 2013, had 150,357 views and 662 comments. Like the example drawn from ElectricDade, the comment featured what would be considered hate speech:


         The comment was dealt with through downvoting, and also triggered a vociferous defense from Ben Schwartz.
          It seems that the three main ways contentious comments are dealt with include: personal responses from the video creator or other users; downvoting; or marking as spam. Downvoting only works as a collective effort - i.e., enough users must hit the thumbs down button to make it effective. (It is not clear how many negative votes are needed to hide a comment. YouTube is probably deliberately unclear on this. Some speculate it requires at least 5 negative votes).  However, personal responses and marking comments as spam can be implemented by individuals.
          According to this, YouTube defines spam as "content and/or correspondences that cause a negative user experience by making it difficult to find more relevant and substantive material. It can sometimes be used to indiscriminately send unsolicited bulk messages to users." YouTube notes: "Please use the 'Mark as Spam' feature with extreme caution, as those who misuse it may be prohibited from using the site." The article also points out that uploaders can moderate comments before publishing, remove comments, and also unmark items flagged as spam.
          Marking something as spam gives users the immediate gratification of hiding a comment and making it a little more difficult for others to see. Out of curiosity about the use of the spam feature vs. downvoting, I decided to examine the first 115 comments on ElectricDade's video. I found that
muffO3O's comment drew disapproving responses (some expletive-laden) from 6 other users. Of the 115 comments examined, 3 were hidden due to negative votes, and 3 were marked as spam. Of the three downvoted  comments,  two asked about the specifics of how Electric Dade became a father and one asked about the direction of transition. One spam comment mentioned werewolves, another called Electric Dade a "freak," and a third was a compliment.
          I wondered about the balance between use of the mark as spam feature vs. downvoting on videos with fewer views. I took a look at violet4151's most-watched video, which is "Hormone Effects MtF." As of April 18, 2013, this video had 261,215 views and  305 comments. Of the 305 comments, 8 were marked as spam and 0 hidden due to negative votes.  Of the 8, 2 were questions:


          One comment marked as spam was a duplicate of a comment that drew 15 upvotes:


            The remainder of comments included a message encouraging users to create videos for Janet Jackson's upcoming documentary on transgender individuals, 3 positive comments, and 1 encouraging the uploader to "model sexy outfits." The seemingly haphazard use of the "mark as spam" feature on this video was puzzling, as it did not fit the pattern seen in the other videos.

Administrator response
           While Kirman (2012) talked about how mischievous behavior online can be a valuable and essential way of expressing creativity and exploring the bounds of a community, the examples above appear to be malicious, not mischievous. However, I believe Kirman's discussion of how users appropriate tools for uses not envisioned by designers is seen in how the "mark as spam" feature is used to circumvent the collaborative requirements of the downvoting system and exert individual power over disagreeable comments.
           For example #1 from ElectricDade, I believe users need a different way to express disapproval of a user's language/comment content. Telling people to click on a link to win a free iPad and using swear words to insult a person are two very different kinds of misbehavior. Users are misusing the "mark as spam" feature to control what they see in the comment section, because at this time they do not have an immediate way to hide comments that they find disagreeable or offensive.
             As administrator, I would add a "mute" feature for individual users. User A could mute muffO3O's comment while viewing ElectricDade's video, which would hide the comment from her view. However, User B would still be able to see muffO3O's comment. This would give User A a sense of control over her personal viewing experience and satisfaction similar to that of using the "mark as spam" feature, while at the same time ensuring that muffO3O is not censored through misuse of the spam feature. This may also encourage users to downvote items, as the addition of the mute feature will remind them that it only controls their personal viewing experience and if they want to  make their disapproval visible to others, they need to hit the dislike button.
            For example #2, downvoting was used to hide the comment. However, this may be a comment that merits removal by the uploader because it adds little to the conversation. If the uploader chooses to leave it as is, a better option may be to implement a collaborative filtering system for comment display.
          Right now, YouTube comments are shown chronologically, with the most recent comments first (after the uploader comment and top comment areas). This sounds simple, but really isn't, as anyone trying to unravel the chronology of a heated exchange via comments will find out.  The confusion is only magnified by the lack of a detailed timestamp. When 15 comments are marked "1 week ago," it's hard to tell which comment came first, particularly if users didn't bother to hit the reply button but are directing their comment at a particular user.
        A collaborative filtering system would push comments that are hidden due to downvoting to the bottom of the comments list. The remaining comments would be presented chronologically, with the most recent first. Users would also retain the option of viewing all comments in the usual chronological order. Cosley et al. (2005) note that "reducing the link between posting a message and getting responses should reduce spammers' and trolls' motivation to make low-quality contributions." Most users will not read all the comments a video receives, so pushing low-quality comments to the bottom will result in less attention. Dibbell (2008) points out that griefers and trolls are motivated to disrupt and anger other users and ruin their online experience. These individuals will still have their say, but individual users gain a little more control over their experience by having the option of pushing hidden comments to the bottom.
            For example #3, I believe the downvoting was used properly and was effective. However, Ben Schwartz's extensive defense may have ended up drawing more attention to the negative comment. I admit to being curious to read the hidden comment that spurred someone to write a response that spans several comments! In this case, I would remind users to reply with care - i.e., consider whether your defense of the uploader will draw more attention to a negative comment. It may be more appropriate to downvote and post your thoughts without replying directly to the troll.
             
Five unwritten rules
1. Agree to disagree - feel free to express a differing or unpopular opinion, but strive to use language  that will extend the conversation. One-word derogatory comments don't extend the conversation, but expressing your opinion fully and in language that will not detract from your message will allow others to better understand your viewpoint and engage in conversation.
2.  Contribute to YouTube's quality of conversation and videos by using the like and dislike buttons on comments. You can improve other users' experience by liking comments that you think are thoughtful, funny, or useful.
3. As an uploader, be proactive in dealing with comments. A well-moderated comment area will encourage others to subscribe, comment, and build relationships with you and other users. We encourage uploaders to monitor what's being marked as spam (and unmark items that aren't spam), to answer questions and respond to positive comments. We know that everyone deals with negative feedback and comments differently, and that some negative comments may not deserve a response, but look for ways to use negative or misinformed comments as a way to assert your beliefs and educate others.
4. Use the "reply" feature thoughtfully. Consider that writing an extensive reply to a troll may only draw more attention to the troll's comment. You may find it better to downvote the comment and express your thoughts in a comment not directly addressed to the troll.
5. Take YouTube for what it is  - and understand what it's not. YouTube is a great way to share your knowledge, experiences and thoughts with a broad spectrum of people. Just as you have the right to share, others have the right to disagree or outright dislike what you have to say. At the end of the day, you are not your videos - so strive to not take things too personally. 

Sunday, April 7, 2013

session 6: motivations of transgender YouTube video creators




PART I: PILOT STUDY
RESEARCH QUESTION: What motivates transgender individuals to post YouTube videos about their personal experiences as transgender individuals? How do transgender individuals use these videos to create identity and form community?
SITE: YouTube
METHOD FOR ANALYZING DATA:  Users' responses were coded to uncover themes and similarities.
INITIAL FINDINGS:
            I searched YouTube using the key word "transgender." Users were selected for initial contact based on how many transgender-related videos they had posted (the more the better) and if they appeared to be current users (most recent video dated 2013).  Channels that featured content from a group of transgender individuals were not included in the study, as the focus was on individual content creators.
            31 users received an e-mail that briefly described the study and asked whether they would be interested in answering questions via e-mail. Of the 31 contacted, 18 responded and expressed interest in participating. These users then received the following 3 starter questions:
  1. Why did you start posting videos on YouTube? Have your motivations/reasons changed or expanded over time? If so, how?
      2. Prior to posting your own videos, how familiar were you with other YouTube content on the subject of transitions/transgender life? How did you interact with this content (did you share it by embedding it or linking it, leave comments, etc)?
      3. What was the most challenging video or piece of information you've shared on YouTube? Why was this video/information so challenging for you? And why did you decide to share it?
            11 users responded in time for inclusion in this pilot study. Their responses to question 1 were coded to identify themes. See the table below for a listing of themes, the number of users whose responses included the theme, and examples of the theme from respondents' answers.


                While the theme of self-benefit/personal was most common (9 of 11 respondents' answers included this), prosocial motivations (desire to provide emotional support/inspiration) came in a close second, showing up in 8 answers.  The prevalence of self-benefit/personal motivation bears out Wellman et al.'s assertion of networked individualism, in which each person is the node of his or her own community (2003). The asynchronous nature of the Internet means that each person has when-I-need-it access to support and resources, as they can draw upon the support and knowledge regardless of time, location of proximity.
              Of the 11 respondents, 8 gave answers that included more than one theme. The majority of users -- 8 -- said their motivations for posting had not changed over time. For the three users who noticed a change or expansion in motivation,  one user said he now wants to
"reach people who are unfamiliar with transfolk. I feel that education will lead to the normalization of trans." A second user noted that due to positive feedback about her videos being useful and helpful, she has a new motivation to create videos to help people. The third user said "I want to become more of [a] voice and advocate for transgender people as I am realizing on this transition that we don't really have anyone."
               A question that arose due to this pilot study is: does the desire to provide emotional support/inspiration lead to off-YouTube connections? For instance, do transgender YouTube video creators find that their YouTube viewers also connect with them on their blogs, Facebook, Twitter, or through e-mail?Also, do YouTube connections translate to offline connections or action, such as meetups with regular viewers or the giving of time, support, or effort to local or national transgender support and advocacy groups?

PART II: READINGS
What is online identity?
                Online identities are created when individuals disclose information about themselves. On YouTube, an online identity is created through one's videos, which feature images and audio. For the majority of transgender YouTube users, these images and audio are of themselves. This is an interesting mingling of offline and online identities, as users allow an unknown audience to see what they look like and sound like face to face. Many transgender YouTube content creators also create timelines or monthly update videos to show changes that result from hormone therapy.  YouTube users also create identity through textual means, such as comments and messages.  Users can also build identity through actions that signal affiliation or interest, such as liking or downvoting comments or videos or subscribing to channels.
                A primary aspect of online identity is that it is constructed and deliberate. Thus, online identities can be a form of self-expression and creativity (the pessimists might say they are also a form of deception). People usually need to disclose a certain amount of information (name, e-mail, age, location - whether factual or manufactured) to begin creating an online identity in most SNSs. This disclosure is an essential first step to setting oneself up at the hub of one's own community, as discussed in Wellman et al (2003). Online identities are not static, but dynamic, as actions (comments, posts, likes, shares, messages), the addition of friends,  and membership in groups or communities accumulate and change over time.
                A second primary aspect of online identity is that it circumvents traditional boundaries such as time, place, and proximity. Wellman et al. (2003) use the term "glocalized" to emphasize how the Internet has displaced local places as the focus of individuals' identities, relationships, and interests;  local connections are now just one piece of the puzzle. An online identity can be crafted to deal with both a cross-section of real-life acquaintances but can also be strongly context-based. This may sound contradictory. However, consider that a person constructing a Facebook profile can reasonably expect to friend a mixture of family, friends, and professional colleagues. These friends have very different interactions with and knowledge of the individual, which may mean that the individual will need to choose signals that resonate with the greatest number of audience members, to maintain identity authenticity (Donath, 2007). An online identity can also be context-based, as it is comprised of people who are connected to an individual and interactions.  A person's trustworthiness can be evaluated not only based on her individual actions, but by the actions of her friends. Also, while Facebook is a rather general SNS, a more specialized SNS - like a dating site for a certain ethnic group/religion - allows people to emphasize certain aspects of their lives. In the case of YouTube transgender users, YouTube allows them asynchronous access to other transgender users' experiences and stories and a low-risk way to gain information and form relationships through messaging, liking, or commenting.

INFORMAL USE SCENARIOS (SUNNY & RAINY):
A sunny-day scenario:
  •          user creates video
  •          user uploads video to YouTube and decides to make it public
  •          user gives video a title that will help viewers find it (useful words include transgender,  mtf(male to female)/ftm (female to male), or transition
  •          viewers find video
  •          viewers respond positively to video by liking it, posting positive comments, and sharing it
  • ·         user feels personally supported and/or believes that she or he has contributed to greater awareness of transgender issues and experiences.
A rainy-day scenario:
  •          user creates video 
  •          user uploads video as publicly viewable
  •          user gives video a useful title
  •          viewers find video
  •          some viewers leave positive comments, but others respond negatively,  using the comment space to engage in name-calling and ill-mannered debates about religious views of transgender individuals, the causes of gender identity disorder, and what makes someone male or female
  •          user feels personally attacked and isolated. User may take down video temporarily or permanently, post videos less often, stop discussing certain issues, or even stop creating videos.

HOW MIGHT YOUR RESEARCH HELP A RAINY DAY BECOME SUNNY?
                Based on my starter questions, many respondents post videos not only for personal reasons, but due to prosocial motivations, as well. This means that transgender individuals may be open to making changes to how they use YouTube if  those changes will help increase awareness of transgender stories and issues.
                I believe that my research could help by analyzing comments to identify common points of confusion. For instance, just from casually viewing transgender videos on YouTube, it appears that many commenters are confused about the physical implications of transitioning (they don't understand exactly what "top" or "bottom" surgery entails and the end results). There are also disagreements about religious viewpoints on transgender individuals as well as confusion about how and why people are transgender.
                While video creators may want to thoughtfully respond to every confused commenter, some videos have so many comments that this is not feasible. It may be useful to have some sort of transgender FAQ/webliography that draws on resources in a variety of formats, from blog posts, articles by psychologists/doctors, and videos. Transgender individuals would be encouraged to assess the resources' quality, comment on the usefulness of the resource, and tag it for easier access. The webliography could also track how many times a resource has been linked to and allow viewers to rate the resource. A video creator faced with a common question or confusion could respond by leaving a link to an appropriate source from the webliography vs. crafting a reply or definition from scratch or spending time hunting for the perfect resource.