8: Thinspiration

Following Braithwaite’s study on the types of messages exchanged in online support groups, we decided to code messages found in a “pro-anorexia” Google group. Although one would typically think of a support group as a place where individuals turn to others as a means to work through an illness or disability, this message board was in fact the exact opposite. Instead of providing encouragement or advice on how to defeat anorexia, members of this social group supported others on how to remain thin by any means possible:

“I kinda feel like the whole "pro-ana" scene has really gotten away from whats important. Losing weight at all costs! I feel like people has gone from 'do whatever it takes' to 'its ok to be a little chunky'. It's not ok! It's SO not ok, and I don't want to let myself slip into that mentality."

Braithwaite used Cutrona and Suhr’s (1992) category system of supratypes and subtypes to code messages as containing information support, tangible assistance, network support, esteem support, and emotional support. Since Braithwaite also frequently found humor in messages, this was an additional category we took into account.

Messages that included emotional support involve relationships, physical affection, confidentiality, sympathy, understanding, encouragement, and prayer (Braithwaite, 1999). In Braithwaite’s study, she found that emotional support messages appeared most frequently in support groups, making up 40% of the messages that were coded. In our study, we found that 85% of the messages included emotional support--twice the percentage of Braithwaite's findings. This may be a result of the fewer number of messages we coded for our study. Moreover, since the nature of this forum leads individuals to push others to stick with an anorexic “lifestyle,” encouragement is inherent in the group we analyzed.

Information support includes advice, referrals to experts, situation appraisal, and teaching messages (Braithwaite, 1999). Information support was found in 31.7% of the messages in Braithwaite’s sample. Although our percentage was much higher, at 95% of the messages coded, this could be due to our inclination to selectively choose threads that seemed extremely content based. We were quick to disregard those threads that had few words and little support or advice. Had our research methods been more systematic, our percentage may have more accurately reflected Braithwaite’s results. This too could be due to the nature of the forum, since group members were eager to pass on their own personal tips and suggestions. A great example includes the thread title “Thinspiration,” an idea by one member to each give a different tip on remaining thin--the most extreme post involving “The Ana Commandments."

Messages containing esteem support includes compliments, validation, and relief from blame (Braithwaite, 1999). We too found esteem support messages to rank third in terms of frequency. Validation, in particular, was found very frequently in our discussion group, as members who were posting were supporting an illness that is not traditionally accepted as a positive behavior that should be encouraged. In contrast, Braithwaite was probably more inclined to find esteem support involving relief of blame or complimenting individuals on the way they overcome their disabilities.

Tangible assistance encompasses performing a direct task, active participation, and expressing willingness (Braithwaite, 1999). Network support involves messages that give Information specifically about networks, including access, presence, and companions (Braithwaite, 1999). Both our study and Braithwaite's study found tangible assistance and network support to appear the least frequently. However, in contrast to Braithwaite, we found more tangible assistance than network support, perhaps due to human error in coding. Perhaps the lack of network support is a result of the rarity of the "pro-ana" support network. The members already contributing to the group feel that they and the other contributors are the few that truly understand this way of life and have no need to connect them to others outside of the group. Additionally, tangible assistance in this case was difficult to find most likely because acts of support were more text-based recommendations rather than a means of physical assistance.

After individually coding messages, we discussed our viewpoints before coming to a consensus. Since there were three of us viewing these messages, whenever one had a different viewpoint then another, it led to disagreement and further discussion. Our inter-rater reliability was 90.83% indicating an extremely high level of consistency. Braithwaite similarly found high inter-rater reliability at 80%. Since training and experience both lead to higher inter-rater reliability, our high result may be attributed to previously learning how to accurately classify messages and already doing a practice coding exercise together in class.

Our results can be further analyzed by applying Walther & Boyd’s (2003) four factors: social distance, anonymity, interaction management, and access. Social distance would encourage socializing online due to the fact that the Internet is a medium in which individuals could be separated by large distances. The Internet provides a wide space for individuals to find “experts” to interact with and seek support from without geographic boundaries. Anonymity allows individuals to feel more comfortable providing revealing information or seeking help. Since the “pro-ana” messages have a stigma associated with them, anonymity helps reduce the fear of sharing this information as compared to sharing this information face to face with close friends and family. Interaction management involves selective self-presentation, in that the asynchronous environment allows people to carefully pick and choose how they would like to be perceived. This is particularly important in the support group we analyzed since members may be especially self-conscious and have low self-esteems. Access refers to the idea that no matter what time of day or where you are, support can be found on the Internet through already posted material or in a later response to your own request for help. Individuals who encourage anorexia can search to find others who have similar needs.

Authors: Alyssa Ehrlich (red), Randi Pochtar (purple), and Lauren Burrick (yellow)

Support Group: http://groups.google.com/group/pro-ana/

Threads Used: Thread 1, Thread 2, Thread 3, Thread 4