Friday, November 2, 2007
9- Facebook: The World at Our Fingertips
Tuesday, October 30, 2007
Assignment 8.1 - narcolepsy
BIANCA GHISELLI (blue blog)
For this week’s post, we visited Google’s support group for narcolepsy. Narcolepsy is a condition mostly characterized by excessive daytime sleepiness. The site listed hundreds of posts from people discussing their experiences with the diseases. Members also asked questions regarding their symptoms and possible ways to deal with it. My partner and I focused on 20 posts and coded them according to Braithwaite’s five scheme: information (advice, referral…), tangible assistance (perform a direct task, active participation…), esteem support (validation, compliment…), network support (access, presence…), emotional support (understanding, encouragement…), and humor (was there an intent to amuse?).
Overall, we agreed on most of the codes. All the posts we encountered shared the common theme of advice and how to deal with different aspects of narcolepsy. As such, the replies focused on very specific discussion themes and let very little room for code interpretation. This is why our results were so similar and thus our inter-rater reliability was so high (95%).
Although most of our answers were parallel, there were also a few instances where we interpreted the post differently. For example, in message four, the creator of the thread stated that under a certain medication he had become increasingly violent. One of the respondents reacted by saying that under the same prescription he had had become so aggressive that he was surprised he was “…not in prison.” My partner rated this to be 1 under esteem support as the respondent was validating the other’s feelings (anger). I, on the other hand, believed that he was simply stating his emotions and was not either agreeing or disagreeing. After a brief discussion, we decided that the commenter was indeed providing esteem support by confirming that he felt the same way, and thus validating the other’s person’s feelings (violence).
Our results were very indicative of the type of group we chose. Compared to Braithwaite’s data they indicated that people were more willing to provide information (100% vs. Braithwaite’s 31.3%), tangible support and esteem support (10% vs. 2.7% and 35% vs. 18.6% respectively). Also, we showed no network support as opposed to 7.1%. Finally, our study reflected more emotional support (45% vs. 40%).
We attributed these differences to the nature of the subjects discussed. Braithwaite’s study examined disabilities (self-care, mobility, communication, socialization and employment) which are widely known and accepted. Our study, on the other hand, focused on a far less “visible” condition. While in the disability case, people are mostly looking for emotional support, in our case, members felt like online groups were the only place where they could find information and comfort, as everyone else judges them to be simply lazy.
All (100%) of the posts contained some form of information sharing (e.g. “IIRC has been withdrawn in the US and is no longer manufactured.”) Also, very few messages offered tangible assistance (10%). Our guess is that while information sharing is mostly based on giving advice, tangible assistance is a more invested form of help. This is something that far fewer people are willing to give. Next, we found esteem support in a few messages (35%). In our case, this was found mostly in cases where respondents had shared the same experiences and validated some of the poster’s points. Network support was never found in any of our posts. Emotional support, on the other hand, was more frequent (45%). Almost everyone provided advice followed by a few words of comfort to show understanding and sympathy (e.g. “you are far from alone!”). Finally, humor was found in about 30% of the posts. In this case, my partner and I agreed that although jokes might help “lighten up the mood”, the authors might not always know if they are well accepted by the other party.
Our study agreed with Walther’s four dimensions of attraction to online social support groups. First,social distance. Narcolepsy is a rare disease and thus members appreciate the great expertise available in the larger online social network. This relates to our finding as 100% of posts provided information, thus indicating the great amount of advice people seek out and give. Second, anonymity. Whether perceived or real, it increases he ability to avoid embarrassment and judgment (narcoleptic = lazy) and increases confidence in providing support. In this case, members might not be as willing to provide as well as look for emotional support (found in 45% of posts) if they were afraid of others’ opinions. Next, interaction management. Users appreciate the ability to craft messages carefully and to read at their own convenience. In the narcolepsy group, members came from all parts of the world (different time zones) and asked information on medications that were not provided in their country. Finally, access. This 24/7 online support system allows members to read and write messages at any time of the day or night.
1-2
http://groups.google.com/group/alt.support.narcolepsy/browse_thread/thread/1fa4252d40d64974/b255eb68e1603b03#b255eb68e1603b03
3
http://groups.google.com/group/alt.support.narcolepsy/browse_thread/thread/cb82eb3e9adf4074/4e8b2d5d44804e3b#4e8b2d5d44804e3b
4
http://groups.google.com/group/alt.support.narcolepsy/browse_thread/thread/ab1a61f3a0cbfa97/c99167a762306dda#c99167a762306dda
5-11
http://groups.google.com/group/alt.support.narcolepsy/browse_thread/thread/09e0650c9dcd699b
12-13
http://groups.google.com/group/alt.support.narcolepsy/browse_thread/thread/b40c03c88bdce72f
14-15
http://groups.google.com/group/alt.support.narcolepsy/browse_thread/thread/e256d90472ce7d2c
16-17
http://groups.google.com/group/alt.support.narcolepsy/browse_thread/thread/da19b8d7b0058ad0
18-19
http://groups.google.com/group/alt.support.narcolepsy/browse_thread/thread/97c0b49e79ee8296/d308fc5cb61f95e4#d308fc5cb61f95e4
20
http://groups.google.com/group/alt.support.narcolepsy/browse_thread/thread/cb43c99f20b38afc/99ec859a203e6164#99ec859a203e6164
8 "My whole life I have known I should have been born female."
Nina Glatthorn (Blue)
Alex Krupp (Blue)
Our own analysis of computer mediated social support was based on the schema of Braithwaite et al. Target messages were taken from the Google Groups interface to the Usenet Newsgroup soc.support.transgendered. The first twenty top-level responses were each examined independently by the three coders and rated on a binary scale for the inclusion of the different support categories. A table summarizing the results of the analysis is shown below.
The concordance factor – the degree to which the coders' analyses were similar - for the sample was .85. Since a significantly high concordance factor indicates a more reliable data set, it is possible to examine the data further. In the twenty messages, there were fourteen (70%) instances of informational support, four instances of esteem support and four instances of emotional support (20% each), and two instances of network support (10%). Within the set there were no instances of tangible assistance or humor.
In terms of the relative frequency of social support categories, our findings were significantly different than those of Braithwaite et al. (p < 0.05). Braithwaite found that emotional support was most common in the messages that she analyzed, while we found information support to be the most prevalent. Braithwaite notes that “information support is most useful and prominent when the recipient can control the situation and put the information to use.” She speculates that emotional support was more prevalent because of the fact that health problems and disabilities are not under the control of the members of the network. In contrast, members of the cross dressing network were often confused about their situation and therefore actively asked fellow group members for advice and information. Another main difference in our data set was that network support was quite rare. Braithwaite speculates that network support was so important to those with disabilities because of the fact that disabilities often constrict social networks. In our situation, the need for social support seemed to be fulfilled by other members of the Google group.
Another possible explanation as to why our analysis yielded different results is the discrepancy in the sample size (n = 20 vs. n = 1472). In order to be validated, our experiment would need to be reproduced over a much larger sample. If such were to occur, it is reasonable to believe that the relative magnitude of support categories would be similar across the two studies, as there are cross-cultural similarities underlying the notion of social support.
Overall the transgendered community has a solid support group. We were surprised. But looking at Walther’s analysis of online support groups, it makes sense. First, Walther is correct about the greater expertise available. Only a very small percentage of people in any given physical community have experience with transgender, so by aggregating expertise the participants gain access to a much richer forum. The asynchronous and distributed nature of Usenet enables a plethora of options for both access and interaction management, another two of Walther’s factors. The benefits here were tangible; our analyses depended on messages from around the world and around the clock.
Perhaps the most important factor though was the anonymity, both perceived and real. The original poster wrote under a pseudonym, which in this case almost certainly increased self-disclosure. (The stranger on the train effect.) This leads to increased self-disclosure in replies, which creates an environment that is more supportive, both intellectually and emotionally.
8 - Where do you hurt?
Saidu Ezike (Yellow)
For our assignment, we analyzed the content of the alt.support.chronic-pain Google Group. As the title suggests, this group was full of threads pertaining to symptoms, treatments, etc. of medically induced chronic pains. We analyzed a total of seven discussions and 20 threads (of course). Many of the discussions pertained to either complaining about their doctor’s pain treatment or expressing how badly they were feeling pain. One with an Indian doctor even referenced how he felt the Indian system of treatment was different than the American. The others were those asking for specific medical advice for the “physical maladies” they had. The following are our tabulated results:
As a whole, with the exception of a statistically insignificant 10 percent difference in emotional support, our results do not align with Braithwaite’s. We feel this difference has a great deal to do with the choice of support group we chose. The posts in our group were a little more medical specific than those chosen by Braithwaite. Also, the sample Braithwaite chose was much larger than ours and with a larger sample, our results would have aligned a little more. On top of that, we defined “disability” regarding our study as those who had pain/physical/medical disabilities. Braithwaite highlighted the behavioral, economic and social effects of
In addition, since the messages in the support groups were transmitted via e-mail rather than a thread in a group, our theory is that the responders had more time to think of and use diction that would really give the person what they wanted to hear as opposed to simply typing their advice along with their emotions away. Judging by the misspellings, the haphazardness of [some of] the threads and the feeling of a more conversational environment, we feel that threads are a little more rich medium than e-mail and therefore, more ambiguous and emotion detectable (Media Richness theory). Also, since this study was done about eight years ago, we feel that people express their emotions slightly different in CMC. For example, on AIM, we went from simple acronyms like LOL in ‘99 to LMMFAO in ’07 (Yes, someone actually has used that in an AIM conversation before believe it or not).
- In terms of information, our results were 48.7% higher than those of Braithwaite. This may be correlated to the fact that this Google Group was one where specific questions about physical maladies were posed and the most effective way to answer them was to give informational support.
- In terms of tangible assistance, our results were 17.5% higher than those of Braithwaite. Once again, this is likely correlated with the fact the questions asked for specific advice and steps and if one disregarded giving tangible assistance, they simply would not be answering the question. For example, people asked “what should I do” type questions and it would be of no help if someone simply showed emotional support and disregarded the tangible assistance.
- In terms of Esteem support, our results were 21.4% higher than those of Braithwaite. Referring the media richness point made in the third paragraph, since people are more likely to express their emotions in this ‘perceived-richer’ medium of google threads, we feel it is easier to detect esteem support since these answers are a little less “professional” than those that would be received in e-mails (especially eight years ago).
- In terms of Network Support, we had no instances and Braithwaite had 7.1% of his sample have network support. This difference is likely due to our choice of group. If we were in a “my computer has a disability!” group of if our group had messages that “appeared to broaden the recipient's social network, by connecting him or her to others with similar interests or situations, including access, presence, and companions” (p.135) then we would have had some instances of network support.
- Finally, in terms of Emotional support, our results were 10% higher than those of Braithwaite. This is likely due to the media richness point made in the third paragraph of this post as well as the Esteem support post. We do, however, feel the reason that this percentage difference is less than Esteem support is because, often times, one is inclined and even obliged to give someone emotional support when they are feeling disable, in pain or depressed. We feel that the person’s human nature factor outweighs the medium of communication.
Our Thread Links:
8 - "But what if I don't want kids?"
Soyong Lee (green)
Ellis Weng (brown)
We selected and analyzed twenty messaged from a thread on alt.support.childfree under Google groups - a “childless by choice” support forum - according to a scheme devised by Braithwaite et al. We focused on the responses following a post written by “Annie” who was debating whether or not to have a child and remain single, and numerically coded them according to six factors:
• Information: the exchange of advice or knowledge of some kind.
• Tangible assistance: aid offered in a very specific way, such as loaned money.
• Esteem support: relief of blame, affirmation/validation of the user’s situation.
• Network support: referral to others that can provide additional support.
• Emotional support: sympathy, empathy, and encouragement.
• Humor: aspect of a message intended to entertain (not part of the original study but shown below).
Our results ranked information highest (60% of the messages), then emotional support, and humor (both 55%), esteem support (40%) and lastly, network support (15%).We found no tangible assistance. Our results compliment Braithwaite’s study, since she found that information, emotional, and esteem support occur with the highest degree of frequency and tangible assistance tends to occur the least; however, her study yielded the highest in the emotional support category instead of information. The difference in the specific numbers can possibly be attributed to a difference in the central focus of the support forums analyzed. Braithwaite et al. focused on Usenet groups for individuals with physical disabilities; living with physical limitations is not uncommon and there is a wealth of information and resources available online. Remaining “childfree,” however, is more of a social issue and therefore the content and tone of each message was radically different. Most of the members are female who dislike children in general, and feel pressured by society to conform to what they perceive as a “mandatory parenting” standard. The posters always sought to inform, but had very set opinions on the subject so they did less supporting and more arguing than was probably intended. Braithwaite didn’t code for humor, but recognized its importance within the groups in her experiment. A couple of the messages we analyzed were sarcastic and spontaneous one-liners posted solely with the intent to amuse, often at another group member’s expense. The absolute lack of tangible assistance is unsurprising given the subject matter and the fact that online spaces do not usually lend themselves to this type of support.
We also found our inter-rated reliability to be 71%, which is slightly above what could be considered “chance” agreement but questionable in terms of statistical significance. We examined this further in an attempt to figure out why our agreement was so low. Working in a group of three caused our inter-rater reliability to be relatively lower than a group of two. In our situation, all three of us had to come to a general consensus to be able to mark “agree.” After analyzing our results, we found that our coding was affected by many factors: our own biases, the type of support, and the length of the message. Our own biases and our individual personalities caused us to judge the message differently. For example, Milan and Ellis found twice as many posts to be humorous because their sense of humor tends to be more sarcastic. The definition of humor can also cause differences in coding because humor is subjective; something can be funny to one person but not to another person. The same goes for information in our topic. What exactly qualifies as “information” when it comes to deciding whether or not to have a child? Because of these vague definitions, humor and information were the least agreed on between the three of us.
Another factor that we found to play a large role in our coding disagreement was the length of a post.
Originally, we noticed that we agreed more when the post was very lengthy. We thought that there might be a direct correlation between agreeability and the length of a post. After counting every word in each of the post and the number of supports that we agreed on in each of the posts, we found that there is not a direct correlation. Instead, we found that it was harder to code posts that were between 20-180 words. Shorter posts, less than 20 words, were actually really easy to code because they usually contained no support whatsoever. The length of a message and the vagueness of a definition of support both played roles in determining our agreeability.
Walther & Boyd's Social Support Theory found four dimensions of attraction involved in online support behavior which were clearly identifiable in our analysis:
• Social distance: the nature of the internet allows for a wide-range of individuals with varying opinions and levels of expertise to interact. The general consensus is that remaining childfree is not something most people would be comfortable discussing with their “breeder” acquaintances since the latter would not be able to understand their position.
• Anonymity: the original poster and those that replied felt comfortable airing their opinions, concerns, and personal anecdotes without fear of negative social consequences due to the relative anonymity provided in online spaces. The affects of anonymity on self-disclosure is a recurrent theme in online social interaction; an increase of the former invariably leads to an increase in the latter. Given that, it can be surmised that the members considered themselves to be mostly anonymous considering they often referenced their work or personal lives.
• Interaction management: Users are able to take as much time as they require to script and edit their responses in CmC which allows for a greater degree of control over what they choose to express. This is also one of the benefits of an online support forum since in FtF interactions one usually doesn’t have as much time to compose their thoughts and words before speaking. Google groups are unique in the sense that they offer “timed” messages – a poster can choose the amount of time that will lapse before their message is automatically deleted from the server. In this way, users control how long others have access to their responses and information (this also ties into anonymity).
• Access: threads are always accessible and users are free to create new ones/reply to old ones whenever they feel it necessary.
8: Social Support for Polygomy
Members: Lina Lee (Blue), Jennifer Yao (Purple)
Description
The study conducted by Braithwaite, Waldron, and Finn documented types and the extent of social support messages exchanged by people with disabilities in a computer-based support group. To determine whether the patterns that were found are generalizable to other groups, we coded 20 postings from a thread that discussed a male who is concerned with dating someone who practices polygamy by focusing on the following six forms of supportive behaviors.
Messages that convey instructions such as a) advice, b) referrals to experts, c) situational appraisal and d) teaching.
Tangible Assistance
Offers that lend physical action such as a) performing a direct task, b) active participation and c) expressing willingness.
Network Support
Messages that appear to broaden the recipient’s social network, by “connecting him or her to others with similar interests or situations.”
Esteem Support
Messages that validate the recipient’s “self-concept, importance, competence, and rights as a person.”
Emotional Support
Messages that include attempts by the sender to express sympathy, support and emotional expressions.
*Humor (not included in Braithwaite et. study)
Messages that are intended to amuse including jokes, sarcasm, and irony.
| | | | | | |
| % inter-rater reliability | 0.825 | | | ||
| | | | frequency | % of msgs | |
| Information | | 16 | 70% | | |
| Tangible assistance | 0 | 0% | | ||
| Esteem support | | 6 | 30% | | |
| Network support | | 0 | 0% | | |
| Emotional support | 18 | 50% | | ||
| Humor | | | 5 | 25% | |
| | | | | | |
Results and Analysis
Based on the analysis of the coding data, inter-rater reliability came out to .825, which is above the chance agreement (.70). This means that agreement on the six forms of supportive behaviors was not by chance and therefore, statistically significant.
The results of our study indicate that in general certain types of mediated social support are more often exchanged than others but the differences between Braithwaite et al. and ours imply that some forms of support used in discussions about relationships compared to messages concerning disabilities are relatively unique.
Unlike Braithwaite et al., information messages were the most frequently found, constituting 70% of total support messages, whereas Braithwaite et al. noted emotional support to be the most prominent. This may be because emotional support is more likely given when the recipient is experiencing distressful circumstances that are not controllable while informational support will be more useful to the male who can choose whether or not to date one who practices polygamy.
Network support was more prominent in messages concerning disabilities (7.1%) than in our study (0%). Network support was high for people with disabilities because due to their physical limitations they cannot find network support ftf as they would online and so it is a highly valued support. The reason why we found 0% is because sexual orientation is less visible and not really talked about and so it may be difficult to even know someone who is polygamous.
Similar to Braithwaite et al, that found a small frequency percentage for tangible assistance (0%). Similar to Braithwaite et al., tangible assistance was least commonly used. Given the lack of physical proximity, offering tangible support is difficult
Unlike Braithwaite et al., we included humor, which came in fourth constituting 25% of the messages. Humor may have been found because people perceive relationship issues as less serious and sensitive and therefore are more likely to include jokes.
Connecting to Theories
According to Wallace, an increased number of people result in reduced helping behavior due to individuals having decreased "noticeability" as well as diffusion of responsibility. This theory contrasts with Walther and Boyd which states that individuals are more prone to social support online. Based on the social support thread based on polygamy, Walther and Boyd were correct. Walther and Boyd's four dimensions were identifiable in the thread; the relevancy is as follows:
1. social distance: The creator of the thread stated that he had never talked to anyone about his own concerns in being with someone who practices polygamy. The social distance of the Google groups mostly likely served as a comfort because it does not require physical interaction.
2. anonymity: Anonymity allowed him to divulge his concerns while those who supported him divulged personal situations as well without fear of embarrassment.
3. interaction management. Users were able to edit their responses and take more time responding, which allows for users to have more control over what they say.
4. access: The thread is available 24/7, which allows the user able read support and advice whenever desired.
Thread:
Assignment 8: Coding Social Support
Group analyzed: alt.support.menopauseThreads used: Threads that we used: http://groups.google.com/group/alt.support.menopause/browse_thread/thread/273950388612d4d1/46959ad61f246a15?hl=en#46959ad61f246a15; http://groups.google.com/group/alt.support.menopause/browse_thread/thread/726e546d763fec18/de4374eca086398e?hl=en#de4374eca086398e; http://groups.google.com/group/alt.support.menopause/browse_thread/thread/36d51bcb7349cdde/42c1ca08be6ffec8?hl=en#42c1ca08be6ffec8; http://groups.google.com/group/alt.support.menopause/browse_thread/thread/dfa3ffac75ccc73c/68ad85f01a017eb0?hl=en#68ad85f01a017eb0
In order to determine the level of social support found or seen in online groups, Braithwaite et al. “coded” messages. In our homework assignment, we read messages on a menopause support group. It was interesting to see how our results were analogous yet different from the results found in Braithwaite et al. (1999).
Our results:
Information Support 85%
Tangible Assistance 0%
Esteem Support 25%
Network Support 0%
Emotional Support 25%
Humor 30%
Inter-rater reliability 95%
Information, or support that individuals seek in areas of advice, referrals, situation appraisal, and teaching, was seen often in the messages our group coded. We believed this was so because of the support group we looked at. Menopause creates many questions for women experiencing it, or women who are curious about that stage in life. We found that many of the messages we analyzed contained questions regarding medical advice—should X get surgery? Should Y go to another doctor? Which doctor should Z go to? The majority of the posts had information-related subjects. Braithwaite et al. (1999) found that ~30% of social support messages contained information support, but we found more than double that percentage!
Because menopause is a very emotional subject that affects the lives of all women at a certain time in their lives, we predicted and found that there was a decent amount of emotional and esteem support. Braithwaite et al. (1999) found ~20% of esteem support and ~40% of emotional support. This is very similar to our findings (25% for both types of online group support). In the messages we coded, we discussed the concept that people needed others to confirm their decision-making skills, and we saw specific examples from some members’ lives. For example, if I needed help making a decision, I would have been supported with confirmations and personal examples why that decision was a good decision. Esteem support is support that strengthens one’s individual importance and significance through compliments, and emotional support is support given through encouragement, prayers, and affection.
Our group found zero tangible assistance and zero network support which is similar to Braithwaite et al.’s results where both tangible assistance and network support was found in less than 10% of all messages analyzed in the study. Tangible assistance is when someone performs a direct task or expresses willingness to aid others and network support is connecting to similar others. We predict that neither of these was illustrated in our results because of the nature of the “online” group. There are few people who really comment and identify with this menopause support group, but most of them probably do not live near one another—it would be much to far and difficult to travel to each other to provide tangible assistance. There was no network support group because of the small amount of people who identified (and by identified, I mean posted) with the group; there is a chance they would have known each other prior to joining the online support group.
We found much humor in the messages we coded. We assumed that occurred because of the connection and bonds that humor creates. There’s a reason why people say laughter is the best medicine, and even though these women may not necessarily be sick, laughter is a cure for many things.
It was shocking to see how high our inter-rater reliability was; I think it is because of the messages we chose to analyze, and that we have pretty similar thought-patterns. It also wasn’t too hard to code the posts because the messages seemed pretty clear.
The major themes we saw revolved around anonymity and reciprocity. The anonymity allows more uninhibited conversations, especially since the people who post only need a username (which doesn’t do much to identify who they are or could be in real life). The anonymity extends even further to the extent that people posting can diffuse the responsibility by saying a friend or a family member is the one who needs the support—there is no pressure to uphold a self-impression and people can be really honest and truthful with one another. Reciprocity was a common theme also because we saw the same few names repeating, and so the same few people interacted with the same few people in this small network.
8: "Although I'd like my daughter to suck less--"
Information. This group seemed like it was primarily a device for other suckers' parents to give advice to the poster. As such, all the messages were informational. (One guy just quoted Dr. J. Brown: "Orthodontia is cheaper than therapy.") On top of advice, the parents appeared to love telling stories about their own kids (see E. Colwell, above), most of which were relevant. It should be noted that this was a far higher percentage than Braithwaite found.
Tangible Assistance. Seriously? Between strangers, on the Wild Wild World Wide Web? Parents whose biggest issue is that their kids keep their thumbs out of their mouths aren't begging for money or any other tangibles, and Original Poster Penny Ginn wasn't offered any.
Esteem Support. This wasn't exactly what Penny was looking for either, and the responses seemed to acknowledge that. No one was saying "your child sucks her thumb, so you're a bad mother"--who knows why children do it, aside from the obvious, immediate benefits of the warm sensation, the interesting taste, the instant sense of security, and the never having to deal with an itchy palate. However, several messages contained stories where parents defended their own esteem, which we felt was aimed at the OP as well.
Network Support. Again, this was a forum for people to share their own experiences and then interpret them--OP Penny didn't need connections to other people who might have knowledge about clinical thumb-sucking disorder. The Amateurs-R-Us, no-credentials-necessary game of thumb-sucker supporting involves no experts or second-uncles-you-should-talk-to. Under the assumptions that the replyers weren't themselves network support and that Dr. J. Brown didn't count, we found no network support messages.
Humor. We were very satisfied with our reading material--fun to read and full of laughs. Suckers seem to be a light enough topic that humor is an appropriate aspect of their description, especially the parents who trust the Internet to forgive them for the awful lies they've told their own children ("what happened to Grandpa's [missing] thumb, Mommy?"). Some posters took themselves completely seriously and didn't even pretend to find any humor in sucking, but we found that half of them had at least one humorous statement in their message.
So why don't our numbers match Braithwaite's? Again, it's the nature of the subject, and probably partly the online environment. This topic in "misc.kids" wasn't a forum of close-knit thumb combat veterans--it was just a bunch of strangers throwing stories around about a "mildly deviant habit." Those restrictions, cutting out any serious, heavy topics and any non-strangers who might feel more comfortable offering network support or tangible assistance, lead Margarethe and to I believe that if Braithwaite had studied only suckers, the results would have mirrored ours closely. And maybe it would have shed some extra light on this most grave of childhood afflictions. After all, there's one born every minute.
--Ken Colwell, Margarethe van der Tuin
8 Online support groups
- Information. Passing on any kind of knowledge of advice.
- Tangible assistance. Performing a direct task or offering direct assistance.
- Esteem support. Validating a user’s importance, competence, and rights.
- Network support. Expanding a user’s social network to others who can provide support.
- Emotional support. Expressing sympathy or encouragement.
- Humor. Any intent to amuse.
All but the last feature was also numerically coded in Braithwaite’s study.
We decided to draw messages from a variety of threads, rather than focus all our message analysis on only a handful of threads. Many posts within a given thread tend to carry a similar tone and support type. This effect may happen because the way in which the thread starter worded his post influences subsequent response content. By choosing from separate threads, we may be able to avoid an unnecessarily biased set of results.
Our results largely agreed with Braithwaite’s analysis. We saw more informational, emotional, and esteem support more than any of the other support types. Furthermore, tangible and network support were extremely rare. Unlike in Braithwaite’s results, we had more informational support than any single other support type (in Braithwaite’s data, emotional support appeared the most). Our data also indicates that humor appears in moderate amounts on support groups. It is definitely more common than tangible or network support, but it doesn’t occur nearly as often as information-related or emotion-related support.Information support might have been so common because of the types of inquiries made by thread starters. Many people knew exactly what type of advice or suggestions they were seeking. This straight-to-the-point kind of post will tend to yield informational responses in order to satisfy the original poster’s request. It should also be noted that most of our messages involved a mental-related issue rather than a physical disability. Braithwaite mentions that those with non-physical disabilities can often do something about their situation, whereas those with physical disabilities cannot. Therefore, mental support messages should be mostly informational, while physical disabilities should generate more emotional support messages. The informational, esteem, and emotional messages we saw were generally intended to assuage any feelings of despair and hopelessness in the subject, which is a primary goal in support forums. Therefore it makes sense that these three types of messages were the most frequent.
Online support group members are generally not within close physical proximity. Thus there is little incentive to provide tangible support, since it would be extremely inconvenient to do so. Network support is also rare because the act of participating in an online support group automatically connects you to the support network of individuals who monitor the group’s activity. Our data only yielded one tangible support example and one network support example, supporting these explanations.
We found online support groups to be responsive to help requests in general. The “bystander effect,” whereby increasing the number of people decrease the likelihood of helping, does not apply in CMC because the medium is asynchronous and connects physically distant users. These two features create the illusion that there isn’t a huge group currently examining a thread without any replies. Thus, diffusion of responsibility is lowered and people feel more inclined to reply.
Walther’s dimensions of attraction online social support can help us explain certain features of our results. Social distance, or the physical distance separating users, helps online support groups by bringing together a larger pool of expertise. Moreover, this expertise can be accessed by anyone at any time, without having to meet in person. This is perhaps the single most valuable asset of online support groups and possibly why our data has so many information support messages. It also explains our lack of tangible support messages – online support groups are utilized precisely because they offer support one cannot get from close proximity. Anonymity allows users to self-disclose potentially embarrassing details without having to worry about what people think of them later on. This may give other users the insight necessary to provide proper esteem and emotional support.
Partner: Sara Jih (Red blog)
Messages:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Assignment 8: Alcoholism, Eating Disorders, Cheating and Sexual Dysfunction ...
BLUE: Sherrie Chavez
For this assignment, our group analyzed a number of support groups through Google. We used discussion boards about alcoholism, eating disorders, cheating, and sexual incompetence, to code the main types of support behavior from the Braithwaite study: information, tangible assistance, esteem support, network support, and emotional support. The messages we read ranged in style from empathetic to angry, from personal to formal, from validating to discouraging. Many threads we found were just irrelevant, unconstructive, inflammatory messages, so we had to be careful to find substantial, positive comments that we could code.
| % inter-rater reliability | 0.8083333 |
| ||
|
|
|
| frequency | % of msgs |
| Information |
| 17 | 0.85 | |
| Tangible assistance | 2 | 0.1 | ||
| Esteem support |
| 10 | 0.5 | |
| Network support |
| 3 | 0.15 | |
| Emotional support | 11 | 0.55 | ||
| Humor |
|
| 3 | 0.15 |
Our inter-rater reliability was high at 80.83% meaning. We believe this is because the topics we coded on were fairly straightforward and easy to understand. Furthermore, Walther and Boyd's theory of social support says that social support online is uniplex and has no nonverbal cues. Therefore, they are simpler and easier to understand.
The percentage of the messages we coded with information was 85%, which was significantly higher than the 31.3% in the Braithwaite study. In our study, the initial poster on the discussion board was asking for direct advice to cope with their problem so most responses included some sort of advice or suggestions (information) to the person. "Message coded as information support appeared to reduce uncertainty or help make life more predictable for the message recipient" (Braithwaite). In general, the people who frequented the message boards we analyzed are people who wanted to share their knowledge from past incidences and difficulties with people who were experiencing them now such as previous bouts of alcoholism or eating disorders, or experiences with cheating partners.
Our results for tangible assistance and network support were low in comparison to the other types of support, similar to the results of the Braithwaite study. Part of what contributes to this is the anonymity involved in CMC. When people are anonymous and don't have direct accountability to help anyone, they will provide less tangible assistance and be less inclined to give them actual contacts to get help. Also, being anonymous makes it harder to make direct connections with people and provide network support. If people don't know the person they are trying to help they have no reason to extend themselves to provide substantial assistance. Another factor of Walther's four dimensions of attraction to online social support is access. With such a variety of people and messages over time and space, it makes it very difficult to provide tangible assistance. Furthermore, the people who are on online message boards are not usually experts or therapists; they are just average people. Therefore, they don't have the knowledge or tools to provide someone with tangible assistance.
The results we had for emotional support were comparable to the results from the Braithwaite study, but the results for esteem were much higher. Both were between 50-55%. This is partially because the topics we analyzed were very conducive to positive feedback and encouragement. People who were cheated on were constantly told "you deserve better than that"; the people who were drinking were complimented on the initiative they took to find help. Also, we think that the levels of esteem and emotional support were much higher than the levels of tangible and network support because those kind of comments only involved words and did not have to be backed up.
The results of our coding shows that while much of our analysis was similar to the results of the Braithwaite study, some of it vastly differed. This can probably be attributed to the difference in the types of topics and people involved. Furthermore, Walther's four factors of attraction to online social support (social distance, anonymity, interaction management, and access) were evident in explaining how much and what types of support people offered.
http://groups.google.com/group/alt.support.eating-disord/browse_thread/thread/1e0c33a0efb36a29/c704006c97171a41#c704006c97171a41
8: Saving Marriages, Online
http://groups.google.com/group/alt.support.marriage/browse_thread/thread/625e9a6c38b75565/0b20232db627c227?q=troubled+marriage&lnk=nl&
Our Data:
| %inter-rater reliability | 0.9 | ||
| frequency | % of msgs | Braithewaite | |
| Information | 11 | 55 | 31.3 |
| Tangible Assistance | 0 | 0 | 2.7 |
| Esteem Suppor | 4 | 20 | 18.6 |
| Network Support | 6 | 30 | 7.1 |
| Emotional Support | 8 | 40 | 40 |
| Humor | 2 | 10 |
We chose to look at a marriage-support thread. We did this because we assumed we would have the most luck with a support thread (naturally), and marriage was one of the active topics on google (says something about marriage). The specific thread we chose was titled "Troubled Marriage" where people can post specific difficulties their marriages are having, and get support and advice from the community.
Information
These include advice, referrals, teachings, and, well, information
Intangible Assistance
These are offerings of actual assistance, ex: “I will come over and help you do _____”
Esteem Support
These comments work to support the author’s individual self-esteem. Ex: compliments
Network Support
These kinds of messages tie in others with similar circumstances.
Emotional Support
This one is pretty self-explanatory. Offering support through affection, understanding, and motivation.
One interesting thing of note is the fact that we chose to also look at humor. While Braithwaite doesn't look at humor, it is interesting to look at as humor is an important part of human interaction. In our thread we found that about 10% of the messages included humor in them. At first glance, this is surprising considering that the posts were responding to a call for help about a struggling marriage. After reevaluating first impressions, perhaps humor isn't that surprising after all. With such an intense topic, perhaps some thread members felt it necessary to lighten the mood a bit before contributing advice. With regards to theories, we weren't quite sure what to attribute this to, but perhaps humor emerges from difficult situations, especially in interactive group settings.
Assignment 8
Matt Rawding (Brown Group)
Emily Etinger (Blue Group)
In this assignment, we examined social support in online spaces. The space we examined was a thread in Google Groups, Eating - alt.support.depression, about a father who had a young daughter who wasn't eating. He was looking for advice on how to get his daughter to eat. This thread contained many messages from various people and each message contained some aspect of support as defined by Braithwaite et al. We found that the most common aspect of social support was "information" in the messages. We also didn't find any messages containing "tangible assistance" or "network support."
There were some similarities between our results and the results in the Braithwaite et al study. In the study, "tangible assistance" and "network support" were the least common traits in support messages, and our findings were similar. Because our sample size was so small, we didn't even have any examples of either trait. In the messages we examined, we found that 40% of them contained emotional support, which is the same amount that the study found. However, the main difference between our results and the results in the study was the percentage of messages that contained "information." Our study found that 90% of the messages had "information" whereas Braithwaite et al only found 31.3% of the messages to contain "information." One reason for this could be that the members of the thread that we examined didn't know each other very well. In a forum where members post often and develop online relationships, it is more likely that the messages there will have "emotional support" or "esteem support" and less "information." There was also one person in this thread who posted very short messages, each with only "information" on how to resolve the problem, but no support in other forms. All of the messages we read exhibited at least one of the social support traits.
The creator of the thread is sharing something very personal over the internet, and strangers are more than willing to respond and give advice. These actions align with social distance theory. People are more willing to open up and share details about themselves over the internet than they are face-to-face. The creator of this thread was able to get advice from strangers without having to leave their home (assuming they own a computer), advice they would not have received if they had spoken face-to-face with someone they knew. Another interesting thing we noticed about threads was that many of them are not very active. Wallace explains this with her theory of numbers on the internet. She says that someone reading a thread may not respond because that person will believe there are many other people willing to help so it is not important for them to respond.