Unfortunately, I missed last week’s classes because I was sick and, consequentially, I waited until Monday too start this assignment, which meant that I was unable to find a partner in time. Because I was unable to get a second coder, there was no opportunity to calculate an inter-rater reliability percentage. However, I did manage to code 20 messages under the same thread. The thread I chose to discuss was a “depression support group.” Here is my table:
inter-rater reliability 0
frequency % of msgs
Information 15 0.75
Tangible assistance 8 0.4
Esteem support 9 0.45
Network support 9 0.45
Emotional support 10 0.5
Humor 4 0.2
Essentially, I found a lot of incoherence from one frequent messager, “David.” For the most part, his messages consisted of short blurbs and random haikus and he rarely provided any focused posts (I still don’t know why). Therefore, doing what I could for this assignment required me to sift through some messages and decide on substantial and meaningful “support-like” messages. In my table, Information was most frequent comprising 15 of the 20 messages. Braithwaite says “messages coded as information support appeared to reduce uncertainty or help make life more predictable for the message recipient.” (Braithwaite et al, 1999) I found that most of the messagers provided responses and feedback to other people within the group. Usually it was set-up where one person would post a desperate message almost begging for someone to help them, and responses would come with inspirational advice, links to websites, and attempts to diagnose the messager’s problems. Also in this particular support group, situation appraisals where a “redefinition of circumstances” was a frequent aspect of information type posts. On rare occasion, I found humor in some messages. These seemed to be an attempt to make light of their depressing descriptions of their situation. I did, surprisingly, find some Tangible Assistance between two messagers that seemed to have known each other outside of the group or could have just been “intersecting frequently” within the group. Usually, it was the expression of willingness to help whether it be meeting each other to discuss issues, or taking someone to a psychiatrist for help. Another frequent and interesting code was Emotional Support. Comprising 10 of 20 posts, I found that Emotional Support came, for the most part in the forms of encouragement and sympathy. In some cases of “David the Haiku Herald” I considered his short, random messages to be emotional support—like a prayer. Braithwaite says that in her study, “any referrals to experts (e.g. physicians, counselors) were coded as informational support.” Therefore, any direct information for a psychiatrist was not considered Network Support. However, I found that some responses were very helpful in the form of Network Support. Often, not only was advice included in the responses, but also direct contact information to other people with depression who were in the group but had not posted recently or other therapy support group links for depression. Esteem Support was frequent, though it was less frequent than I expected from a depression support group. Often, respondents offered Esteem Support in the form of relief from blame. When there was Esteem Support, it made the group seem more cohesive and reciprocal—something Wallace deems important in social support.
In relating my findings to Braithwaite’s study, I found that they were somewhat similar. My results indicated high frequencies in the forms of Informational Support (this being the most frequent), Emotional Support, and Esteem Support much like that of Braithwaite’s results. Also, Walther & Boyd’s study was also applicable. I fid that the increased anonymity in online support groups (both perceived and real) were especially important when avoiding embarrassing or negative messages and I believe that in the case of responses to depression people were more apt to provide positive feedback with helpful support. In one case, someone used “%” as their username and still received the same amount of support and confidently provided their personal circumstances when dealing with depression. I also think that Walther & Boyd’s description of “access” of support groups became important in groups for depression. Granted that users may live in different time zones, I’m sure that some if the 4:55 a.m. posts were really in the middle of the night or that early in the morning and I’m sure that it helped to elicit quicker responses and positive feedback to some people who were in extremely desperate situations. This 24-hours/7 days a week online accessibility becomes very useful when people are suffering from insomnia or even thoughts of suicide! Ultimately, I felt that the online support groups were a good way for someone who may be a bit more timid than most in such a vulnerable position (I think we all are in these type of situations) to discuss their problems and search for help.
Link:http://groups.google.com/group/alt.support.depression/topics?lnk=srg
Tuesday, October 30, 2007
GIMME GIMME GIMME GIMME (I think he has ADHD) Assignment 8
Due to personal issues occurring this weekend I was unable to find a partner within 245 to do this project with however, I asked one of my friends to read the material (Braithwaite et al) and help me out with the coding.
Braithwaite et al (1999) used "the five-category system developed by Cutrona and Suhr
(1992): information support, tangible assistance, esteem support, network support,
and emotional support" to code over 1400 messages.
Using the method outlined in the aforementioned study, my friend and I proceeded to separately code 20 messages in one thread.
The thread we chose dealt with diagnosing a 3 year old with ADHD. This sparked much conversation about the disease itself and less about how to actually help the woman. This speaks to the volatility of the topic itself. The fact that the person who recommended that the child be looked at was his preschool teacher also sparked a discussion.
Here are our results:
We found information support to be the most prevalent, followed by emotional support unlike the Braithwaite study which found emotional support to be the most prevalent followed by informational support. This was due to many of the messages arguing back and forth about the situation resulting in more of a situational appraisal or advice type message rather than emotional support type messages.
Unlike Braithwaite, we found no messages offering tangible support or network support. This may have to do with the specific situation at hand since there was not much tangible support that could have been offered that would have truly helped this woman. I wasn't surprised that network support had such low numbers since people on support threads often tell stories about situations they have heard of but rarely give you a way to contact the person who they are referring to.
There was not much humor in our messages either but that has to do with the fact that the discussion took a serious turn and since the subject was so volatile, humor would have seemed rude.
The percentages we found overall differed significantly than that of the study but than that was to be expected due to the fact that we focused on only one small thread. Braithwaite's study focused on more of a variety of messages.
Walther and Boyd's four dimensions of attraction to online support are prevalent in this study. Social distance allowed the people to speak their mind freely without physical reprimand. Although one person was called out on their misinformation, for the most part, people stayed friendly and the discussion moved peacefully.
Anonymity helped in allowing the woman to post a question about her personal life without fear of someone being able to tell who she was. Anonymity also helped in allowing other group members to respond to her post truthfully.
The idea of access is also prevalent. By posting on an online group thread, the woman is ensuring that she is reaching anyone who is in the group as well as random Internet surfers who stumble upon the thread. This allows her question to reach people all over the world and gives her access to more information and advice than she could have found in FtF interactions.
Interaction management was shown when people responded separately to specific parts of a post or when they could take as much time as needed before posting. This allowed a better message to be posted since the message could be specifically tailored.
We used the following thread:
http://groups.google.com/group/misc.kids/browse_thread/thread/8936bebf00b78aee/c5d79eb11e7d8abe#c5d79eb11e7d8abe
Braithwaite et al (1999) used "the five-category system developed by Cutrona and Suhr
(1992): information support, tangible assistance, esteem support, network support,
and emotional support" to code over 1400 messages.
Using the method outlined in the aforementioned study, my friend and I proceeded to separately code 20 messages in one thread.
The thread we chose dealt with diagnosing a 3 year old with ADHD. This sparked much conversation about the disease itself and less about how to actually help the woman. This speaks to the volatility of the topic itself. The fact that the person who recommended that the child be looked at was his preschool teacher also sparked a discussion.
Here are our results:
| % inter-rater reliability | 85% | |||
| frequency | % of msgs | |||
| Information | 16 | 0.8 | ||
| Tangible assistance | 0 | 0 | ||
| Esteem support | 1 | 0.05 | ||
| Network support | 0 | 0 | ||
| Emotional support | 8 | 0.4 | ||
| Humor | 2 | 0.1 | ||
We found information support to be the most prevalent, followed by emotional support unlike the Braithwaite study which found emotional support to be the most prevalent followed by informational support. This was due to many of the messages arguing back and forth about the situation resulting in more of a situational appraisal or advice type message rather than emotional support type messages.
Unlike Braithwaite, we found no messages offering tangible support or network support. This may have to do with the specific situation at hand since there was not much tangible support that could have been offered that would have truly helped this woman. I wasn't surprised that network support had such low numbers since people on support threads often tell stories about situations they have heard of but rarely give you a way to contact the person who they are referring to.
There was not much humor in our messages either but that has to do with the fact that the discussion took a serious turn and since the subject was so volatile, humor would have seemed rude.
The percentages we found overall differed significantly than that of the study but than that was to be expected due to the fact that we focused on only one small thread. Braithwaite's study focused on more of a variety of messages.
Walther and Boyd's four dimensions of attraction to online support are prevalent in this study. Social distance allowed the people to speak their mind freely without physical reprimand. Although one person was called out on their misinformation, for the most part, people stayed friendly and the discussion moved peacefully.
Anonymity helped in allowing the woman to post a question about her personal life without fear of someone being able to tell who she was. Anonymity also helped in allowing other group members to respond to her post truthfully.
The idea of access is also prevalent. By posting on an online group thread, the woman is ensuring that she is reaching anyone who is in the group as well as random Internet surfers who stumble upon the thread. This allows her question to reach people all over the world and gives her access to more information and advice than she could have found in FtF interactions.
Interaction management was shown when people responded separately to specific parts of a post or when they could take as much time as needed before posting. This allowed a better message to be posted since the message could be specifically tailored.
We used the following thread:
http://groups.google.com/group/misc.kids/browse_thread/thread/8936bebf00b78aee/c5d79eb11e7d8abe#c5d79eb11e7d8abe
8
In this survey, we took a thread that a father is worried because his daughter doesn't eat food. http://groups.google.com/group/alt.support.depression/browse_thread/thread/467e40122ae80a49/3b883a7a712e8ca5#3b883a7a712e8ca5 And we coded 20 messages. The result is information support-13, tangible assistance-0, esteem support-5, network support-0, emotional support-6, humor-2. And that means information support is the prevalent as 65%, emotional support is far second 30%, followed by esteem support 25%. This is different from Braithwaite's result. And we think there are several reasons contributing to this difference.The first one, Braithwaite's experimental object is an inability community. Most people leave messages or reply to show their care and the willingness to help. Except for this emotional care, most people actually can give little virtual help. For example, we common people mostly are impossible to cure a handicapped person, but we really can give a lot of emotional support and encouragement through messages. So at first, we want to find a thread that focus on "homesick". Because we think although we can't get the homesick people back to their home, we can give our support through words. We assume that would come out a similar result to the inability community in Braithwaite's experiment. But unlucky, we didn't find one. In that thread, the father is worried about the daughter's eating problem. In this case, most people can give some advice on how to settle this problem, like go to a doctor, don't force her to eat, or play with her when eating. So this seems a problem we common people really can help to solve, so people give more informative support. And as we also think the father is really a great one, and we give esteem support and emotional support. And we also find that, the first few messages are all informative messages focus on how to make her eat, only after the father reply "She is eating an apple now. I am relieved", more people began to change their focus on "You are a good dad". Thus we conclude that people give more informative support when they can solve the problems; and tend to give more emotional support when they can't give actual help.The second one, the number of the sample is very different between mine and Braithwaite. In the survey, there are only 20 messages. As we mentioned above, people first tend to give solutions, and them tend to encourage the father and appreciate him as a good father. So, we think it’s safe to assume that, at first, people give informative support messages. But after a few message, maybe 30, maybe 50, they find that most people can find only the same 5 or 6 ways to let the little girl eat. So then they instead tend to encourage the father, tell him not to be upset, tell him to be optimistic, and tell him he is a good father, that is, emotional support. Because at this situation, they can't find any more solutions and what they can give is emotional support. But, in fact, as there are not too many messages here, we can't check whether it goes on as we assumed. Maybe when there are 400 messages replying the father, the result is totally different.The third one, we can't tell how many and what supports one message gives very accurately. And as the sample number is really small. A little mistake may result in a great difference when turned into percentage. And we didn't find any tangible and network support in the messages, we can't tell how much these factors really weigh.So, as a whole, the category of the object community, the hardness for us to give actually physical or material support; the volume of the sample, number; the validity of telling every support contained in a message really affect the result a lot. And in the survey, informative support is the prevalent support.
Posted also by Yu Tian(Purple)
Posted also by Yu Tian(Purple)
8 People with Issues
Austin Lin (green)
Nick Fajt (green)
Melissa Bernard (purple)
Nick Fajt (green)
Melissa Bernard (purple)
We coded five messages in four separate threads from four different groups: ocd help, HELP! Our son is acting bizarre!!! ...Very worried parents, Women are b****es, and Adderall Problem.
Unlike the Braithwaite study, we found informative support to be the most prevalent, while emotional support was a distant second. This result is, however, consistent with the Cutrona and Suhr (1992) study. Braithwaite attributed the differences in her study and the Curtona study to the different experimental settings examined, claiming the Curtona and Suhr environment was “nonemotional.” Our group examined four different online spaces that were, arguably, very emotionally driven. Therefore our contrasting results are probably not a result of the messages we examined, but instead the way that we used the coding system. Our group found every message we examined to contain informational support. One reason for this may have been because we rated the first five posts on four separate discussion threads which may have been more informative in support than later, more complex threads after the basic information has been laid out. It is also possible that we were being too liberal (or Braithwaite was too conservative) in our coding of this category. However, based on our understanding of Braithwaite’s “informational support,” it is difficult to imagine a support statement that doesn’t include an informational or instructive element. This leads directly into the limitations of this coding process. As evidenced above, the coding structure and style is heavily reliant on the individual’s understanding of each coding category. Another limitation factor that was found in our examination and referenced by Braithwaite, is the inability to incorporate the perspective of the senders, because we lacked self-reported data. We were also hindered by the inability to incorporate the way messages were perceived by the recipients because none of the recipients in the posts we examined ever replied to the support they were given. This is a significant limitation as the state of mind of both sender and receiver is what we are, in many ways, examining.
Despite these limitations, the coding method did allow us to gain a reasonable understanding of the types of support offered in online groups. Our group noticed that the strongest forms of support generally included multiple categories of coding, while the weakest forms of support just exhibited information support. The support responses that appeared most thoughtful and thoroughly engaging tended to exhibit informational, esteem and emotional support. As tangible assistance was only witnessed once and network support was never witnessed it is hard for us to really weigh in on how they might affect the strength of a support statement. Braithwaite’s reasoning aligns closely with ours in that tangible assistance often requires physical proximity and that users are fulfilling their neet for network support just by participating in a the discussion. The humor element didn’t really appear to add or detract significantly from the strength of support, and in many ways the humor seemed to appeal more to third party readers rather the recipient themselves as they did not add to the content of the post.
Threads
alt.support.anxiety-panic
ocd help
http://groups.google.com/group/alt.support.anxiety-panic/browse_thread/thread/16a3e744e1af418d/
alt.psychology.personality
HELP! Our son is acting bizarre!!! ...Very worried parents
http://groups.google.com/group/alt.psychology.personality/browse_thread/thread/b07af8d974b301d5/
alt.romance
Women are b****es,
http://groups.google.com/group/alt.romance/browse_thread/thread/a9c41677f1c3a62a/
alt.support.attn-deficit
Help--Adderall Problem
http://groups.google.com/group/alt.support.attn-deficit/browse_thread/thread/4e5aeece5090b1be/
A8 (Group). I've Got A Feeling. I Feel It :: alt.support.ocd
by: Andrew C. Meehan (Green), Brian Isett (Brown), Robert Jerry (Red)
Self-help and social support groups—like many socially-focused practices—have found a home on the “boards” and forums of the internet. Braithwaite, et al, examined the integration of social support groups into the internet medium to observe how this “permanent fixture in American culture” had adapted itself to the CMC environment (Braithwaite, 1999). The research team coded over 1,400 support messages according to five social support supratypes, outlined by Cutrona and Suhr (1992). The original supratypes’ sub-categories were slightly modified by the research team to best fit their specific medium and topic of research.
In the manner of the Braithwaite study, our team of three separately coded 20 messages located in the “alt.support.ocd” forum of Usenet. The results, including the inter-rater reliability of the coders, are included below:
Figure 1
| % inter-rater reliability | 0.875 | |||||
| frequency | % of msgs | Modified % | ||||
| Information | 13 | 0.65 | 34.21 | |||
| Tangible assistance | 0 | 0 | 0.00 | |||
| Esteem support | 10 | 0.5 | 26.32 | |||
| Network support | 10 | 0.5 | 26.32 | |||
| Emotional support | 5 | 0.25 | 13.16 | |||
| Humor | 4 | 0.2 | ||||
A statistical comparison of our frequency and percent distribution with that of Braithwaite’s revealed a significant difference between the respective scoring distributions among supratypes. Our inter-rater reliability supports the nation that each coder understood the distinction between the supratypes and could code accordingly.
Analysis
Using a Chi-square test, or multinomial goodness-of-fit test, with 4 degrees of freedom—Humor was excluded because Braithwaite did not score for it—we found our P-value was less than statistically significant: 0.005 (alpha level = 0.05).
Hypotheses of Test:
Null Hypothesis: There is no significant difference between the distribution of
support messages scored by us and the distribution of support messages
scored by Braithwaite, et al, in the 5 supratypes.
Alternative Hypothesis: There is a significant difference between the distribution of
support messages scored by us and the distribution of support messages
scored by Braithwaite, et al, in the 5 supratypes.
The Null Hypothesis was rejected (P-value less than alpha level), which allows us to conclude that our distribution among the supratypes was significantly different than Braithwaite’s. Quantitatively, we attribute this to our much lower Emotional Support percentages (circa 27% lower) and higher Esteem, Network, and Information Support percentages.
The nature of the support groups coded for—Persons with Disabilities versus Persons struggling with OCD—can partially account for the discrepancy. Braithwaite denoted three “key life functions” most directly related to social support situations involving the disabled: mobility, communication, and socialization. For Obsessive Compulsive Disorder, mobility and communication may not be altered as substantially, and socialization may be the greatest barrier to overcoming an OCD. Furthermore, preventative and/or palliative support can be lent online in the case of OCD, whereas Disabilities are often permanent and CMC advice might tend to lean toward an exchange of “empathy” (Emotional Support) or peer-to-peer stories. Walther and Boyd’s attractive dimensions of online support accounts for these types of exchanges: Social Distance allows people in similar situations to empathize or emote in ways that non-disabled persons could not. Conversely, OCD sufferers rely more heavily on Information and Network Support for aid – here, the dimensions of Interaction Management and Anonymity might become as vital as Social Distance. Many messages on the alt.support.ocd board were referrals to experts’ books or admonishments for professional consultation (more tangible help is suggested).
Tangible Assistance did not factor in our or Braithwaite’s coding, due to the very reasons online support can be helpful – Social Distance and Anonymity. Posters/commenters are not likely to lend physical support to OCD or Disabled patients due to the importance of the aforementioned dimensions of online support.
Our additional supratype, Humor was not scored by Braithwaite, though we found it to be present in 1/5th of the scored messages (see Figure 1). This suggests that there may be a sixth
category that needs to be considered when examining support groups (or the replacement of Tangible Support by Humor for Online Support supratypes). It is easy to see how humor could provide palliative support in these situations; Walther and Boyd’s dimensions, as well as the Hyperpersonal Model, suggest that the freedom of conversation online might allow for a greater exchange of Humor. Inexpertness, as Walther suggests, might lead to personalized but unhelpful messages in family or friend support.
Threads:
OCD fool needs advice badly
OCD?
Mad Cow!
Relationship breakups...
Getting Engaged to an OCD woman
1.6% of us pretend
Help stamp out abuse and torture by Psychologists
Monday, October 29, 2007
8-Calling All Quitters
Rachel Newman-Green
Jamie Hacker-Blue
Using Google Groups, we discovered a blog regarding people that are trying to quit smoking. We used two different message threads since neither had twenty messages. However, they each had to do with support for quitting smoking.
http://groups.google.com/group/alt.support.stop-smoking/browse_thread/thread/2dc2c7cca1c09f71/82629203c95420aa#82629203c95420aa
http://groups.google.com/group/alt.support.stop-smoking/browse_thread/thread/59ec39569520fa3d/7d70219d8eafcd82#7d70219d8eafcd82
% inter-rater reliability 0.791666667
frequency % of msgs
Information 14 0.7
Tangible assistance 1 0.05
Esteem support 7 0.35
Network support 6 0.3
Emotional support 18 0.9
Humor 5 0.25
Most of our inter-rater reliability percentages were very different from the values that Braithwaite found. For information support, we had a 70% correllation rate, which is very different from Braithwaite's 31.3%. Clearly, these values are highly differentiated which can be explained by the environment of the blog itselsf. This thread for people trying to quit smoking is specifically meant for support purposes. Thus, the entire idea behind the thread is for people to constantly be there for one another and provide support when it is needed. This includes support like information support, which includes providing advice, referrals, situation appraisals or teaching. The nature of this thread implies that advice should be given, and the reality is no different. Almost everyone who wrote on the thread provided some sort of information support in the form of advice. Common pieces of advice inluded people telling each other that reading the posts and writing new ones daily would be extremely helpful in refraining from smoking. This is why our inter-rater reliability was so high. The inclusion of information support and advice was evident.
Braithwaite's tangible assistance value was measured at 2.7% while ours was not much higher at 5%. Tangible evidence refers to support in which a respondant actively participates with the writer by performing a direct or indirect task or expressing a willingness to assist others. Most of the thread posts were filled with advice or encouragement rather than specific instructions or tasks as to how to quite smoking. The numbers generally coincide with Braithwaite's research. This is most likely because people are more likely to give out encouraging words of wisdom rather than telling someone to perform a task. There were a few people on this thread that gave suggestions as to how to quit by mentioning what worked for them. However, these suggestions were not all worded in ways that tell the writer to perform a task, but some of them were slightly suggestive.
In measuring esteem support, our value was significantly higher than those of Braithwaite, with 35% of our messages containing esteem support compared to only 18.6% of Braithwaite's messages. Our thread contained instances where the people needed validation that what they were doing was in fact good for them and that they shouldn't blame themselves for waiting so long to quit. Instead, they should embrace their decision to quit and be showered with compliments for making the right decision. This therefore led to the higher percentage of esteem support found in these messages.
Our values for network support were 30%, while Braithwaite's were only 7.1%. Although Braithwaite thought that people didn't need network support because they were already involved in a thread, the people in these messages wanted to reiterate that they were there for each other and ready to lend any help possible. They were aware of the network they were in and how to use it, but the members still felt the need to really express that they were always going to be there for one another, and that all the members should actively use this resource.
Emotional support was extremely prevalent in our findings while only present 40% of the time in Braithwaite's study. The posts that we read have obvious mentions of emotional support because they mainly focus on encouragement, sympathy, and being understanding or empathetic to the writer's quest for non-smoking status. In comparison to Braithwaite, we found such a hugh inter-rater reliability because quitting smoking is all about getting constant encouragement that "you can do it!" which is mentioned constantly in the thread. Almost all of the posts provide lots of variations on encouraging thoughts as to how to keep it up and preventing the urge to start smoking once again.
Even though humor was not given quantitative date by Braithwaite since it seemed to be common to most social support messages, we found in our study that about 25% of the messages contained humorous or satirical content. Some people wrote about their "tragic" weight gain post quitting or their interest in having someone else quit because it would "get [them] off the top of the list!" that everyone fills out with when they had their last cigarette. In a thread such as this, it may help to hear people being sarcastic in order to quit. Everyone can mention encouraging word after encouraging word and that may not work for everyone. For some people, the idea that the ability to quit smoking could be looked at as humorous, could actually be a positive influence on the reality of them quitting.
Although Wallace thought that "increased numbers reduce helping behavior," we found that there were many people that were willing to help despite the large amount of members in the group. Some people replied multiple times to a certain thread, across many threads, and there were some threads that even had over 60 posts, showing that the numbers did not inhibit people from posting.
According to Walther and Boyd, there are four dimensions of attraction to online social support that are social distance, anonymity, interaction management, and access. We found all of these to be present in this thread for people that have quit or want to quit smoking. The social distance was evident where people felt they weren't being judged no matter what they said or did. One woman had tried to quit but had been unsuccessful and was now trying again. Regardless of her previous failure, everyone was completely confident she could succeed this time and they showed no judgment whatsoever. Anonymity was also present because this person could admit her failure without everyone knowing who she really was. This allowed her to express the truth while knowing that she would not be embarrassed by it. Interaction management was especially helpful because each person could craft their post to truly get across the support they wanted. Some posts were quite lengthy, which enabled them to get their ideas across in a clearly contructed and helpful manner. Also, users could post whenever they wanted and take as much time as they needed to respond. Lastly, and most importantly, this thread allowed access to its users every day at all times of the day. Each post also stressed how important it was to express your feelings at any time of day and there would always be someone there to respond and offer encouragement and guidance.
Being able to quit smoking requires a lot of will-power but support from others is essential in order to give the quitter the confidence and advice they need to feel like they are not alone and quit successfully. Braithwaite's measurements of support were all present in these messages, which shows that they are all important in providing the social support that one needs.
Jamie Hacker-Blue
Using Google Groups, we discovered a blog regarding people that are trying to quit smoking. We used two different message threads since neither had twenty messages. However, they each had to do with support for quitting smoking.
http://groups.google.com/group/alt.support.stop-smoking/browse_thread/thread/2dc2c7cca1c09f71/82629203c95420aa#82629203c95420aa
http://groups.google.com/group/alt.support.stop-smoking/browse_thread/thread/59ec39569520fa3d/7d70219d8eafcd82#7d70219d8eafcd82
% inter-rater reliability 0.791666667
frequency % of msgs
Information 14 0.7
Tangible assistance 1 0.05
Esteem support 7 0.35
Network support 6 0.3
Emotional support 18 0.9
Humor 5 0.25
Most of our inter-rater reliability percentages were very different from the values that Braithwaite found. For information support, we had a 70% correllation rate, which is very different from Braithwaite's 31.3%. Clearly, these values are highly differentiated which can be explained by the environment of the blog itselsf. This thread for people trying to quit smoking is specifically meant for support purposes. Thus, the entire idea behind the thread is for people to constantly be there for one another and provide support when it is needed. This includes support like information support, which includes providing advice, referrals, situation appraisals or teaching. The nature of this thread implies that advice should be given, and the reality is no different. Almost everyone who wrote on the thread provided some sort of information support in the form of advice. Common pieces of advice inluded people telling each other that reading the posts and writing new ones daily would be extremely helpful in refraining from smoking. This is why our inter-rater reliability was so high. The inclusion of information support and advice was evident.
Braithwaite's tangible assistance value was measured at 2.7% while ours was not much higher at 5%. Tangible evidence refers to support in which a respondant actively participates with the writer by performing a direct or indirect task or expressing a willingness to assist others. Most of the thread posts were filled with advice or encouragement rather than specific instructions or tasks as to how to quite smoking. The numbers generally coincide with Braithwaite's research. This is most likely because people are more likely to give out encouraging words of wisdom rather than telling someone to perform a task. There were a few people on this thread that gave suggestions as to how to quit by mentioning what worked for them. However, these suggestions were not all worded in ways that tell the writer to perform a task, but some of them were slightly suggestive.
In measuring esteem support, our value was significantly higher than those of Braithwaite, with 35% of our messages containing esteem support compared to only 18.6% of Braithwaite's messages. Our thread contained instances where the people needed validation that what they were doing was in fact good for them and that they shouldn't blame themselves for waiting so long to quit. Instead, they should embrace their decision to quit and be showered with compliments for making the right decision. This therefore led to the higher percentage of esteem support found in these messages.
Our values for network support were 30%, while Braithwaite's were only 7.1%. Although Braithwaite thought that people didn't need network support because they were already involved in a thread, the people in these messages wanted to reiterate that they were there for each other and ready to lend any help possible. They were aware of the network they were in and how to use it, but the members still felt the need to really express that they were always going to be there for one another, and that all the members should actively use this resource.
Emotional support was extremely prevalent in our findings while only present 40% of the time in Braithwaite's study. The posts that we read have obvious mentions of emotional support because they mainly focus on encouragement, sympathy, and being understanding or empathetic to the writer's quest for non-smoking status. In comparison to Braithwaite, we found such a hugh inter-rater reliability because quitting smoking is all about getting constant encouragement that "you can do it!" which is mentioned constantly in the thread. Almost all of the posts provide lots of variations on encouraging thoughts as to how to keep it up and preventing the urge to start smoking once again.
Even though humor was not given quantitative date by Braithwaite since it seemed to be common to most social support messages, we found in our study that about 25% of the messages contained humorous or satirical content. Some people wrote about their "tragic" weight gain post quitting or their interest in having someone else quit because it would "get [them] off the top of the list!" that everyone fills out with when they had their last cigarette. In a thread such as this, it may help to hear people being sarcastic in order to quit. Everyone can mention encouraging word after encouraging word and that may not work for everyone. For some people, the idea that the ability to quit smoking could be looked at as humorous, could actually be a positive influence on the reality of them quitting.
Although Wallace thought that "increased numbers reduce helping behavior," we found that there were many people that were willing to help despite the large amount of members in the group. Some people replied multiple times to a certain thread, across many threads, and there were some threads that even had over 60 posts, showing that the numbers did not inhibit people from posting.
According to Walther and Boyd, there are four dimensions of attraction to online social support that are social distance, anonymity, interaction management, and access. We found all of these to be present in this thread for people that have quit or want to quit smoking. The social distance was evident where people felt they weren't being judged no matter what they said or did. One woman had tried to quit but had been unsuccessful and was now trying again. Regardless of her previous failure, everyone was completely confident she could succeed this time and they showed no judgment whatsoever. Anonymity was also present because this person could admit her failure without everyone knowing who she really was. This allowed her to express the truth while knowing that she would not be embarrassed by it. Interaction management was especially helpful because each person could craft their post to truly get across the support they wanted. Some posts were quite lengthy, which enabled them to get their ideas across in a clearly contructed and helpful manner. Also, users could post whenever they wanted and take as much time as they needed to respond. Lastly, and most importantly, this thread allowed access to its users every day at all times of the day. Each post also stressed how important it was to express your feelings at any time of day and there would always be someone there to respond and offer encouragement and guidance.
Being able to quit smoking requires a lot of will-power but support from others is essential in order to give the quitter the confidence and advice they need to feel like they are not alone and quit successfully. Braithwaite's measurements of support were all present in these messages, which shows that they are all important in providing the social support that one needs.
8. Coding for Suicidal Thoughts
Marisa Lusthaus (Green)
Skyler Sourifman (Brown)
Skyler Sourifman (Brown)
In Braithwaite’s article, “Communication of Social Support in Computer-Mediated Groups for People with Disabilities” he discusses the different types of social support messages exchanged by people in a CMC disability support group. His study of social support was based upon three main issue which include: 1) What types of supportive messages are communicated in CMC support groups serving persons with disabilities? 2) To what extent is social support manifested in the messages of CMC social support groups serving persons with disabilities? and 3) How do the patterns of support messages exchanged by persons with disabilities in a mediated context vary from those reported in previous literature?
In our examination of social support messages we choose to analyze an online depression support discussion group, more specifically a thread on suicidal thoughts. Depression is an illness shunned by our society and can be difficult to discuss face-to-face, however, an online environment can foster a more feasible support system. According to Walther and Boyd, there are four dimensions of attraction that lead to online social support: social distance, anonymity, interaction management and access. All four of these factors are apparent in the thread we chose to analyze and contributed the thread’s success. The dimension of access was extremely applicable in the suicidal thoughts thread because when a person is contemplating suicide, they need immediate support. The CMC environment provides the necessary help without depending on a typical daily schedule like a routine therapist appointment.
After coding the messages, we found an 86.67% inter-rater reliability, which is statistically significant. Although some factors were more apparent than others, all six facets of Braithwaite’s coding scheme were evident in our analysis of the suicidal thoughts group. Our most frequent factors were information and network support, which were both present in messages 85% of the time. Our findings do not follow suit with Braithwaite’s percentages for information (31.3%). These results are what we would expect given the nature of suicide and depression. Many posts implied the importance of hearing others’ experiences and how they were able to overcome suicide; they provided the necessary information to cope with suicidal thoughts. In particular, they shared their experience with various therapists and types of therapy and offered each other advice that could not easily be learned sitting in a therapists’ office. Braithwaite found 7.1% for network support, which is significantly different from our findings. Network support aids in the transfer of information by connecting others with similar situations. When it comes to helping people with depression, it is most important for people to know that they are not alone and the have network to depend on and turn to for help. One particular post praises the importance of this particular network when a member states, “ Yes. And every node strengthens the network”.
In addition, we found emotional support to be 55%, which support Braithwaite’s findings of 40%. Typically, people who are suffering from suicidal thoughts are in need of empathy and understanding that they have people who care about them. As one person states, “I might have mentioned before, "no, bad Lisa! - you have a family that needs you, so you are not allowed to do that!". For esteem support we found it to be prevalent in 45% of the messages while Braithwaite only found it to be evident 18.6% of the time. This discrepancy can be explained by the difference in help offered to those with a disability vs. a mental illness, namely depression. People with suicidal thoughts need to understand that its not their fault and need that extra reassurance that they are worthy of living. Our analysis found tangible assistance to be evident 30% of the time while Braithwaite found it to be 2.7% of the time. The large discrepancy in our findings can be expected because suicidal thoughts have a more immediate need for tangible help than people with disabilities. Also, contemplation of suicide requires more drastic action and expressing willingness, to ensure that others know that they have someone to talk to and not act on their emotions. For purposes of our analysis, we also studied humor (10%), which is logical give the severity and seriousness of the issue. Overall, the discrepancies in our findings when compared to Braithwaite’s findings may also be contributed due to the fact that our sample was very small (20) and Braithwaite’s sample was very large (1,472).
Link To Suicidal Thoughts Forum
#8: Coding For Child Rearing
Alice Choo (green)
Yang Zhang (purple)
Sue’s post
http://groups.google.com/group/misc.kids/browse_thread/thread/32665b5d88ffcc27/1551ef44eff22687#1551ef44eff22687
Funmum’s post
http://groups.google.com/group/misc.kids/browse_thread/thread/98630506c3319f13/46ef6101a323094a#46ef6101a323094a
Phil’s post
http://groups.google.com/group/alt.support.step-parents/browse_thread/thread/4dfff213af102230/a894452e07116c68?hl=en&lnk=st&q=child+raising#a894452e07116c68
For our assignment, Yang and I decided to examine posts that dealt with raising children. The first group we examined was misc.kids, where we found two threads to evaluate. The first thread we evaluated was originally posted by Sue, a mother trying to find ways to get her ten-year-old daughter to sleep in her own room. Sue was frustrated in constantly having to assuage her daughter’s fears. The second thread in misc.kids that we evaluated was by “funmum,” a mother who was bored of entertaining her two-and-a-half-year-old with the same activities every day; thus, she asked for advice as to new activities they could engage in. The last thread was in the group alt.support.step-parents. Phil, a father to thirty-month-old twins and a stepfather to a nine-and-a-half year old, was having issues with his wife as to how much of a role he should have in disciplining his stepdaughter, with whom he had lived since she was three years old. Phil wanted to know other people’s experiences in disciplining their stepchildren.
Unlike Braithwaite, the highest number of support messages we saw were informational support messages; they consisted of 100% of the posts. The other types of messages were much lower than in frequency than that of the informational support messages. Humor came in second in terms of frequency with 25% of the posts, and esteem support came in third with 20% of the posts. Emotional support, network support, and tangible assistance were not very prevalent in our study, appearing in 10%, 5%, and 0% of posts, respectively.
One reason that informational support messages dominated the threads may have been because of the online environment. According to Braithwaite, “information support…is most useful and prominent when the recipient can control the situation and put the information to use.” The original posters were all parents who could easily make use of the advice that other posters gave them. Indeed, the original posts were all focused on getting information and advice rather than any other support.
Even though emotional support was highly prevalent in Braithwaite’s study, we did not find many instances of it. Braithwaite asserts that “emotional support is more likely to be given when the recipient is experiencing distressful circumstances that are not subject to his or her control. This type of support can promote comfort and healing in such circumstances.” Contrary to Braithwaite’s observations of Support Network, Yang and I studied an environment where parents did have control over their situations; thus the need for emotional support was lessened. Indeed, parents did not seek “comfort” or “healing”; they were simply searching for possible solutions to their problems. Their posts concentrated on getting the information that they wanted rather than fulfilling emotional needs.
Esteem support was third in terms of message frequency, which matched Braithwaite’s analysis. Posters sometimes validated the original poster’s experiences. For instance, with Sue’s post about her daughter refusing to sleep alone, one poster identified with Sue’s frustration by saying, “I can see how you [Sue] would be over it!” With esteem support, posters aimed to lift the original posters’ spirits by validating their feelings and experiences.
Our findings for messages related to network support and tangible assistance were similar to those of Braithwaite. As Braithwaite mentioned, the low number of network support messages may be attributed to the fact that “members met their needs for network support simply by participating” in social networks in the first place. Because they are already engaged in a support network, they do not feel an urgent need to expand their networks. Tangible assistance was unseen in our study. Braithwaite attributes the lack of tangible assistance to the “lack of physical proximity of the participants to one another.” Because participants may be on opposite ends of the globe, tangible assistance is difficult to offer unless participants are willing to make the effort to meet each other.
Humor ended up being the second most frequent type of message that we found in our evaluations. Braithwaite notes that humor can be used “in self-depreciating ways, in sarcastic ways, and as a way to diffuse tension and decrease discomfort in early stages of relationships.” Self-depreciating and sarcastic humor were certainly evident in the posts that we examined. As a response to Phil’s concerns about how much influence he should have in disciplining his stepdaughter, one poster described her experience in raising her own stepdaughter. In mentioning how she and her husband were stricter than her stepdaughter’s biological mother, the poster said, “Somehow SD [her stepdaughter] has survived the trauma of being raised by us!” By employing humor, posters were able to build bonds with the original poster.
The four dimensions of attraction to online social support, established by Walther and Boyd, tie in neatly with our analysis of the parenting support threads. The social distance factor means that the parents feel more inclined to ask for advice in these threads because they feel there must be someone with expertise in the vast number of people connected to these threads. At least one of them will have a solution to whatever problem presented. Indeed, parents were eager to find solutions to the original poster’s problems.
Furthermore, anonymity, both perceived and real, promotes participation from both people with questions and people with solutions to come forth. People with questions do not feel embarrassed to ask their questions because they are not likely to be judged by people that they actually know in real life. People supplying answers feel more secure knowing that even if their answers are wrong, there is virtually no way they have to own up to the responsibility.
Even though a lot of the messages are not carefully crafted to create specific impressions (many messages have typos in them), interaction management is still an important factor. The messages can be read at one’s leisure, and parents do not feel obligated to stick around and answer other people’s questions. In fact, a lot of the threads have a “hit-and-run” feeling to them; posts are often short and concise. Furthermore, anyone with a question can start a thread to get solutions, and that person may then return to dormancy (“lurking”) after he or she gets enough information.
Walther and Boyd’s final factor is access. For parents, it is important and comforting to know that no matter what time of the day they post their question, someone is likely to see it and answer it—someone is always awake on the Internet. Posts were usually answered in a timely manner; some responses to the original post came in as quickly as a few minutes.
Yang Zhang (purple)
Sue’s post
http://groups.google.com/group/misc.kids/browse_thread/thread/32665b5d88ffcc27/1551ef44eff22687#1551ef44eff22687
Funmum’s post
http://groups.google.com/group/misc.kids/browse_thread/thread/98630506c3319f13/46ef6101a323094a#46ef6101a323094a
Phil’s post
http://groups.google.com/group/alt.support.step-parents/browse_thread/thread/4dfff213af102230/a894452e07116c68?hl=en&lnk=st&q=child+raising#a894452e07116c68
For our assignment, Yang and I decided to examine posts that dealt with raising children. The first group we examined was misc.kids, where we found two threads to evaluate. The first thread we evaluated was originally posted by Sue, a mother trying to find ways to get her ten-year-old daughter to sleep in her own room. Sue was frustrated in constantly having to assuage her daughter’s fears. The second thread in misc.kids that we evaluated was by “funmum,” a mother who was bored of entertaining her two-and-a-half-year-old with the same activities every day; thus, she asked for advice as to new activities they could engage in. The last thread was in the group alt.support.step-parents. Phil, a father to thirty-month-old twins and a stepfather to a nine-and-a-half year old, was having issues with his wife as to how much of a role he should have in disciplining his stepdaughter, with whom he had lived since she was three years old. Phil wanted to know other people’s experiences in disciplining their stepchildren.
Unlike Braithwaite, the highest number of support messages we saw were informational support messages; they consisted of 100% of the posts. The other types of messages were much lower than in frequency than that of the informational support messages. Humor came in second in terms of frequency with 25% of the posts, and esteem support came in third with 20% of the posts. Emotional support, network support, and tangible assistance were not very prevalent in our study, appearing in 10%, 5%, and 0% of posts, respectively.
One reason that informational support messages dominated the threads may have been because of the online environment. According to Braithwaite, “information support…is most useful and prominent when the recipient can control the situation and put the information to use.” The original posters were all parents who could easily make use of the advice that other posters gave them. Indeed, the original posts were all focused on getting information and advice rather than any other support.
Even though emotional support was highly prevalent in Braithwaite’s study, we did not find many instances of it. Braithwaite asserts that “emotional support is more likely to be given when the recipient is experiencing distressful circumstances that are not subject to his or her control. This type of support can promote comfort and healing in such circumstances.” Contrary to Braithwaite’s observations of Support Network, Yang and I studied an environment where parents did have control over their situations; thus the need for emotional support was lessened. Indeed, parents did not seek “comfort” or “healing”; they were simply searching for possible solutions to their problems. Their posts concentrated on getting the information that they wanted rather than fulfilling emotional needs.
Esteem support was third in terms of message frequency, which matched Braithwaite’s analysis. Posters sometimes validated the original poster’s experiences. For instance, with Sue’s post about her daughter refusing to sleep alone, one poster identified with Sue’s frustration by saying, “I can see how you [Sue] would be over it!” With esteem support, posters aimed to lift the original posters’ spirits by validating their feelings and experiences.
Our findings for messages related to network support and tangible assistance were similar to those of Braithwaite. As Braithwaite mentioned, the low number of network support messages may be attributed to the fact that “members met their needs for network support simply by participating” in social networks in the first place. Because they are already engaged in a support network, they do not feel an urgent need to expand their networks. Tangible assistance was unseen in our study. Braithwaite attributes the lack of tangible assistance to the “lack of physical proximity of the participants to one another.” Because participants may be on opposite ends of the globe, tangible assistance is difficult to offer unless participants are willing to make the effort to meet each other.
Humor ended up being the second most frequent type of message that we found in our evaluations. Braithwaite notes that humor can be used “in self-depreciating ways, in sarcastic ways, and as a way to diffuse tension and decrease discomfort in early stages of relationships.” Self-depreciating and sarcastic humor were certainly evident in the posts that we examined. As a response to Phil’s concerns about how much influence he should have in disciplining his stepdaughter, one poster described her experience in raising her own stepdaughter. In mentioning how she and her husband were stricter than her stepdaughter’s biological mother, the poster said, “Somehow SD [her stepdaughter] has survived the trauma of being raised by us!” By employing humor, posters were able to build bonds with the original poster.
The four dimensions of attraction to online social support, established by Walther and Boyd, tie in neatly with our analysis of the parenting support threads. The social distance factor means that the parents feel more inclined to ask for advice in these threads because they feel there must be someone with expertise in the vast number of people connected to these threads. At least one of them will have a solution to whatever problem presented. Indeed, parents were eager to find solutions to the original poster’s problems.
Furthermore, anonymity, both perceived and real, promotes participation from both people with questions and people with solutions to come forth. People with questions do not feel embarrassed to ask their questions because they are not likely to be judged by people that they actually know in real life. People supplying answers feel more secure knowing that even if their answers are wrong, there is virtually no way they have to own up to the responsibility.
Even though a lot of the messages are not carefully crafted to create specific impressions (many messages have typos in them), interaction management is still an important factor. The messages can be read at one’s leisure, and parents do not feel obligated to stick around and answer other people’s questions. In fact, a lot of the threads have a “hit-and-run” feeling to them; posts are often short and concise. Furthermore, anyone with a question can start a thread to get solutions, and that person may then return to dormancy (“lurking”) after he or she gets enough information.
Walther and Boyd’s final factor is access. For parents, it is important and comforting to know that no matter what time of the day they post their question, someone is likely to see it and answer it—someone is always awake on the Internet. Posts were usually answered in a timely manner; some responses to the original post came in as quickly as a few minutes.
8.0: "Bipolar disorder is causing my divorce"
Hello Green bloggers (and non-Greens here for the first time). I have decided to do this blog alone, for I could not find a partner in time for the post.
I decided to analyze and code a forum/thread for Bipolar Disorder (since Braithwaite, Waldron & Finn focused on people with disabilities). Bipolar disorder can affect many people's lives in many different ways. This includes some of the things Braithwaite et al. emphasized: self-care, communication, and socialization.
I used the coding system by Braithwaite et al. and coded 20 reply posts on a thread from different people experiencing the disease or with the disease. I coded whether the replies contained information, tangible assistance, esteem support, network support, emotional support, and humor. I had a low inter-rater reliability of 0.46 and my table is pasted below (sorry about the size).
Information, esteem support, and emotional support were the most common form of support on the thread. The made up 70%, 65%, and 80% (respectively) of the responses. This data goes along with Braithwait et al. and their research. They said that emotional support was most common online, then information, and then esteem support. The data presented in the table supports their findings.The reason the data may be leaning towards these three types of support is because of the nature of the thread. It began with a wife/mother telling a story about how her husband with BPD is driving them to divorce. Many of the replies to her post came from individuals suffering the same thing. They gave her advice telling her what to do to help (information), they empathized with her situation (emotional support), and gave her compliments about how she is a strong woman for dealing with her husband (esteem). There was little to no humor (rightfully so) because the nature of this woman's post was not a laughing matter.
Some other posts on the thread were other people sharing their stories as well. These stories scored a zero in all the support catagories, for it was just a person reaching out for help or describing their situation, and not giving advice to others.
An interesting thing that I noticed while reading the thread was that there was one prominent poster named "Anna". She is a woman with BPD and posted many times to give advice to and support people struggling directly or indirectly with the disease. Many of the people in the thread looked to her for guidance and used her as their main source of support.
8. Assigning Blame with Online Support
This past weekend I was out of town and didn’t have the opportunity to work with a group here in 245, so I took the opportunity to teach my friend Juan (a Psychology major from Potsdam) about the Psychology of Social Computing.
We decided to look at a GoogleGroup for marriage support and focused on a thread titled “His and Her Stuff”, which was started by ‘Phil’ who was looking for advice about how to deal with his wife’s pack rat tendencies: http://groups.google.com/group/alt.support.marriage/browse_thread/thread/ca0761ec7c6e9545/f950b6957836dd4d#f950b6957836dd4d
Using the coding system established by Braithwaite and Waldron we coded 20 separate messages from unique authors on information, tangible assistance, esteem support, network support, emotional support, and humor. We had an inter-rater reliability of 0.96 and our findings are summarized in the table below.
We found that information was the most prevalent form of support found in this thread, with it being present in 100% of the messages that we analyzed. Esteem support was the second most common with 35% of the messages, followed by emotional support (15%) and humor (15%). None of the messages that we coded contained any tangible assistance or network support.
These results are somewhat in contrast to the findings of Braithwaite and Waldron’s research from 1999. In their study they found that emotional support was most prevalent online (40%), followed by information (31.3%), and then esteem support (18.6%). However, like the previous study tangible assistance and network system were not very prevalent in the online support messages.
One reason for this difference may simply be the content of the actual messages. Braithwaite and Waldron focused on support groups for individuals with disabilities, a highly personal and emotional issue. On the other hand, we looked at an issue in which someone would be considered right or wrong; which easily leant itself for individuals to weigh in and give their own input and advice, which would be coded as informational support. Additionally, Phil, the original poster of the thread, did not seem very distraught by the issue at hand and even concluded his post by saying he was “venting”, sending a message to those who responded that he was probably looking more for advice than emotional support.
One other interesting trend that we noticed while reading this thread was that there seemed to be somewhat of a “reverse esteem support”. Messages seemed to validate Phil’s wife’s position or validate her actions rather than Phil’s. This was also coded as advice and helped lead to the high information score, but we would argue that there is something more to it than simply information. However, it does not neatly fit into the category of “esteem support” because it is not Phil who is being validated. Furthermore, it does not fit into “emotional support” because again, the understanding being expressed is not for Phil, but his wife. While construed as advice, because the message had more charged language in the forms of validations it’s important to look and see if there may be another dimension to online support messages that is not present in Braithwaite and Waldron’s article.
8: Coding Social Support
Jacob Chase: Green
Emily Docktor: Green
Dana Klion: Yellow
Lauren Kramer: Blue
http://plainjanemom.com/2007/10/09/i-didnt-sign-up-for-this- particular-new-attitude/#comments
Our group analyzed a support group focused on the woes of a mother, Erika, who is fed up with her 3-year-old son’s foul attitude. Plain Jane Mom is an online community of mothers and fathers who share a great deal of common interests and experiences as parents. Users can submit default profile photos, and a short description of the user’s family often appears next to that user’s photo. User profiles may also include more detailed family descriptions and contact information. The attributes of openness and common ground seems to make Plain Jane Mom a very successful community for those who participate in it.
The first 20 messages in response to Erika’s original post were coded. The most frequent type of supportive behavior found on this particular support group was information support (coding frequency: 15). Information support was followed by humor (coding frequency: 13), then emotional support (coding frequency: 6), and then esteem support (coding frequency: 3). Our inter-coder scores for the supportive behaviors of tangible assistance and emotional support resulted in final frequencies of 0 once discrepancies were resolved. Our group agreed on a substantial majority of the presence (or lack there of) of each social support category. Inter-rater reliability was 67.5%.
Braithwaite found that emotional support messages were most frequently enacted, with 40% of the total number of support messages showing this characteristic. Informational support messages were second most frequent, being found in 31.7% of the messages Braithwaite examined. This is quite different from our statistics, which found informational support in 75% of the messages on Plain Jane Mom. Esteem support frequency came in third, as it was found evident in 18.6% of Braithwaite’s study. This is similar to our findings of 15% frequency of esteem support. In Braithwaite, network support had a frequency of 7.1% and tangible assistance accounted for 2.7% of messages. Our group research reiterates that these two categories are of lower frequencies.
In conclusion, our findings were rather similar to Braithwaite’s analyses, even though Braithwaite’s study was much more thorough. Braithwaite studied almost 1,500 messages, but we only looked at 20. His inter-rater reliability was 80%, whereas ours was a bit lower at 67.5%. This could be a result of our sample size, or perhaps due to a lack of formal code training.
Emily Docktor: Green
Dana Klion: Yellow
Lauren Kramer: Blue
http://plainjanemom.com/2007/10/09/i-didnt-sign-up-for-this- particular-new-attitude/#comments
Our group analyzed a support group focused on the woes of a mother, Erika, who is fed up with her 3-year-old son’s foul attitude. Plain Jane Mom is an online community of mothers and fathers who share a great deal of common interests and experiences as parents. Users can submit default profile photos, and a short description of the user’s family often appears next to that user’s photo. User profiles may also include more detailed family descriptions and contact information. The attributes of openness and common ground seems to make Plain Jane Mom a very successful community for those who participate in it.
The first 20 messages in response to Erika’s original post were coded. The most frequent type of supportive behavior found on this particular support group was information support (coding frequency: 15). Information support was followed by humor (coding frequency: 13), then emotional support (coding frequency: 6), and then esteem support (coding frequency: 3). Our inter-coder scores for the supportive behaviors of tangible assistance and emotional support resulted in final frequencies of 0 once discrepancies were resolved. Our group agreed on a substantial majority of the presence (or lack there of) of each social support category. Inter-rater reliability was 67.5%.
Braithwaite found that emotional support messages were most frequently enacted, with 40% of the total number of support messages showing this characteristic. Informational support messages were second most frequent, being found in 31.7% of the messages Braithwaite examined. This is quite different from our statistics, which found informational support in 75% of the messages on Plain Jane Mom. Esteem support frequency came in third, as it was found evident in 18.6% of Braithwaite’s study. This is similar to our findings of 15% frequency of esteem support. In Braithwaite, network support had a frequency of 7.1% and tangible assistance accounted for 2.7% of messages. Our group research reiterates that these two categories are of lower frequencies.
In conclusion, our findings were rather similar to Braithwaite’s analyses, even though Braithwaite’s study was much more thorough. Braithwaite studied almost 1,500 messages, but we only looked at 20. His inter-rater reliability was 80%, whereas ours was a bit lower at 67.5%. This could be a result of our sample size, or perhaps due to a lack of formal code training.
8 Narcoleptic Loneliness
Our group decided to investigate support communication within newsgroups for narcolepsy and loneliness. We observed two posts: one from alt.support.loneliness and another from alt.support.narcolepsy. From the two posts that were analyzed, we were able to successfully code a total of 20 support responses.We found 86% inter-rater reliability, which is statistically significant. The top three response characteristics we found were information (70%), humor (35%), and esteem support (25%). In contrast, Braithwaite’s article entitled “Communication of Social Support in Computer-Mediated Groups for People with Disabilities” found that emotional support was most prominent, followed by informational and finally esteem support. Both Braithwaite and ourselves found that network and tangible support were least commonly found.
A few reasons could have contributed to this discrepancy between our findings. Firstly, our sample size was only 20, whereas Braithwaite examined approximately 1472 responses. It is possible that if we were allotted a greater sample size, our findings may have been more similar. Also, we noticed that the two newsgroups that we analyzed featured a small number of subscribed users, suggestive of small communities of people that are familiar with each other. This dynamic would hijack many of the responses, as users would deviate from the original topic as they developed their own conversations within the thread – usually consisting of shared meaning, inside jokes, and community references.
Another reason that our results may have differed from Braithwaite, is that we analyzed different types of issues than Braithwaite. Whereas Braithwaite coded disability support groups, we analyzed narcolepsy and loneliness. The nature of the forum, as well as the level of concern of the original poster influences the type of responses. For example, the narcolepsy support group contained more emotional and information support than did the loneliness support group. This is due to the fact the original poster in the narcolepsy thread seemed more nervous and upset about his problem than the loneliness poster. Consequently, the narcolepsy support group elicited more emotional and informative support.
We found that our support groups obtained many responses from frequent users, while there were also a few sporadic users present within the group. This contradicts Wallace’s claim that less support is elicited with more people present. Another theorist, Walther, created the four dimensions of online support, namely, social distance, anonymity, interaction management, and access. Within this theoretical framework, social distance and anonymity are two particular support attributes that contributed to the level of self-disclosure present on the loneliness newsgroup. Similarly, in the narcolepsy newsgroup, people were more likely to discuss the specifics of their medical conditions with strangers, because of the perceived anonymity afforded them in CMC environments.
Thread 1
Thread 2
Dina Halajian
Robert Grue - Purple
Robert Rowland - Brown
Sunday, October 28, 2007
8. Spinal Cords and Social Support
The Braithwaite et. al paper entitled "Communication of Social Support in Computer-Mediated Groups for People With Disabilities," analyzes online support communities by coding the types of messages users leave. Messages are categorized as information, tangible assistance, esteem support, network support emotional support, and using a coding system.
We replicated this study (on a much smaller scale) by analyzing 20 messages in an online support community for those with Spinal Cord Injuries (SCI). We also included the additional cateogry of humor as a type of support. The chart below indicates our results after observing messages on three different threads. The agreement of our individual analyses (inter-related reliability) was 88.3% (88.33333%).
Our findings also supports Walther's theory regarding online community support. Walther defines four factors of online communities: social distance, anonymity, interaction management, and access. Within the online communities we researched, three of four of the factors were certainly present. It was clear that social distance, anonymity, and access were contributing factors in individual posts and replies. The amount of experts on the forums were substantial when compared to what these potentially immobile individuals would have in FtF reach. Secondly, the CMC communication gave the posters as much or as little anonymity as they desired. For example, one poster known only as “friend” had lots of great information but we knew nothing about him/her. Other users posted their phone numbers, email addresses, or even personal WebPages. Lastly the access to this online community was unlimited. It could be accessed any time during the day and as much as one so desired.
In conclusion we came to the agreement that these findings do not support Wallace’s predictions. Much support was offered to each one of these individuals despite there being a large volume of people online. Although going back to the subject matter, it is a touchy subject people get very upset about. If we were to examine a less touchy subject perhaps not so many individuals would offer such support.
We replicated this study (on a much smaller scale) by analyzing 20 messages in an online support community for those with Spinal Cord Injuries (SCI). We also included the additional cateogry of humor as a type of support. The chart below indicates our results after observing messages on three different threads. The agreement of our individual analyses (inter-related reliability) was 88.3% (88.33333%).
The three support categories that were most often found in the messages analyzed in the Braithwaite et. al study were emotional support, information, and esteem support. These three categories were also found most frequently in the responses we analyzed in the SCI online community. However, while Braithwaite et. al found the percentages to be 40, 31.3, and 18.6 respectively, our percentages for the three categories overall were much higher: 85, 50, 74 percent. Additionally, information was the lowest of the three categories in our personal study. Unlike the Braithwaite study, we found a much larger incidence of tangible assistance, as demonstrated when users offered their screen name or phone number the person in which they were replying so that they coudl have more direct contact.
One reason for these differense in the choice of subject matter. We focused on a distinct disability (spinal cord injuries) rather than disabilities in general, as noted in the study. The Braithwaite study states that "the less control you have over the disability, the more emotional support is given." This aspect greatly applies to spinal cord injuries because patients do not have any control over the accident, and once an injury is sustained, there are very limited physiological fixes. (Compare this to something such as depression, which can be fixed by different types of medication, counseling, etc.). Also, a higher percentage of tangible assistance may have been found because many of the users wheelchair bound and may choose to have more contact with other persons in similar situations.
One reason for these differense in the choice of subject matter. We focused on a distinct disability (spinal cord injuries) rather than disabilities in general, as noted in the study. The Braithwaite study states that "the less control you have over the disability, the more emotional support is given." This aspect greatly applies to spinal cord injuries because patients do not have any control over the accident, and once an injury is sustained, there are very limited physiological fixes. (Compare this to something such as depression, which can be fixed by different types of medication, counseling, etc.). Also, a higher percentage of tangible assistance may have been found because many of the users wheelchair bound and may choose to have more contact with other persons in similar situations.
Also, we only analyzed 20 messages, whereas the Braithwaite study analyzed 1472. Since we didn't use the same number of messages, we don't know if we would have had the same results.
Furthermore, 16 of the messages analyzed were a part of the same thread on a single message board. The original post was "I would like to be able to talk to people with similar situations on a on-line chat room if possible." The variety of messages may have been limited because they were all related to the very same request.
Furthermore, 16 of the messages analyzed were a part of the same thread on a single message board. The original post was "I would like to be able to talk to people with similar situations on a on-line chat room if possible." The variety of messages may have been limited because they were all related to the very same request.
Our findings also supports Walther's theory regarding online community support. Walther defines four factors of online communities: social distance, anonymity, interaction management, and access. Within the online communities we researched, three of four of the factors were certainly present. It was clear that social distance, anonymity, and access were contributing factors in individual posts and replies. The amount of experts on the forums were substantial when compared to what these potentially immobile individuals would have in FtF reach. Secondly, the CMC communication gave the posters as much or as little anonymity as they desired. For example, one poster known only as “friend” had lots of great information but we knew nothing about him/her. Other users posted their phone numbers, email addresses, or even personal WebPages. Lastly the access to this online community was unlimited. It could be accessed any time during the day and as much as one so desired.
In conclusion we came to the agreement that these findings do not support Wallace’s predictions. Much support was offered to each one of these individuals despite there being a large volume of people online. Although going back to the subject matter, it is a touchy subject people get very upset about. If we were to examine a less touchy subject perhaps not so many individuals would offer such support.
Subscribe to:
Posts (Atom)