Our results were similar to the study in the Braithwaite article in that our interrater reliability was 78.3%, and the article's was 80%. Our data regarding % of total messages per category differed, but since our interrater reliability is so similar, we can speculate that the differences are based on factors such as the type of support group, rather than our own faulty coding. The actual differences between our results and the article's are that 1) our % of total messages per category are much higher in all categories, and 2) our results showed that the highest % of total messages was the "information" category, while the article's results showed that the highest % of total messages was the "emotional support" category.
There are a lot of possible explanations for these differences. Our results might have had higher total %'s overall because unlike the support group in the article which had the broad topic of disabilities, the support group that we analyzed had a very specific and focused topic. The members of the group felt a very close connection to one another since they had such similar experiences on the unit, and therefore, offered more support in their responses. They might have also felt even closer to one another than average support groups because they disclosed their full names, and some one them knew each other from their stay on the unit. Another reason for the high %'s overall in our results could be that eating disorders are generally female dominated, and according to Wallace: "The situation in which women may be more likely to help on the Internet are...in the support groups in which people are sharing personal problems." (p.200) The female dominated group will be more willing to offer specific and personal support. All 20 of the messages that we coded were from females, and there was not a single post from a male. This also adds to the common ground of the support group; the members were related through their eating disorder, the fact that they had been on the Princeton unit, and that they were all female. This also adds to the argument by Braithwaite, Waldron, & Finn that the members of the support group are "all in the same boat," and therefore, would be more willing to share personal information and support each other.
A reason to explain why our results showed that the most % of total messages was for "information" as opposed to "emotional support" (2nd highest) category, which was the reverse result of the Braithwaite experiment, is that according to the article's model, "information support is most useful and prominent when the recipient can control the situation and put the information to use." (p.142) The majority of the members of the support group in the article all suffered from disabilities that they couldn't control, but eating disorders can be controlled to a certain degree, especially with support.
Humor, which was not an official category but was very prominent in the results from the article, was only present in 5% of the interactions that we coded. A possibility explaining why it was so scarce is that eating disorders are a serious issue, and the females posting the supportive messages were truly trying to get through to their recipients, and were trying to be serious. This is because the illness is psychological, and if it is taken lightly, those with the disorder won't believe that it is fatal.
No comments:
Post a Comment