Kvansy and Igwe

I thought the article on the African-American blog community response to the ABC News report on AIDS in Black America was really interesting. Overall, I thought the article was excellent, but right now I want to talk about a few things that confused and troubled me.

One thing that I wanted to learn more about in this article was the methodology. While the authors explain that they coded the blog, they provide a confused indication of what served as their unit of analysis. They explain that they coded a number of comments, but also explained that they analyzed direct quotatoins, and threads of conversation. I'm wondering how they segmented the data into chunks, how they determined what the parameters of those chunks were. Also, the authors explain rather vaguely that "quotations were coded and analyzed according to themes that they represented" (577). How were these themes chosen/deciphered? How did the authors negotiate differences/similarities amongst the "candidate codes" that they came up with when they coded separately? What was their interrater reliability?

More importantly, it seems like the authors were expressly looking for signs of "resistance" in these blogs. They say that one of their research questions was to find if and how the "community provided an oppositional interpretation" of the ABC report (576). It seems, that if you're looking for these oppositional interpretations of the news report, you'll be likely to find and code for them. Thus, it's not a surprise that the discussion portion of the article emphasizes these signs of resistance, while perhaps understating many of the other categories that emerged from this analysis.

I'm not saying that the authors were merely finding what they were looking for. But I am saying that their description of the methodology makes this an important question to ask. One problem I see is that the authors don't indicate the percentage of comments or quotations or threads (or whatever unit of analysis they were coding) that fell under each coding category. In their discussion, they make it appear that bloggers' comments resisting ABC news were overwhlemingly present in the data: "They [the bloggers] questioned the statistics, provided explanations for why the figures overstate the proportion of HIV infections among African Americans compared to other racial/ethnic groups, and critically analyzed how science is often misued to legitimize negative portrayals of Black people....The agency to resist these ascribed identities is situated in and often in opposition to the institutional power structure of existing AIDS discourse" (588). However, the authors provide no way of knowing if these oppositional interpretations of the ABC report were the most salient feature of the blog that was analyzed. They don't compare the number of such comments to the number of comments that support and uphold the news report. For instance, how many comments gave "props" to ABC for giving the report and shining light on this issue? How many comments neither questioned nor challenged the findings of the report? There's also no way of knowing if the "resistance" category was more or less salient than the "ineffective leadership" or "Black cultural practices" or "individual behavior" categories.

Finally, let me take small issue with one potential category that the authors did not code for, but which they mention in passing. The authors notice "clear linguistic markers such as "they" and "us", ABC and BET to demarcate the outgroup and the ingroup, which serves as additional evidence of a shared group identity" (582). I realize that this was not one of the researchers primary questions, but it would have been interesting if they had coded for demarcations of an "us/them" binary. It would have been interesting to try to understand where and when these demarcations occur. And I think it's important to ask if, by making such demarcations, this blog community sustains a kind of unnecessary separation b/w black and white, us and them, BET and ABC. To me the creation of this demarcation is double-edged. On the one hand, it serves to promote a kind of shared group identity, as the authors note. On the other hand, it serves to promote a kind of Otherizing of White people, White news sources, and White culture. This is certainly understandable given the history and persistence of racism in our country; moreover, it is understandable given the real bias in the White-controlled news reports (see Teun van Dijk's Elite Discourse and Racism). Still, part of me cringes when I see us/them lingusitic markers, and I think researchers have a responsibility to expose these binaries and comment on thier potentially divisive implications.