Lindsey’s Notes on This Week’s Reading

General tip: As you read your assigned article, I strongly suggest you make an outline of it–try to understand how it is put together. This practice is one that will eventually help with both speed and comprehension. While scholarly articles are often written according to discipline-, language-, or even university-specific templates or guidelines, the structure of academic writing is generally somewhat standardized. Learning the patterns makes it easier to get to the content. 

Noortje Marres

The Redistribution of Methods: On Intervention in Digital Social Research, Broadly Conceived

Marres’ point of departure for this paper is that when research on people, cultures, and communities is conducted in a digital space, it is a “redistributed” form of research that relies on multiple tools, people, and habits in order to be successful.  Digital social research, then, is internally diversified in a way that could productively differ from offline methods.

The introduction to this paper argues that responses to technology are polarized, both in general and in research communities specifically.  People tend to be either highly optimistic about the impact of technology, or highly pessimistic.  This is also true of people who conduct social research. There is a general sense that new technology has changed the field, but that agreed-upon observation makes some people excited and some people anxious.

(In this, Marres’ observation of her own field is a lot like what we saw in the articles we read before our first class meeting—there’s a similar excitement-anxiety continuum among humanists, and we’ve even experienced some of that in our own class thus far.)

Marres wants to “unsettle this opposition,” to try and think beyond the optimistic and the pessimistic viewpoints and think instead about the variety of spaces and people involved in digital social research, and what implications that might have for how we study groups and communities. She also wants to apply this effort specifically to social science methodologies—to argue that new technologies change who provides what quantitative data, as well as how it is provided.  To do this, she presents the history of sociological debate over the ways in which technology has shaped the field, then looks at two case studies of web tools that have taken offline social science practices and “translated” them for online use.

One of the game-changing technology developments Marres first considers is analytics—the ability to see who is looking at what and where, online. (Echoes of the NSA, anybody? Though this article was published prior to those revelations.) While social theorists have had a difficult time explaining what access to these analytics actually means for society, it is clear that they have great weight—not the least of which is serving the profit motive of capitalists. But Marres’ point about analytics is that if we focus too much on the kind of data we can gather about digital habits, we risk not seeing who is doing that gathering, and we might miss out on how those different agents relate to one another.

This is where the idea of “redistribution” comes in—the point being that it “highlight[s] processes of exchange” rather than overemphasizing precisely what data is changing hands. The parties involved should be as important to us as the content they are sharing. Marres wonders just how much the changing digital research field has allowed for the renegotiation of traditional research relationships—if, for example, the line between the subject of a study and the person conducting a study could be blurred productively as a result of digital intervention.

Marres has four core avenues for this “redistribution” of research methods:

1. Examining internal connections rather than external ones
2. Focusing on who does what labor in digital social research
3. Questioning the line between what is studied and the context in which it is being studied
4. Rethinking the role of a social scientist under these new conditions

When it comes to thinking about how the methods used to study social groups themselves might change due to the growth of digital spaces and tools, Marres again sees a spectrum of thought. The most conservative thinkers see very little change in how to conduct research in the technological age—basically arguing that our new technologies are based on traditional methods in any case. On the other end of the continuum, some scholars argue that the increase in data available in digital research means that new models of social analysis not only can be developed, but must be. While Marres seems in general to trend towards agreement with this end, she feels like this perspective privileges math over people, and emphasizes the objectivity of numbers at the expense of qualitative data.

The other changes she summarizes here are internal debates about qualitative digital data—one style emphasizing the role of individual subjects in digital social research, and the other, the Digital Methods Initiative, attempting to make the life of various research technologies (think here about how all the tools we use online change over time, offering new features and retiring old ones) transparent to everyone involved.

After this lengthy look at the state of the debate, Marres covers two tools in detail: Issue Crawler (check out the Google Scraper!) and co-word analysis. Here, I think the article gets more accessible;. When it comes to Issue Crawler, Marres considers how search engines and other sorting tools decide what results are most popular—including, most notably, the number of other pages that link to a particular page. This focus on links between sources of content, she argues, has a parallel in earlier offline questions about source citation, in that it can replicate the popularity effect (basically, the more popular a site is in a given network of links, the more popular it will continue to become).  And so focusing solely on links in this way means that a whole host of research questions will be omitted or obscured—the tool cannot ask as many questions of its data as she might prefer.

Her second case study looks at co-word analysis—which again, as an offline tool, was an attempt to measure the strength of certain ideas or keywords, but to do so in a way that wouldn’t simply reinforce the popularity effect. Instead of looking at what came up the most often, researchers instead looked at how many variations of a single keyword or phrase were mentioned, at how many different times that keyword might be connected to other, different content.  While the tool being created to replicate this online seems to be in development at the time this article was mentioned, Marres discusses the process of trying to create that tool, and whether or not it will be possible to create a tool that can effectively do this kind of research across different digital spaces.

Overall, I think the most important takeaway from Marres’ article is the question of whether or not the distribution of research sites and users and subjects really is going to change the traditional divisions of labor in sociological research. There’s a lot here to think about with regards to participatory research methods. Check out the Wikipedia entry on participatory action research, or the Graduate Center-based Public Science Project, for more on the kind of research I think Marres is hoping to see happen online.

Nicholas Hookway

‘Entering the Blogosphere’: Some Strategies for Using Blogs in Social Research

Hookway begins this article by summarizing the primary trends in social science research that has been conducted on digital cultures. This article is from 2008, and the language used throughout reflects a distance between the time period in which this was written and the time we live in today—even the use of the word “blogosphere” (defining a community or group by their status as bloggers) feels antiquated? I’ll be curious to hear what value this article still has in 2014, in your eyes. In any case, at the time this article was written, Hookway saw most social science Internet research as being (overly) focused on questions of community development and online identity production.  And while his peers were beginning to consider some of the more sophisticated questions of content and context that are discussed by Marres later on, he asserts here that social scientists have yet to consider the full research potential of blogs.

His definition of a blog is very time-specific, claiming that the term “refers to a website which contains a series of frequently updated, reverse chonologically ordered posts on a common web page, usually written by a single author.” While I understand the benefit of choosing such a clear definition, I think this considers blogs as a product rather than a tool—and I’ll probably bring up that distinction in class, as it’s worth thinking about further. In any case, Hookway sees blog analysis as a means of collecting qualitative research material that could not easily be collected in offline environments. Researchers can survey more blogs than they could people, and they don’t have to bother with any of that nasty transcription business. Blogs are relatively anonymous, relatively global, and archived—three qualities which will make them a consistent source of research data.

Hookway’s first focal point after his introduction compares blog research to offline diary research—an established strategy in history and anthropology. He explains his own current research (into the ways in which Australian bloggers blog about everyday morality) as a means of situating this idea. Here, Hookway primarily touts “access” as a central reason for shifting to online analysis. He notes at the end of this section, however, that blogs are written for an audience—at least an implied one, if not a specific one.

In the second core section of the article, Hookway considers how to validate data collected from blog research—that is, how to make sure it meets the standards of his discipline when it comes to data collection and processing. One of his primary concerns is that blogging is performative—it’s a conscious act of self-presentation. But he feels that “the online context,” as he calls it, is relatively anonymous—and that people feel safe to be more honest when they blog than they might be in person. (Given how much blogs have been monetized since 2008, I’m not totally sure I agree with this idea, but I’ll be curious to hear what y’all think.) He also considers the ease with which a blogger may be “deceptive” (I prefer to say,  “playful”) when it comes to how they present themselves.

But Hookway isn’t sure that it matters so much whether or not bloggers are “truthful”. If you’re researching blogs to compare them to a sense of offline “real life,” then you’re setting up a skewed experiment to begin with. Like Marres, Hookway argues for a more nuanced approach to context—blogs should be analyzed within the “blogosphere” more than in comparison to offline interaction. Besides, as he points out, how do you know people are being truthful when you ask them to participate offline—to fill out a survey or answer your research questions?

In his final area of interest, Hookway discusses ways in which he thinks blogs might become a more effective source of research material for more social scientists. He emphasizes the centralized nature of blogging platforms, some of which may be searchable, and the aggregating feature of RSS readers and subscription services, which can collect content on the basis of topic or from pre-selected sources. Finally, he describes the details of his own blogging research and the features of the blogs he studied as part of his own project. Finally, he touches on ethical questions of permissions and access to content, on what is “private”, and copyright law. I feel like this should have been a bigger part of the article–do you agree?

As I reread Hookway’s article I’m struck by how out-of-date some of this feels, and how tentative some of it is. I hope that the student assigned this article will ask questions of our guest speakers, and compare how they conduct blog research with Hookway’s own process.

Adam Edwards, William Housley, Matthew Williams, Luke Sloan, and Malcolm Williams

Digital Social Research, Social Media and the Sociological Imagination: Surrogacy, Augmentation, and Re-Orientation

This article considers the growth of interactive web spaces and portable devices and the impact that both will likely have on research in the social sciences. Edwards et al survey the work of other digital social scientists and categorize their thoughts on technology and sociological research into three categories, each of which builds upon those which precede it: surrogacy –> augmentation –> re-orientation.

First, some social scientists see digital tools as being analogous to offline tools for data collection and analysis, such as surveys and interviews. Edwards et al suggest that researchers who use digital research tools as they would use traditional research methods are seeing the digital tools as “surrogates”—in this case, simply standing in for offline methods, and not really doing anything new. They look at Twitter analyses for this segment, questioning whether or not this is an appropriate or useful use of technologies, noting that this replicates traditional distances between the subject being studied and the scientist doing the studying in almost a colonial pattern. In some respects this sort of digital social research is akin to the explorer studying the native population, with all of the problems of that sort of observation.

In contrast to this model, Edwards et al focus on ways in which digital research enhances, or “augments,” traditional research methods. In this scenario, digital social research has a special role to play: it can collect data on entire populations at the moment that data is generated, rather than through time-bound traditional channels. For example, a real-time analysis of Tweets about hair care products is in some ways a better snapshot of current population-wide trends in hair care than is a survey, which asks people hair care related questions. One of them is immediate, the other one is mediated by time and distance. (The hair care example is mine, not Edwards et al, and it’s because I’m thinking about washing my hair as I write this post.)

The reason that Edwards et al emphasize the research value of the “augmentation” model is because they see it as being a more honest extension of the field. They go all the way back to C. Wright Mills’ 1959 formulation of the “sociological imagination” to provide support for this methodology’s correctness. Essentially, they say, augmented digital social research is truest to the origins of sociology, and is closest to the ideals of the discipline… which is what they consider more deeply in their third and final level of analysis:

Finally, another group of social scientists are arguing that beyond the new kinds of data that can be collected, new technologies generate new sites of study and new kinds of questions. This group, Edwards et al argue, sees the influence of technology upon social science research as a change in ideological direction, a “re-orientation” of the attention of social scientists towards new lines of inquiry. This is initially an improvement upon the limits of the surrogacy model, because it allows for access to groups that have otherwise been underrepresented in sociological research.  As an extension of the augmentation model, or the natural outgrowth of its ideology, “re-orientation” holds the most promise for the future of the field.