What do research methods do? Research methods are routinely understood as objective techniques for getting to know the world. Yet they may be more influential and socially significant than this, particularly as more digital methods are being developed and deployed. So what, too, do digital methods do? And why is this important for researching digital media and learning?

In the field of digital media and learning, we use an array of social scientific methods in order to get to know as much as we can about learning situated in contexts that are increasingly understood to be networked and technologically mediated. Putting it very simply, for psychologists in the field, methods are techniques for getting to know the inner lives of learners. For sociologists, methods help to illuminate the social structures within which education takes place and learners' identities are formed. For learning scientists, methods are resources for explaining the immediate technical and social processes of learning.

However, underpinning the technicality of methods is the assumption that they are able to capture and represent the world just as it is. Methods are understood rather like a photographic device that can capture, freeze-frame and reproduce a facsimile of reality. As researchers, we can say we've done a good job if our methods have been up to the job of capturing a picture of an objective reality as it really is—or at least pretty accurately so.

Neuroscience is a good example. Thanks to brain scanning devices researchers are now able to produce images of the brain which appear to show actual sensory and motor functions occurring at the neural, molecular and cellular level. But much the same can be said of anthropological ethnographers returning from fieldwork. Their fieldnotes, photographs, dictaphone recordings, transcripts and video data are much like the neuroscientist's CAT and PET scans. They represent a reality—a human brain, a culture, whatever—that has been recorded and made presentable enough for interpretation.

But are research methods really so objective? Or do they do other things?

Methods are Social Too!

An important argument needs to be made here. It is that our research methods may in some important ways fabricate the very things we want to observe.

Mundane instruments like questionnaires, observation schedules, coding schemes and video data of course frame the ways in which data can be collected and later analyzed. They are empirical constructions of reality. More importantly, our disciplinary perspectives, whether in various schools of psychology, sociology, anthropology, computer science, or interdisciplinary blends such as the learning sciences, arm us with particular theories and methods for observing, recording and understanding things. Methods are not neutral or innocent tools but necessarily construct, shape, configure, frame and make up the social worlds they study—methods help create society.

This is why I've become so interested in recent work on the “social life of methods,” a program of research studying the historical development of social scientific research methods. The research examines, for example, early forms of social scientific research in 18th and 19th century map-making and census-taking, as well as popular contemporary methods such as sample surveys and focus groups, and the emergence of new digital methods in the 21st century. The implications are more significant than merely historically fascinating. The researchers make two central points: first, that methods are shaped by the social world, and second, that methods help shape the social world.

To take the first point, methods are social because they are shaped by the social world in which they are located—they are shaped by the social and political circumstances in which they have been produced. Methods are designed, obviously enough, for particular purposes, and they come into being through the work of sponsors and advocates. From this perspective, for example, making maps can be understood as a way in which governments could understand and manage their territories. Likewise, the creation of censuses enabled countries to become aggregated statistically as populations which could consequently be calculated and acted upon. More recently, the history of focus groups shows how this technique has been largely adopted for marketing research purposes in the commercial sector. Such details demonstrate the importance of recognizing the social life of methods. These are not neutral tools but politically charged instruments.

Methods are also social, however, because they in turn help to shape that social world—or, as it's put in the social life of methods program, methodologically speaking “what you see is what you get.” The point being made here is that all research methods come with ready-made assumptions about the character of the social world that they are studying. This means that the discoveries made by social scientific research are not merely objective facsimiles or representations of an existing world, but play a part in making it up.

The focus group, to continue an example, comes preloaded with the assumption that people have attitudes and opinions, that they are able to make choices, that these choices will be rationally made, and, furthermore, that the attitudes and choices of individuals may be aggregated into statistical collectivities and populations. In large part, such assumptions about the characteristics of people are what have made focus groups so attractive to commercial market researchers. We've learned to become people with attitudes and opinions that marketers can elicit from us and aggregate into target groups.

Making Digital Data

Important questions are raised for research in digital media and learning by these insights. Newer forms of digital methods are now being developed and deployed that will enable researchers to make data on learning in new kinds of ways. Technological methods like track-and-trace technologies, digital mapping, software visualization, transactional data, data mining, social network analysis, digital databases, brain scanning, wikis, web2.0 and open source social analytics are all beginning to change the ways in which learning can be tracked, recorded, visualized, patterned, documented and presented. Researchers of learning, armed with these instruments, will be able to produce descriptions and images of learning and learners that really didn't exist before. Such insights make it possible to intervene in education in new ways. Methods, to reiterate, are shaped by the social circumstances in which they are created, but they also help to create and circulate new social worlds.

Is this a big deal? If methods allow us to know more, then doesn't that mean we can intervene more effectively to improve learning? Isn't making new social worlds an admirable aim? Maybe so. But let's remember that methods are also socially produced for particular purposes by particular sponsors with particular interests; they come politically preloaded. Many of the new digital research methods and techniques come from commercial organizations rather than from innovation in social science. Their purposes and their sponsors and advocates lie outside of social science, and outside of the field of learning, and are concentrated in the internet industries. Amazon, Google, Facebook and all the rest are continually generating data as by-products of many millions of human interactions and transactions. In an important way, such companies are now doing social science research. At present, they may be the future of social science itself.

Transactional Politics

Perhaps the key point to be made about many such digital methods is that they generate transactional data without the awareness or intervention of research subjects—we are being aggregated as research data based on our transactions online without even thinking about it. That is to say that whereas social science research makes people the center of its accounts, digital methods record data transactions between parties, many of which aren't even people at all. If focus groups made us into people with attitudes then Amazon recommendations and the like have made us into “transactional actors” within dynamic flows of data exchange and interaction processes.

There are political and ethical consequences of doing research based on people's web transactions, data trails and on what they do online rather than on what they actually say. “Transactional politics” describes how new kinds of experts and gatekeepers are now involved in the collection and analysis of transactional data—a form of politics which involves commercial companies but increasingly, thanks to web 2.0 devices, allows anyone to access and analyze transactional data for themselves. So in some ways these devices have the potential to transform social science research—to move the object of study from the human and the social to the transactional.

These are perhaps extreme examples but the important point to recognize is that innovation in research methods is now happening in the space of digital media development and the generation of transactional data perhaps more than in social scientific disciplines. If we take it that learning is happening in such transactional, networked spaces, too, then we need to attend to the ways in which such digitally networked learning is being inseparably shaped through these new transactional methods. And if we actively mobilize such devices and methods in our own repertoires of research techniques then we need to recognize how we may be taking part in a process of shaping the future of learning that is genealogically rooted in commercial interests and non-humanist transactional politics.

Prototype Futures

The implications for the field of digital media and learning may be especially acute because its concerns are routinely with the place of technology in building the future and shaping society, with what could be, or indeed should be, if things were “otherwise.” Digital media and learning research traces learning processes as they occur in new digital and networked spaces where they are inseparable from transactional data. What we are doing here, then, is inherently political. We are making normalizing assumptions about the kinds of social worlds, the kinds of learning, and the kinds of learners we want to help make more real. We are participating in the prototyping of society itself.

Yet one risk, as we have seen, is that the rise of digital methods has begun to emphasize transactional data over human participation in research. Deploying transactional methods in the study of digital media and learning risks shaping a society in which social relationships are reduced to non-humanist networks of transactions and data exchanges.

Banner image credit: Gilderic Photography http://www.flickr.com/photos/gilderic/4749214748/