A nice article on some of our current efforts came out today in Oregon Sea Grant’s publication, Confluence. You can read the story on-line at http://seagrant.oregonstate.edu/confluence/1-3/free-choice-learning.
One of the hardest things to try to describe to Nathan Gilles who wrote the article (and to the folks who reviewed the draft) is the idea that in order for the lab to be useful to the widest variety of learning sciences researchers, the cyber-technologies on which the museum lab are based have to be useful to researchers coming from a wide range of theoretical traditions. In the original interview, I used the term “theory agnostic” in trying to talk about the data collection tools and the behind-the-scenes database. The idea is that the tools stand alone independent of any given learning theory or framework.
Of course, for anyone who has spent time thinking about it, this is a highly problematic idea. Across the social sciences we recognize that our decisions about what data to collect, how to represent it, and even how we go about collecting it are intimately interwoven with our theoretical claims and commitments. In the same way that our language and symbol systems shape our thinking by streamlining our perceptions of the world (see John Lucy’s work at the University of Chicago for the most cogent explanations of these relationships), our theories about learning, about development, about human interaction and identity shape our research questions, our tools for data collection and the kinds of things we even count as data.
Recognizing this, we struggled early on to develop a way to automate data collection that would serve the needs of multiple researchers coming from multiple frameworks and with interests that might or might not align with our own. For example, we needed to develop a data collection and storage framework that would allow a researcher like John Falk to explore visitor motivation and identity as features of individuals while at the same time allowing a researcher like Sigrid Norris to document visitor motivation and identity as emergent properties of mediated discourse: two very different notions of identity and of best ways to collect data about it being served by one lab and database.
The framework we settled on for conceiving of what kind of data we need to collect for all these researchers from different backgrounds is focused on human action (spoken and non-spoken) and shaped by a mediated action approach to understanding human action. Mediated action as an approach basically foregrounds agents acting in the world through the mediation of cognitive and communicative tools. Furthermore, it recognizes that such mediated action always occurs in concrete contexts. While it is true that mediated action approaches are most often associated with sociocultural theories of learning and Cultural Historical Activity Theory in particular, a mediated action approach itself does not make strong theoretical claims about learning. A mediated action framework means we are constantly striving to collect data on individual agents using physical, communicative, and cognitive tools in concrete contexts often with other agents. In storing and parsing data, we strive to maintain the unity of agent, tools, and context. To what extent this strategy turns out to be theory agnostic or learning theory neutral remains to be seen.