Building the “new data science of learning” – #eli2014 reflections

I’ve heard this phrase over and over in the last few months…”we are building the new data science of learning.” It sounds exciting, doesn’t it? The possibility of exploring emerging digital learning environments, with data points at scales that make education researchers’ eyes light up, and using all of this to inform and account for this thing we call “learning”…well, it’s very appealing.

Here’s my concern: The enthusiasm for research that uses large data sets to offer predictive and explanatory power has moved us further and further away from research models aimed at understanding the human condition and, thus, the learner herself. And, uncover what we may via algorithms and regressions, learning is ultimately still deeply a relational, emotional, and contextual human experience.

This concern is what drove Dr. George Veletsianos and me to embark on a mission to elevate research methodologies that target the human experience in this “new data science of learning.” You will see in our ELI 2014 presentation slides below, we’re both frustrated (slide 2) and worried (slide 3) these research methodologies—that generally fall under a qualitative research umbrella—are not being included in conversations or designs in the latest push for online learning research.

Now, I do not intend to re-incite the vitriolic discourse of the paradigm wars of the 1970s and 1980s (for a good examination of these wars, check out Denzin & Lincoln’s The Landscape of Qualitative Research). Instead, I want to encourage discussions that start from a place of caring and empathy—for the learner, and for each other. In this first post (first of several, I’m sure), I want to wrap up with three challenges for building the “new data science of learning:”

1. Evolve ethnography in digital learning environments – In the 1960s and 1970s, ethnographer Shirley Brice Heath changed the way educators viewed children’s language development through her deeply embedded observations of children in the towns of “Roadville” and “Trackton” (described in the must-read Ways with Words). Ethnography has a rich history in education and social science research and we would do well to continue its evolution within and outside of digital learning environments. Michael Wesch has given us great examples of digital ethnography on his blog Mediated Cultures; Laura Pasquini recently shared some thoughts on the topic as well. I haven’t heard of anyone doing ethnographic studies in large-scale digital learning environments but, if you have, please share.

2. Don’t ignore politics and power in learning research – It’s too easy for studies that fish in the data exhaust (Justin Reich’s term) of digital learning environments to assume that politics and power do not play a role in that research. Qualitative research paradigms can help inform our understanding of how power, privilege, and social identities shape research from start (in research questions, assumptions, methodologies) to finish (in analyses, conclusions). In our ELI 2014 presentation, George and I talked about not trading empathy for efficiency in research. This sentiment applies here.

3. Get excited about praxis – One of the hallmarks of qualitative research (and one of my favorite things about it) is its inextricability from action. Here at Stanford, we talk a lot about the importance of doing research in Pasteur’s quadrant, or research that has both scientific and societal benefits. Yes! That’s exciting. We can also learn from action research paradigms to collaborate with research participants to enact or sustain the benefits they would like to see. I like this approach because it does not assume that we know what’s best for participants.

As I said before, I expect to be writing more about this topic. In the meantime, I welcome your thoughts and ideas on building the “new data science of learning.”

Side note: My primary methodological training is in narrative inquiry and analysis. From that lens, I would be excited to do a research study on learners’ personal journeys through MOOCs, as told via their stories. George V. did a nice study along these lines with his students (more here). I would love to hear from others what they are interested in studying.

Photo Credit: 7D-Kenny via Compfight. Permissions granted via CC-BY-NC

5 Responses

  1. Thanks so much for sharing this. I’m currently starting a mixed methods study of learning to learn in MOOCs for my PhD project and am really struggling to convince some of those I’m working with that such a study can’t simply be done by “data mining”. They were baffled when I mentioned using ethnographic methods, as they couldn’t figure out how I’d “go and live with” the MOOC participants I’m researching. What’s even more interesting is that at the same time I’m having to defend my use of quantitative data to my department, which is so set in the qualitative ‘paradigm’ that any research mentioning ‘learning science’ or ‘evidence’ is inherently distrusted and looked askance at.

    1. Hi Ashley, thanks for your comment. I’m sorry to hear about the paradigm struggles you are facing for your PhD project. My department had similar struggles; I remember being told that the only way qualitative “worked” was if it was paired with quantitative methods. I was lucky that my dissertation chair was a qualitative researcher, but I had to go outside of my department for the rest of my committee, which actually turned out to be a really good learning experience. Best of luck with your project! Cheers!

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Back to Top