Lately I’ve been thinking about data and stories, design and luck, prediction and intuition, faith, empathy, beliefs and truth, scale and thinking small, political and personal, power, control, freedom and social justice. And how all of these ideas crashing into one another inside my head. Conflicting and challenging one another. Creating turmoil, unease and discomfort. Uncertainty. A traffic jam of thoughts and ideas.
And into that confusion, I’ve inserted a deadline and personal challenge. (Thank goodness for deadlines). As a organizer of the Learning Analytics and Knowledge Hackathon, I suggested that we dedicate some of the time to a discussing some of discussing these ideas. The other organizers agreed and said, “OK, let us know what you come up with.” As of today’s meeting, I had nothing. Except now some newfound urgency.
So here it goes.
Data-driven vs. person-driven stories
Never so much in the past year have I felt as uncomfortable talking about using data among forward-thinking learning folks. In a year when data-driven algorithms influenced the news (propaganda) folks read, the polls got it wrong on critical questions, and both humans and robots spewing hate online became seemingly unstoppable, I don’t blame them. At the same time, we have have some recent examples of person-driven narratives whose goal to mislead (or gaslight – More notes on gaslighting by Tressie McMillan Cottom).
In the same year I’ve watched my phone faithfully tracked the steps I take and where I drive. Those data points, are simply micro-facts. If I used a wearable device, you could improve the accuracy of those facts by identifying all the times I forget to carry my phone (another micro-fact). If you gathered up enough micro-facts, you could create a story of me. What that story would say would of course depend on the micro-facts you chose to include and how you chose to interpret them but it would likely emphasize my day-to-day routines.
Still in 2016, I’ve started sharing my thoughts and stories on several blogs and gathering the narratives of others on whenIneededhelp.com. Some of these stories are gut-wrenching accounts of injustice and marginalization. In these stories, myself and others tend to filter the micro-facts and focus on a relatively small number of life events that have had a larger than normal impact on our lives, macro-facts. In fact, I recently noted that I only tend to write on my personal blog when things are falling apart. If you gathered up these macro-facts, you could create another story of me. What this story would say would of course depend on the macro-facts I chose to include and how I chose to interpret them but it would likely emphasize my non-routines.
These are only two types of stories. I think the above story dimensions could be combined in a variety of ways depending on the information gathering and filtering methods. My point is no matter which facts we choose to use to create our stories, the lesson I want to learn for 2016 and early 2017 is the importance of why and how stories are both told and heard.
Ambiguity and answers
In a recent event at Thompson Rivers University called “Towards Indigenizing the Curriculum,” I listened to Rob Matthew describe the importance of stories. He explained that “anything worth learning can be put into a story.” He went on to explain stories are powerful and the role of listener is powerful and needs to be learned. “Stories involve ambiguity and not one right answer” which can be frustrating for students expecting one right answer. That frustration needs to be discussed.
According to Matthew, the listener has a shared responsibility to interpret the story being presented to them and many students become frustrated in that role. How did we arrive to a point where post-secondary students believe that there is one “right” answer to a question? How can such an approach possibly help any of us to make sound decisions when faced with increasingly complex and often conflicting information flows? How do we learn to decide better using both data-driven and people-driven stories?
Deciding Better
What does any of this have to do with analytics and the LAK Hackathon? The goal of analytics is to gain knowledge which can be used to make improvements or changes in by discovering, interpreting and communicating meaningful patterns.
Where businesses tend to have agreed on what they are trying to achieve (usually higher levels of profit) and who is the audience of their stories (usually internal operations, consumers and regulating bodies), within learning organizations and educational institutions there tends to be less consensus on both the goals and audiences within education.
What might we trying to achieve with our stories?
- Increased graduation rates and completion rates
- Revenue generation
- Efficient transmission of content and use of LMS
- Alignment of course content with prescribed learning outcomes
- Assess efficacy of new content (like open textbooks) and educational technology
- Improve learning design
- Teach skills to meet market demand
- Increase student engagement
- Innovate approaches to learning
- Improved access to educational opportunities
- Reduced reliance on the LMS
- Increased digital and data literacy
- Evidence of knowledge generation
- Better decision-making skills
- Empowering marginalized groups
- Challenging societal assumptions and remedying inequalities
- Increasing our ability to empathize with others
- (What else??)
Who might want to be our listeners?
- Institutional leadership
- Institutional operations
- Teaching faculty
- Research faculty
- Prospective students
- Current students
- Industry partners
- Accreditation organizations
- Government organizations
- Training organizations and trainers
- (Who else ??)
Learning Better
What if instead of defining our work by information gathering techniques, we define it by the goals of a particular project?
What if we made a practice of seeking out conflicting sources of information that challenge us to accept ambiguity? How might that change our opinions of stories, data, educational research and learning analytics? What types of conversation might we need to have to move such an approach forward?
Hoping to get some additional input, ideas and citations that I can use to generate conversation on this topic. (Please).
I’m trained in narrative practices following a therapeutic model (the model derived from the work of Michael White) and this has taught me that stories are individual and community practices of decision-making, and they are essentially projects of the reflective self, or group.
Business storytelling takes a different line: that with enough data and processing power, businesses can claim to be able to tell the stories of others.
This is what worries me about both health analytics and learning analytics: autogenerated storytelling of the other. So whoever we are, and whoever we think our audience might be, our stories are not our own — they are stories of others, and very often we want to use them to reform the behaviour of others. And we do this because analytics introduce a presumption of rigour that underestimates (or actively undermined) the other’s fitness to tell their own story.
I really want to resist this presumption: I want to see what conversation can emerge that’s prompted by data, but not finalised by it.
This is such a thought-provoking post, thank you for making it.
Thanks for taking the time to read and comment Kate. I value your thoughts tremendously.
Your comment reminded me that my entry into analytics was in business but for internal purposes – we created data-based stories of our efforts, our successes (and failures). We spent lots of time in the data that did prompt fascinating conversations either supporting or challenging what “we all knew” about learning and training.
It also put into words to many of my recent fears and misgivings: “Autogenerated storytelling of the other… to reform the behaviours… that underestimates (or actively undermines) the other’s fitness to tell their own story.” I’m now going to spend some time thinking of ways to put that sentence to good use 😉 Thanks again.
It’s taken me awhile to get to this but the wait has been worthwhile. I like it. I particularly like both the table of the two story types and everything from the “Ambiguity” header on. There’s not much I would really differ or object to, although I would suggest considering the two different types of stories to be “-focused” and not “-driven” as in person-focused and data-focused. I’m not sure why, but it seems to me that -focus captures it better than -driven. After all, it’s humans driving (creating and telling) both types of stories no matter what the mathematicians and data “scientists” claim with their algorithms.
And that brings me to the question you posed. “How did we arrive to a point where post-secondary students believe that there is one “right” answer to a question?” I would answer: we taught them that. In fact, many in the academy believe that. Look at the obsession with fact-checking in politics today. It’s a lot of very educated folks including professors that are obsessed with “the facts”. But there are no facts without context – even in science! One of my favorite examples is how context & scale matter in determining physical facts. That wall I’m staring at looks pretty solid. It’s a fact (at the scale of my whole body) that it’s solid and I can’t pass through it. But is it solid? Is it impenetrable? My wifi radio signal passes through it easily. If I zoom to down to micro level, I might notice large gaps of empty space between the parts of the atoms that make up the wall. Is that “solid”? If I zoom down even more and take Schrodinger with me, I might find the wall is or isn’t there depending on whether I measure it or locate it. We teach them that there is one right answer when we teach them that facts are objectives entities independent of the humans or context that define them.
And that leads me to my last main observation. Unlike Kate, my background is in social science & business. I’m an economist. I spent 30+ yrs using data to help formulate stories in business. We called them “strategies”, “plans”, “market analyses”, and “performance/financial statements” – but they were always based on data. We, both business people and economists, always make a big deal about the data and claiming that’s what we do. We do facts. We do data. But there really isn’t such a thing as data. It’s all just stories. Period. The data doesn’t exist without a story that uses it. And the data doesn’t exist without a story, a story usually unstated or implied, about how that data could be captured or measured.
See in my little story above about me being a data-driven economist & biz folk, I left out that I also studied rhetoric. One of the most influential articles & book of my grad yrs was “The Rhetoric of Economics” by D. McCloskey. See http://www.jstor.org/stable/2724987?origin=JSTOR-pdf&seq=1#fndtn-page_scan_tab_contents for the paper – I heartily recommend it and it’s very readable. (sorry, JSTOR paywall though). In it, she makes the point that everything we do and say in economics – and she does mean EVERYTHING – is just a story or a metaphor. Even the data tables are really stories. The equations, the calculus, etc, is all just stories.
So I urge you to beware of falling into a trap that I feel too many folks, often from a humanities or ed persuasion, fall into. Perhaps they feel intimidated by the “data” and all that magical, confusing mathematics and the claims of facts! facts, I say! They seem too willing IMO to separate narrative and stories into a separate genre of research or knowledge. That only helps people associate stories with soft, fuzzy, subjective things that can deceive as opposed to those facts and data. After all, facts and data are objective, right? Except stories aren’t a different genre, they’re the superset of which data-focused stuff is a subset.
One last suggestion if you should get the time, is to read a recent book called “The End of Average” by Rose. (i think that’s the name). It really explains how the data we use, the measurements, are all basically shorthand for stories that are unstated. In particular, most of the stats – almost anything relying on the mean (average) as the norm – is garbage and is lying to us. Really excellent book and particularly useful to anyone in ed analytics.
Again, thanks for the post. I can tell now that when I’m doing my long commute to work in coming days I’m likely to be wondering how the story-type taxonomy you’ve started could be expanded.
Thanks so much for taking the time to reply – Did I really say “data-driven”? I guess I did… I have been using “data-supported” and think I like “data-focused” even more.
At one point a Star Trek: Next Generation meme of Data driving a car while apparently grooving to music. *That* is the only way I ever see the phrase “data-driven” now. In the world of blogs, I suppose I could correct it, but I’ll leave it as a reminder of my error, lol.
I’m working on putting together the LAK Hackathon presentation now and your thoughts combined with Kate’s are very, very helpful 🙂
Thanks again!