Some people are more fluent in the humanities than in quantitative or computational discourses. No matter. The future of education will be re-created by the language of math and its many relatives. Data, analytics, and AI are not passing fads, nor are they simple augmentations to what will remain a largely unchanged and age-old educational enterprise. As such, whether you dream in narratives or numbers, whether your heart language is that of literature, history, and philosophy or theorems, derivatives, and algorithms; you need to venture into the world of quantifiable data and numbers to understand and be a co-creator of the future of education. That is not easy for me to write or admit. I have a deep and longstanding passion for the humanities, and if you stay with me throughout this series of articles, you will read what I hope to be a strong case for their role in this new world of numbers, but that does not change the fact that quantitative and computational thinking and the results of that thinking will be a powerful driver of fundamental change in education.
As a student of futures in education, I am not a fatalist. I see ample evidence that individuals and groups of individuals can influence what happens in the future. The future direction of data, analytics, and AI in education is something that you and I can influence. If I didn’t think this to be true, I wouldn’t bother to write this article. There are and will be strong corporate, government, political, social, and philanthropic influences in this future, but your voice and actions matter as well. It is with this confidence that I make this next statement. Whether data, analytics, and AI will empower or simply control and influence learners is partly up to the decisions that we collectively and individually make in the next 5-10 years. I am here to advocate for an approach that strives to promote a better, more hopeful, more humane, and more empowering educational ecosystem, and it begins with a challenge for us to better understand the landscape. That is the goal of this first article, to offer a simple framework for beginning to recognize and understand the role of data, analytics and (in some cases) AI in education, now and in the emerging future.
In a recent presentation, I guided a group of higher education colleagues through five categories (I prefer to use the metaphor of “buckets” for some reason) for thinking about the role of data in modern higher education. Describing those five will give you some ways to make sense of what might otherwise look like an overwhelming pile of numbers, metrics, and statistics; but there is most certainly method and strategy in the numeric and data-immersed madness of our higher education age.
College Search and Connection Data
These data relate to the connection and relationship between higher education institutions and prospective learners. However, because I am thinking of this as a two-way connection, this might also be the data that learners review, collect, consume, and share about institutions. On the prospective student side, this might include data points like leads (people who have inquired or expressed interest), qualified leads (leads that have been vetted to some extent…people use this terms a bit differently from one context to the next), applications, completed applications, deposits, enrolled, and ultimately starts (How many and which people actually enrolled and started their first course?).
These data also include a collection of data points related to cost. These might include return on investment metrics for marketing and recruitment strategies; along with common online marketing measures like site traffic, source of that traffic, returning visitors to the site, cost per click and/or impression for online ads, conversion from visitor to qualified lead from various sources and marketing tactics.
However, since I also include University data in this bucket, at least to the extent that it plays a role in what school(s) a prospective student chooses (or rejects), we can think of data like location, price, features, graduation rate, demographics, selection of majors and programs, ratings and rankings, reviews (not numeric, but reviews are often turned into numeric rankings and ratings), and much more.
This is a simple introduction to the subject, but in my work in higher education and my interaction with various organizations involved with higher education, I can say that there are indeed several hundred relevant data points in this first bucket. There is also a dominant metaphor for thinking about and using these data today, a sales and shopping metaphor. It is not the only metaphor, however, and I see growing evidence of that metaphor changing, a topic that I will explore in a future article in this series where I focus exclusively on a few cautions, opportunities, and considerations about the role and use of college search and connection data.
Student Retention & Success Data
These data relate to features, behaviors, indicators, attributes, and metrics that are attached to whether students persist, stop out, drop out, or graduate. It might include data that allows one to track or identify risk factors like attendance, performance on tests/courses/assessments, financial status, high school or prior GPA and test scores, performance in what have been identified as “gateway courses”, measures of student interest/motivation/engagement, and much more. At the same time, this can include factors associated with retention and student success like faculty responsiveness to student questions; the quality, quantity, timing, tone, and nature of feedback on student work; the scheduling and timing of coursework and how that aligns with student life demands and needs; and again, so much more. If it is data that people use for insights or indicators related to student retention and/or success, then I put it in this bucket.
Given ongoing concerns about the cost of higher education, enrollment challenges for schools, questions about the relevance of the college degree, demands for more evidence about the return on the higher education investment of students and the public (especially where tax dollars contribute to the schools), and concerns about the disruption of higher education from various sources; this is one of the more highlighted areas for data and analytics today. In addition, while arguably in its infancy, the use of predictive analytics and rudimentary forms of what some today are defining as artificial intelligence, student retention and success is an area where many institutions are gaining the most traction around the use of data.
Yet, not everyone is comfortable reducing data about the student experience in college to measures about retention and graduation. IT feels a bit to corporate for some in education. There are still interest, however, in indicators that students are flourishing, engaged, experiencing a sense of belonging, or fully and actively engaged in the community.
In a future article, I will highlight some of the promising practices, but also some of the cautions and important considerations as this area develops.
Student Learning, Progress, and Mastery
While related to the last category, these data are more about what students learn, how well they learn; their motivation; interest; the conditions under which they learn best or worst; learning gaps and strength; data that compares performance across individuals/cohorts/populations; and their progress toward mastery, progress, proficiency, competency, or perhaps a more nuanced and sophisticated approach to a given discipline or subject. I offer the different terminology at the end of that last sentence to recognize that these data embody more than a single philosophy or approach to teaching and learning.
I’ve convinced that data/analytics/AI related to this bucket will be one of the two most significant in transforming or truly disrupting formal education as we know it. At a minimum, it will have a significant impact, even (or especially) if most of the research and applications for these data happen in learning environments outside of the hallowed halls of academia.
As with the other categories, I will dedicate a separate article in this series to data/analytics/AI and student learning. It may well be the most challenging area for some to openly consider, especially as a I begin to get into what will sound like far-fetched science fiction to some readers. I will make what I hope to be a convincing argument that it is neither far-fetched nor science fiction…at least not the type of science fiction that has no grounding in a current or emerging future reality.
What does it mean to be a healthy and viable higher education institution? Ask any group of school leaders and, regardless of the mission, financial viability will be in the top three or four in the list, often in the top two. Of course, there will also be answers related to being faithful and committed to the distinct mission and core values of a given higher education institution. Put all of those answers in a list, find the data points that are solid indicators for each of those items, and you get a good understanding of what I’m referring to in this category. These data points are increasingly represented in scheduled (weekly, monthly, quarterly, annual, but sometimes real-time) reports and dashboards. Just like we use a dashboard when driving a car, people are increasingly using them to steer organizations, or at least to gauge how things are going, seeking actionable insights for next steps.
Learner Connections With People and Organizations
This is the most undeveloped category in higher education, but I contend that this, along with the category on student learning/mastery will become one of the most important in the future. Austin Kleon wrote a short, informal, and highly influential book about life and work in a digital age. In it, he argued that if you are not online, you don’t exist. He was referring to the means by which we are and our work is discovered, how we connect with other people and organizations, and how we tell meaningful stories and learn from the meaningful stories of others in a digital and connected age. Consider that each data point related to a person is a possible connecting point in a world of analytics and artificial intelligence, and that is already changing how we meet people, network, find jobs, shop, learn, volunteer, collaborate, solve problems, and create our public identity. Right now, most higher education institutions are hardly even thinking about this category of data, not beyond a few outliers, some workforce development efforts, internship programs, experimentation with micro-credentials and portfolios, and some promising 21st century caliber job board and networking technology. There is already a great deal of data in this area, but few have developed a deep understanding or clear vision for what it means for the organization and especially the learners (and graduates). Again, I will dedicate a separate article to unpack this, as I plan to do so with the others.
As I’ve tested these five categories with colleagues and at least one group of people, I’m moderately confident that they represent a useful way to categorize the data that has relevance in current higher education conversations as well as ones that I’m confident will become increasingly important in the near future. The purpose of this first article was to introduce the categories and the topic, to build a basic vocabulary or framework. In the next five articles in this series, I will explore each category in greater depth. I have no intention of providing an exhaustive commentary for each of these five categories, but I will highlight the role of the category, offering a couple promising practices, some challenges, as well as suggesting some possibilities for the future. Then I plan to fill out the series with a seventh and final concluding article, summarizing and highlighting some of the insights in the series, and suggesting some next steps.