Why Data Is the Most Important Tool for a Higher Education Leader

• 07/20/23

CT: When you're working toward analytic maturity as an institution, is there an end goal that is achievable and you're done? Or is it more of a process of continuous improvement?

Simon: There is a continuum of maturity that we like to strive for, based on the work of Tom Davenport, who is considered the father of modern-day analytics. At the very lowest end of the spectrum, it's hand-entered information, hand-scraped, retyped, etc. That's even before the reporting. The next phase is your basic regurgitation of what we already know — autopsy data. It already happened. You can't fix it. Then we move up, in terms of maturity, to alerting: We think something's about to happen, or we might have observed something just happen, and we have an opportunity in the short term to leverage data to begin to address it and intervene. Then we get into predictive analytics — instead of thinking about what happened in the past, we try to determine what is the best that could happen in the future. That's an inflection point where I see a lot of institutions of higher education begin to turn the corner on addressing some of these real big challenges that are ahead of them. They're taking their destiny into their own hands and they're running simulations of "what if" predictions and really trying to change policies, practices, and procedures related to that. Then lastly, machine learning and artificial intelligence are at the upper end of that spectrum of maturity.

ML and AI is an area where institutions with smaller staffs or smaller budgets might begin to leverage those practices a little differently to enable targeted interventions or focused support systems for students. But I would say most institutions of higher education are not there yet. The majority of institutional leaders who I talk with or call me for guidance, they're still either in the alerting phase or just barely starting on predictive analytics.

CT: How do you think emerging tools such as generative AI are going to impact how institutions approach analytics? Could AI help institutions move toward their analytics goals faster?

Simon: I like to call myself a healthy skeptic or a supporter with an asterisk. I'm very much focused on how natural language processing tools like ChatGPT are moving us ahead in the space. But I would say that many of these technologies are still transactional. They are good at regurgitating information, but not necessarily synthesizing it and using storytelling to help an administrator at a particular institution really do something with the data.

For example, ChatGPT can help individuals scour historic data in institutional fact books, and identify fairly quickly what that data is and what the trends might be, but it's not necessarily going to help with more complex questions. What policies and procedures would we need to change in order to see a 2% increase in retention and what impact would that have on our net tuition and revenue? That's not something that ChatGPT can answer in its current form. Someday, yes, it should.

At the same time, I've seen ChatGPT be incredibly useful to institutional researchers who are looking for ways to bring a different level of persuasion to our work. We can utilize visual NLP tools like Adobe Firefly or Midjourney, because a picture is worth a thousand words, right? For instance, if we have qualitative research with quotes from students at our own institution, we can utilize those tools to provide a face to the words while protecting the anonymity of the actual student.

I believe we need to teach data professionals about the technology — but just like a hammer could either be used to build a home or be turned into a weapon to do harm, it's still the hammer. It really comes down to how we as professionals in technology and data are trained in how to use it. AI and NLP can help in an institution's journey toward analytic maturity, but only when the institutional culture and those who are leading the data areas have an interest, have the time, and have the desire to learn more about the technology and apply it to higher ed.

CT: What do you hope people will take away from your session?

Simon: I hope that they learn how to leverage the culture of their institution, the partners that they can build across their campus, and the obstacles that they can break down, to achieve a higher state of analytic maturity.

Attendees will walk away with a potential roadmap on how to begin to have these conversations across their institution. Because at the end of the day, data's just data. For me — I'm a first-generation college student — and really it's not about the data. It's about the decisions and the policies and the practices and the culture that can either help students get through and graduate quicker with less debt, or continue some of our practices that might make higher education more expensive or prolong students' time to graduate.

I'm hoping attendees will see the connection between the challenges that are ahead of us in higher education and the need to act now, grab this by the horn, and really champion the data conversation. Even if you're not the primary decision-maker at your institution, I believe everyone has an opportunity to impact the state of their data culture, and I hope that the session gives attendees a key push on the first one, two, three short-term things that they can do to make it better.

To hear more from Jason Simon, register for Tech Tactics in Education at TechTacticsInEducation.com.