A former healthcare data guru steps up to drive an analytics platform and application strategy for a major edtech company. 

INTERVIEW | by Victor Rivero

As a data scientist for a cloud-based predictive analytics platform for colleges and universities, David Kil recently joined Civitas Learning to drive the buildout of a sophisticated and sustainable platform, to guide the development of applications that use strategic data tools to engage students, empower faculty and support advisors. A former healthcare data guru, Dave brings an interesting perspective to the education technology industry. Most recently, David was the chief science officer at HealthCrowd. Prior to HealthCrowd, he founded and was the CEO of HealthMantic, a company focusing on lifestyle-medical informatics. Prior to founding HealthMantic, David was chief scientist at SKT Americas and chief science officer at Humana, responsible for the development and deployment of healthcare informatics applications. At Humana, he led the design and development efforts in the enterprise knowledge engine, predictive modeling and outcomes analytics. He holds numerous patents for his data and analytics work in the healthcare and other enterprise sectors. He has also published or co-published numerous peer-reviewed, academic

There is a tremendous amount of innovative work being done in the world of analytics and data science, and so much of that work will pay tremendous dividends in the education sector.

articles on the topic of analytics and system design, including a book on pattern recognition and prediction published by Springer Verlag. “There is a tremendous amount of innovative work being done in the world of analytics and data science, and so much of that work will pay tremendous dividends in the education sector,” says Dave. “Civitas Learning is both on the cutting edge of analytics and has a deep understanding of the challenges facing students and faculty today – the best of both worlds. I’m really looking forward to helping students make the most of their learning journeys.”

Victor: What are some of the main discoveries in healthcare analytics that can apply to the world of education?

Dave: One key discovery in healthcare is that, in order to produce positive outcomes, it is not enough to know impactful conditions (lifestyle illnesses), predict future health costs, and then have nurses call patients based on predicted scores. We had to understand engage-able moments and what interventions would be effective for what population segments under what situations. We did a number of outcomes analyses to drill down deeper as it was difficult to find macro-level interventions proven to be highly effective for all. In other words, focusing on the individual patient with the right intervention at the right time was key.

Furthermore, in today’s world of smartphone sensors and expectations for real-time recommendations, we had to find ways to speed up analytics and scale up implementations so that we could track health progress, provide action recommendations, and measure outcomes all in real time to facilitate fast learning.

Victor: How can analytics really help students be successful?

Dave: Analytics can help students by helping us better understand and guide them on their diverse education paths. Analytics can encompass a wide range of activities. On one hand, advanced time-series signal processing is required to transform transactional data into temporal events linked over time for knowledge discovery. At the other end of the spectrum, analytics can help us deliver the right micro interventions to the right student at the right time based on the likelihoods of student receptivity and impact. There are numerous personalized ways to improve learning and influence student outcomes. Analytics will play an important role in finding the best ways to help students succeed over time – one student at a time.

Victor: How important a role does analytics play in any intervention effort with a student?

Dave: I used to tell clients that if they just collect customer data, and run predictive models, it is not all that useful. Analytics must facilitate a fast-learning feedback loop by linking data to insights to actions to outcomes in a rapid cycle. Otherwise, why bother collecting so much data?

There are some schools of thought first popularized in the early 1980’s that expert-driven rules can lead to effective interventions. Given today’s multiple learning modalities and complexities today’s young and adult students face, it will be a daunting task to develop these rules by hand, not to mention evaluating their effectiveness scientifically given the fickle nature of user engagement. Instead, we use various analytics techniques to help our partner colleges and universities identify and prioritize engage-able moments and test interventions that can lead to student success.

Victor: Are there different types of analytics (for example, predictive analytics) that play a given role at any given time during a student’s academic journey?

Dave: At Civitas, we divide what we do into insight analytics and action analytics. They play off of each other and provide synergy. Insight analytics encompass student lifecycle predictions, clustering/segmentation, understanding key behavioral/intervention influencers for each cluster/segment, specialty models for deep domain-specific applications, and outcomes analytics. They are designed to provide insights. They put spotlights not only on targets of opportunities, but also on how various interventions work and how they can be improved.

Action analytics focus on identifying opportune moments of engagement through complex event processing, predicting highly impactful interventions, and prioritizing these interventions based on the principles of considerate computing. They also help us deliver the right micro interventions to the right students at the right time using the most effective channels, — such as email, SMS, and push notifications — and feeding exhaust data to the insight analytics for even greater insights and helping us design more effective interventions working with our clients.

Victor: What does the future of analytics hold for the education sector?

Dave: It depends on how we can hold analytics accountable for outcomes. Are insights generated from analytics powerful enough to not only help students heading in the wrong trajectories but also lift those who are doing well? How can we link insight analytics with action analytics so that we can create a virtuous cycle of better insights leading to improved outcomes in a fast learning environment? At the same time, we must respect student privacy and minimize student, faculty, and advisor burden for participation.

Analytics mean different things to different people. To us, we must turn data into personalized, effective interventions to help students succeed. There is a time compression factor, meaning today’s users want more useful insights and action recommendations delivered just at the right time. We are working hard to make this possible in the near future.

—

Victor Rivero is the Editor in Chief of EdTech Digest. Write to: victor@edtechdigest.blog

***

WANT TO EXPAND your edtech company rapidly? Leading administrators say that one of the best ways to really connect with them is to show them examples of where and how a product or service is working for their peers. Rich, alive and informative case studies are the perfect way to get attention, attract interest and deliver your message. Get the word out now to increase your revenues. Request more info.