Saturday, May 8, 2010

A New Amenities Arms Race: Student Success

Donald M. Norris
President, Strategic Initiatives, Inc.

For the past two decades, American higher education has been engaged in a sort of “amenities arms race” of epic proportions. As institutions vied for the attention and favor of ever-more sophisticated student consumers, they have acquired a generation of shiny new academic and research facilities; plush athletic stadiums, practice fields, and wellness facilities; commodious residence halls, condominium-like student apartments, and student unions with every feature from rock climbing walls to bistros to juice and coffee bars.

The aftermath of the Great Recession has slowed the pipeline of campus construction projects and dimished future prospects. But it has given fresh impetus to the race to offer an amenity of enhanced importance: improved student success. Students are demanding better feedback on what they need to do to succeed and fulfill their learning objectives on time and within their budget. Shrewd and purposeful institutions have developed a range of analytics-based tools, practices, alerts, and interventions than enable them to develop policies that improve student performance and better advise and inform students. They are also acting decisively and in real time to alert and assist students whose performance deviates from the patterns that have characterized cohorts of previously successful students.

Consider the following examples of action analytics in practice in support of student success in courses:

Purdue University has been one of the pioneers in applying predictive modeling, longitudinal data and large-scale data sets to student success.
They have mined data from systems that support teaching and learning to provide customization, tutoring, or intervention within the learning environment – this is what they call “Actionable Intelligence.”

One of their most successful efforts has been the Signals program, featured in an MSNBC news clip. Using historical data, Purdue deployed predictive modeling to identify patterns of behavior and performance in introductory gateway courses that led to success and compared them to current student efforts. Students receive a red, yellow, or green indicator to show them where they stand. Starting as a pilot, this effort has scaled to 500 gateway course sections enrolling 11,000 students at a cost of $47 a student. John Campell demonstrated this application at the First Symposium on Action Analytics.

While technology is the enabler, the Purdue Team feels strongly that it is the capacity of the organization - people, skills, and processes - that makes the difference in a successful intervention system. This system has made tradeoffs between predictive perfection and scalability; it focuses on actions that are made possible by the analytics. Purdue has partnered with SunGard to develop a commercial version of Signals that is available to other institutions.

Capella University embeds analytics in every aspect – academic and administrative – of the student experience. Capella is a market-drive (for-profit) university that uses predictive modeling to increase application rates, enrollment, and course attendance; to improve academic performance and the learning experience; and to increase persistence. Capella’s leadership and faculty are dedicated reflective practitioners. They have studied student behavior and success and understand both the characteristics of online learners and the elements of successful online learning experiences for their students. Five factors differentiate the online learning experience as a platform for predictive modeling:

• Online learning generates a huge amount of data
• The data arrive in a cyclical manner,
• Capella has a long-term engagement with learners
• They can monitor learner behavior on different time scales
• Learners have more freedom in managing their time


For Capella’s online learners, the first week is everything; students who get off to a bad start seldom catch up. So Capella perpetually monitors students, uses predictive modeling to match them to past patterns, and sends tailored messages to students to get them on track and keep them there. Predictive modeling and artificial intelligence are key elements of the management of every student’s learning experience. Alex Ushveridze of Capella University demonstrated these modeling techniques at the First National Symposium on Action Analytics.

These analytic mindsets and approaches also affect learning outcomes. Capella’s offerings are based on competence. They utilize embedded templates, rubrics, and analytics to enable students to acquire competences and demonstrate them in ways that are understandable to employers. Jeff Grann and Kim Pearce demonstrated Capella’s approaches atCapella is sharing its ideas and practices through the Action Analytics in Education Partnership and the Action Analytics Community of Practice, seeking to advance transparency and accountability in higher education.

Many institutions have developed “home-grown” predictive modeling tools for recruitment, keeping students on track, and improving student success.
Most leading-edge community colleges have similar versions of these tools to manage and improve student success. Last year, Ken Moore from Sinclair Community College described his institution’s retention and student success practices and at this year’s National Symposium, Vernon Smith from Rio Salado Community College described his institution’s efforts to develop open source solutions to student classroom progress and success. There are many models to emulate and improve upon - and these approaches have a high return on investment, when done well.

Powerpoint presentations of the Purdue University and Capella capabilities can be found at the Public Forum for Action Analytics website. They will be featured in an upcoming Webinar, along with other cases, in a Webinar announced by John Hammang of the American Association of State Colleges and Universities (AASCU). The time and date for which will appear on the Public Forum for Action Analytics.

These developments were discussed on May 5-7 at the Second National Symposium on Action Analytics in St. Paul Minnesota. At this meeting, institutional leaders and practitioners, policy makers, and foundation representatives discussed how to deploy and leverage analytics that will provoke action to improve student access, affordability, and success and enable the rediscovery of financial sustainability.

The sense that emerged from this meeting was that the imperative for concerted, aggressive action to improve student success has grown dramatically given the current state of the economy and family finances. Students are demanding better feedback and support in improving their odds of success. Their demands are destined to grow and they are likely to vote with their feet – and their clicks – if they are unsatisfied. Learners can be counted on to ratchet up their demand that the current generation of intervention and advisement tools be enhanced and extended.

No comments:

Post a Comment