Strategic Directions | Feature
A Collaborative for Student Success and Institutional Comparisons
A Q&A with PAR Chief Research and Strategy Officer Ellen Wagner
The Predictive Analytics Reporting Framework (PAR, http://parframework.org) began in 2011 as a research project to investigate the potential of learning analytics for student success, and was administered by WCET under the auspices of the Western Interstate Commission of Higher Education (WICHE). Now in 2014 PAR is set to receive its 501.c.3 nonprofit status by the end of the year and is operating on its own as a member-supported, not-for-profit, analytics-as-a-service provider.
PAR's innovative work includes software and database development, but focuses more on community rather than developing new software tools. Joined by more than 20 member institutions to date, the organization is a growing collaborative venture that pools normalized (and anonymized) data to support research and create predictive models and strategies for intervention. With 1.8 million-plus records in its dataset, PAR is offering its members highly reliable data modeling for student success, as well as a new potential for comparative institutional research. CT talked with Ellen Wagner to learn more about PAR's strategy.
Mary Grush: As PAR moves forward, will the organization be building new, more innovative software tools for analytics?
Ellen Wagner: I know that during this next year we will continue to build off of our core, and yes, in fact we will see more product development. But that's not really the point. While we couldn't do all we've been talking about without tools to work with, our primary interest has never been in creating brand-spanking-new, never-before seen analytics software or platforms, per se. We are really looking more at what we can do to change perspectives on student success, and using analytics so that decisions about supporting students are driven by evidence. We've been keeping our interests focused on what we can do as practitioners to leverage technology we already have on campuses.
Getting people to think differently about using analytics and evidence to drive decision making — rather than taking a top-down, authority-driven approach — is hard enough! We don't need to create a deeper level of disruption by adding more technology that people quite possibly would not know what to do with. There's no point in making this as hard as we possibly can.
At PAR, we are both education practitioners and experienced software developers. And as software developers, we know that we can come up with great new tools, but if we don't understand how people are going to deal with them, our objectives are going to be compromised. We are not putting together a patchwork quilt of new software and hoping that it will turn into something that we like. Instead, we are looking first at ways we can actually encourage better adoption of analytics. Then, the introduction of new tools will make a lot more sense.
Grush: How do you focus on adoption, then?
Wagner: With our first grant to start PAR in 2011, Josh Jarrett at the Gates Foundation challenged us to think not so much about creating technology, but about creating a movement. Analytics were going to be big enough, and new enough, that simply talking about the new technology was not the thing that would tip us into adoption. The fact that people — including of course, our members — believed that we were doing things differently really was more powerful.
Also, our decision to be transparent has been very helpful. We realized that we were not the only ones working in analytics in higher education. It was clear that if we can all be focused on success — student success, institutional success — then in sharing our work openly we can more easily figure out the parts we need to play to move everything forward.
Grush: Will PAR function like a consultancy to its member institutions?
Wagner: We certainly will provide the support that members need to get started, stay engaged, and make sure that they can deploy the results of analyses at their institutions. However, we never expected that we would ask people to turn over their analytics work to us so we could do everything for them. PAR is not about outsourcing services, because analytics are not the kind of things that an institution can turn over to a consultant and have them figure it all out. And while we know that we could provide general services, we are not trying to be a consultancy so much as we are a collaborative.
The way that PAR works, and the way we've always imagined it, is that data-contributing institutions bring [normalized] student records that we add to the PAR dataset. All the variables are commonly defined, and because we have been very systematic about how the data goes in, the value of this resource is that it is the only place you can go to find a giant dataset that has been normed, and put through quality assurance. We can do all kinds of analyses on this dataset.
And even though we have this great resource, we are not a regulatory agency and we have no desire to go forth and tell people "how it should be." We would not insist that all of higher education come along with us, either, although we do think that would be a great idea, of course! We've been around higher education long enough to know, that's not the way analytics adoption is going to work!
Grush: What areas of decision making could the PAR dataset support, beyond student success and intervention?
Wagner: Institutions want to know more about what is going on with students and whether the services and support they are providing hit the mark. They want to determine customized benchmarks for their campus. And because the PAR federated dataset is set up at the national level, it is possible to get comparison samples that are institution-specific focused on, for example, retention, completion, progression, or other national issues, such as online learning efficacy, serving adult learners, and navigating the performance-based funding landscape.
Grush: What would you identify as PAR's highest value at this point in time?
Wagner: PAR is about helping people figure out together how to make better decisions, using data that supports institutional effectiveness and student success. Data is being used in numerous different ways, and for different types of decision making. This push toward analytics is going to be as profound a shift for most of us working in education as moving computers into the classroom and into administration was for us.
[Editor's note: The PAR Framework will receive a 2014 CT Innovator award at Campus Technology's annual summer conference in Boston, July 28-31. See CT's description of this innovative work and then visit PAR's Web site.]