- What is the Predictive Analytics Reporting (PAR) Framework?
- Why is the PAR Framework unique?
- Why does PAR matter?
- What is Analytics?
- Which U.S. institutions helped establish PAR?
- What is involved in joining PAR?
- How can my institution become a member of PAR?
- How will PAR benefit my institution?
- What steps have PAR taken to ensure the security and privacy of student records?
- Who manages the PAR Framework?
- What types of professionals work for the PAR Framework?
- Does PAR plan to expand the number of participating PAR Framework institutions in the future?
- What types of data does PAR analyze?
- What types of programs are included in PAR institutional samples?
- Have PAR findings been published in peer-reviewed journals?
The Predictive Analytics Reporting (PAR) Framework is a membership organization that provides learner analytics-as-a-service to its members. PAR provides: unique national benchmark reporting based on student outcomes; conducts predictive analyses on variables contributing to student risk; and explores best practices in student retention through the use of intervention measurement. During the past 5 years, PAR has developed a massive collection of student records in a quest to understand the variables affecting student loss (e.g., what signs indicate that students may be in danger of dropping out), and student momentum (e.g., what signs indicate that students are making progress toward program completion). The PAR Framework employs predictive techniques commonly found in business intelligence settings to aid in educational decision-making.
PAR is unique among higher educational analytics providers for three reasons. First, PAR uses common data definitions so that data from each and every member institution can be efficiently and reliably aggregated into the single federated dataset. Using common data definitions make it possible to conduct predictive analyses where findings can generalize within and across heterogeneous institutions. Second, PAR uses data that is readily available for all students at each of the participating institutions–adding both to the generalizability of findings and the opportunities to scale solutions across sub-populations of students, regardless of where they attend school. Third, PAR uses industry-leading software platforms including SAS, Oracle, and Amazon Web Services that are already found on today’s campuses, reducing the need to invest in new platforms (and related consulting services), and it requests educational stakeholders to reconsider how to better use data to support decision-making.
Student retention and analytics initiatives are most often based on institutional-specific situations or activities. By federating and aggregating independent institutional datasets into a single federated data resource, PAR makes it possible to conduct predictive analyses on a vast collection of de-identified student records and apply the findings of those analyses to find interventions shown to be effective with particular student populations at particular points of need. PAR is unique in using predictive analytical insight approaches commonly practiced in business intelligence to aid in higher-education decision-making. Perhaps even more important, PAR links predictions of risk to interventions used by campuses to mitigate risk. Using techniques such as A/B testing, PAR supports intervention measurement and takes the guesswork out of finding students at risk, which gives decision-makers the tools they need to directly support student success at the point of need.
Analytics refers to business intelligence methodologies used to support data-driven decision-making. Analytics uses information collected from all over an enterprise to keep track of operational and strategic activities and evaluates how well activities are contributing to essential enterprise goals. As a derivative of business intelligence methodologies, analytics used in higher-educational settings helps inform institutional decision-making on three unique fronts:
- Learning Analytics focuses on the best ways to teach and learn;
- Learner Analytics focuses on the best ways to support students;
- Organizational Analytics focuses on the best ways to operate an institution.
PAR is a learner analytics provider that finds students likely to be at risk of dropping out of school, identifies what variables are contributing to risk, and then helps institutions find the best interventions for mitigating specific risks, based upon the results of intervention measurement. By knowing what interventions are most likely to yield the best results with what kind of student populations, it is possible to focus energy and resources at points of demonstrated need to support student success.
PAR began as a research project, funded by the Bill & Melinda Gates Foundation, to see if it was possible to use predictive analytics as a way to find students in the U.S. post-secondary pipeline at risk of dropping out of college. The PAR “proof-of concept” grant was supported by six forward-thinking institutions, several of which contributed important ideas from their own proprietary predictive platforms to further this important conversation. PAR’s founders were instrumental in launching learner analytics as a legitimate disciplined inquiry in U.S. post-secondary education.
Founding Partners include:
- American Public University System
- Colorado Community College System Online
- Rio Salado College
- University of Hawaii System
- University of Illinois Springfield
- University of Phoenix
During the Implementation grant, an additional 10 schools were recruited to help PAR build a data model robust enough to accommodate 2-year, 4-year, public, proprietary, traditional, and progressive institutions, including online and competency-based programs. Implementation Partners include:
- Ashford University
- Broward College
- Capella University
- Lone Star College System
- Penn State World Campus
- Sinclair Community College
- Troy University
- University of Central Florida
- University of Maryland University College
- Western Governors University
During the Transition grant, four more schools joined the PAR Framework to further extend validation of the PAR predictive models and services, national outcomes benchmarks, and the Student Success Matrix (SSMx) intervention measurement tools. These institutions include:
The PAR Framework offers a multi-institutional lens for focusing on the risks to student success in targeted, actionable ways, gaining new perspectives and identifying unique, actionable strategies for improvement. Your institution will be asked to:
- Provide senior-level engagement from academic administration, student success/student affairs, and IR/IT. With this level of engagement, along with the PAR member tools and insights, you’ll have the keys to unlock measurable improvement in student success at your institution.
- Deliver your institutional data to our cloud-based platform for analyses and insights into institutional performance, opportunities for interventions, and identification of the students who can most benefit from support services.
- Participate in the collaborative exploration and application of data-driven decision-making in U.S. post-secondary education.
PAR membership is open to accredited 2-year, 4-year, public, and proprietary institutions. Interested institutional partners have the opportunity to join PAR with a 2-year membership commitment. To express interest in becoming a member or to learn more, fill out our contact form or email.
PAR Framework participation benefits campuses in the following ways:
- Providing a unique opportunity to engage with like-minded thinkers, committed to student success.
- Delivering institutional and intra-institutional benchmarks for assessing student achievement, engagement, progress, and other critical success criteria.
- Predicting which students at your institution are at-risk for not succeeding in critical courses (or not being retained at the institution), what puts them at risk, and delivering tools to help identify what can be done to help them succeed.
- Offering validated tools and frameworks that yield the insights on how to treat at-risk students. These tools help inventory, map, and measure the effectiveness of student intervention efforts.
With the implementation of the PAR Framework, institutions will be able to make in-the-moment decisions to support students in successful progress and completion of a higher-education credential. Effectively targeted interventions can help students stay in school, which leads to higher success rates for students and institutions alike.
The PAR Framework team is committed to data privacy. PAR data handling includes IRB approvals from each of the participating member institutions and well-documented policies to ensure data privacy, record security, and research integrity. PAR utilizes a double-key approach to separating identity from records. Each member of the PAR team is certified by the Collaborative Institutional Training Initiative (CITI) in conducting human-subjects research in an ethical manner. The PAR Framework has also a member-led governance group setting the policy for utilization of the PAR Framework dataset for both researchers within institutional participants and for those from institutions and organizations not contributing data to the dataset.
The PAR Framework began as a grant-funded project housed at the Western Interstate Commission for Higher Education (WICHE). In 2015, PAR began operations as a North Carolina independent 501c3 non-profit organization. PAR is overseen by a board of directors and operated by an independent professional staff. PAR provides data services to institutions that join on an annual basis.
The PAR Framework team members bring expertise from data science, institutional research, statistical analysis, quantitative and qualitative research, cognition and instruction, online learning, higher-education administration, educational public policy, project management, and commercial software development. They come from the public and for-profit sectors. They include administrators from community colleges and 4-year comprehensives, including state-funded, private, and for-profit institutions. Several PAR institutional members/partners are award-winning, acknowledged practice leaders in the domain of analytics.
The PAR Framework works with 77 variables under the major headings of student demographics (i.e., date of birth, gender, etc.), course-level records (i.e., start date/end date, size of course, etc.), student-level academic records (academic status, transfer credits, etc.), and institutional records (i.e., type, business model, etc.). For a full list of the current PAR Framework variables, view The PAR Approach.
At its inception in 2011, PAR analysis began by examining data primarily from online programs. Since then, PAR has expanded the types of course activity being measured; PAR data includes student, course, and program data for undergraduates taking online, on-ground, and blended courses, as well as non-traditional programs of study.
The PAR Framework was featured in the “Learning Analytics” issue of EDUCAUSE Review. View the full article at EDUCAUSE Review.
In June of 2012, the work of the PAR Framework Proof of Concept was published in the Journal of Asynchronous Learning Networks.