Updated article originally published September 27, 2016.
For over a decade, higher education has looked to big data for supporting teaching and learning practices. While institutions have access to enormous unprecedented amounts of information, they have struggled with how to meaningfully use the massive amount of data collected about every student. Educators find themselves wondering What does the data really say?
Big Data in Practice: Taking a Lesson from Healthcare
American physician-economist, Christopher Murray, was a pioneer in using big data to change global healthcare policy, practice, and funding. Understanding that the devil is in the details, Murray used individual data points to identify the true causes of issues and the best solutions to address them. Until this point, best practices for healthcare had organizations using broad data to inform their focus, often resulting in fighting or funding the wrong thing.
A Forbes post discussing big data argues that higher education is at risk of misinterpreting and using data the same way. The increased pressure institutions face in reporting on educational technology and innovative teaching may result in broad solutions designed to assuage funding bodies. However, without careful consideration of what the data is actually saying, big data applications face the risk of being “based on everyone and relevant to no one.” The key to education not falling into the same trap is making good use of macro- vs. micro-level data.
Using Big Data to Improve Student and Teacher Performance
In 2012, City University of New York (CUNY) educators identified eight categories for applying educational big data. These categories ranged from evaluation of students’ performance to testing and evaluation of curricula.
According to this study, several institutions used macro-level data to implement learning analytics applications. These systems included:
- Northern Arizona University’s Grade Performance System, which sent emails to students to inform them of academic issues as well as share positive feedback.
- Purdue University’s Course Signals System, which was designed to bolster student performance and learning outcomes with real-time, frequent, and ongoing feedback.
- Ball State University’s Visualizing Collaborative Knowledge Work, an analytics application that encouraged continuous formative evaluation among collaborators. While these systems have faded from the spotlight since their original launch, they built the foundation for big data innovations in higher education teaching and learning.
For instructors who are interested in using micro, or classroom and individual, data, EDHEC Business School and SmartData Collective highlight starting points for intervention in the classroom.
Adapting Teaching Style
Discussions of big data can be overwhelming, and typically do not feature enough details regarding teachers. This oversight in teacher relevancy can be rectified by looking at how big data may assist teachers in reconceptualization of course structure.
Traditionally, lesson plans are created at the beginning of a term and remain unchanged as class progresses. However, course analytics and assessment design can play a key role in the identification of misunderstood content. Analytical measures can determine the difficulty and comprehension of a specific question, allowing instructors to identify when a class, as a whole, is grasping a concept.
An instructor can, then, change a teaching plan based on information about the class’ educational progress. For concepts that are proving more challenging, instructors can assign additional practice and readings as well as asynchronous learnings to ensure core competencies are learned. Similarly, instructors can pace the course faster at points when the class shows mastery.
Personalizing Student Learning
In the same way that educators can assess student performance and learning outcomes at a class level, they can determine challenges and successes on an individual basis. Once issues have been identified, educators can intervene by sharing extra resources to help struggling students improve their understanding of core concepts. Similarly, high-performing students can be given the option to engage more with a topic via supplementary content.
The Future of Data
Data is already challenging preconceived notions about the ability to predict and evaluate student success. This growing resource fuels both hope for better learning and concerns about data privacy and data bias. We must be careful as we make decisions about its integration in our institutions.
Of note, too, is that despite its great potential, the data is only as good as the data collection method. Designing meaningful, performance-based assessment as well as employing appropriate analysis is essential to successful teaching and learning with big data.
