This article is a selected piece from the 2014 NMC Horizon Report For Higher Education
Learning analytics is an educational application of “big data,” a branch of statistical analysis that was originally developed as a way for businesses to analyze commercial activities, identify spending trends, and predict consumer behavior. As web-tracking tools became more sophisticated, many companies built vast reserves of information to individualize the consumer experience. Education is embarking on a similar pursuit into new ways of applying to improve student engagement and provide a high-quality, personalized experience for learners.
Learning analytics research uses data analysis to inform decisions made on every tier of the education system, leveraging student data to deliver personalized learning, enable adaptive pedagogies and practices, and identify learning issues in time for them to be solved. Other hopes are that the analysis of education-related data on a much larger scale than ever before can provide policymakers and administrators with indicators of local, regional, and national education progress that can allow programs and ideas to be measured and improved. Adaptive learning data is already providing insights about student interactions with online texts and courseware. One pathway to creating the level of data needed for effective learning analytics is seen in creating student devices that will capture data on how, when, and in what context they are used, and thus begin to build school-level, national, and even international datasets that can be used to deeply analyze student learning, ideally as it happens.
Since the topic first appeared three years ago in the far-term horizon of the NMC Horizon Report: 2011 Higher Education Edition, learning analytics has steadily captured the interest of education policymakers, leaders, and practitioners. Big data are now being used to personalize every experience users have
on commercial websites, and education systems, companies, and publishers see tremendous potential in the use of similar data mining techniques to improve learning outcomes. The idea is to use data to adapt instruction to individual learner needs in real-time in the same way that Amazon, Netflix, and Google use
metrics to tailor recommendations to consumers. Analytics can potentially help transform education from a standard one-size-fits-all delivery system into a responsive and flexible framework, crafted to meet the students’ academic needs and interests. For many years, these ideas have been a central component of adaptive software, programs that make carefully calculated adjustments to keep learners motivated as they master concepts or encounter stumbling blocks.
New kinds of visualizations and analytical reports are being developed to guide administrative and governing bodies with empirical evidence as they target areas for improvement, allocate resources, and assess the effectiveness of programs, schools, and entire school systems. As online learning environments increasingly accommodate thousands of students, researchers and companies are looking at very granular data around student interactions, building on the tools of web analytics. Pearson Learning Studio, for example, provides an LMS infrastructure that is aggregating data from the millions of learners using their systems, with the aim of enabling school leaders and national policy makers to more effectively design personalized learning paths.
Similarly, a group at Stanford University is examining vast datasets generated by online learning environments. These efforts are taking place through the Stanford Lytic Lab, where researchers, educators, and visiting experts are currently building an analytics dashboard that will help online instructors track student engagement in addition to conducting a study of peer assessment in a MOOC on human-computer interaction, based on 63,000 peer-graded assignments. In April 2013, the Bill & Melinda Gates Foundation awarded Stanford more than $200,000 in funding to support the Learning Analytics Summer Institute, which provided professional training to researchers in the field.
Relevance for Teaching, Learning, or Creative Inquiry
Learning analytics is developing rapidly in higher education, where learning is happening more within online and hybrid environments. It has moved closer to mainstream use in higher education in each of the past three years. Sophisticated web-tracking tools are already being used by leading institutions to capture precise student behaviors in online courses, recording not only simple variables such as time spent on a topic, but also much more nuanced information that can provide evidence of critical thinking, synthesis, and the depth of retention of concepts over time. As behavior specific data is added to an ever-growing repository of student-related information, the analysis of educational data is increasingly complex, and many statisticians and researchers are working to develop new kinds of analytical tools to manage that complexity.
The most visible current example of a wide-scale analytics project in higher education is the Predictive Analytics Reporting Framework, which is overseen by the Western Interstate Commission for Higher Education (WICHE), and largely funded by the Bill & Melinda Gates Foundation. The 16 participating institutions represent the public, private, traditional, and progressive spheres of education. According to the WICHE website, they have compiled over 1,700,000 student records and 8,100,000 course level records in efforts to better understand student loss and student momentum.
Companies such as X-Ray Research are conducting research in online discussion groups to determine which behavioral variables are the best predictors of student performance. The tools reflect the potential of analytics to develop early warning systems based on metrics that make predictions using linguistic, social, and behavioral data. Similarly, studies at universities are proving that pedagogies informed by analytics can improve the quality of interaction taking place online. At Simon Fraser University in British Columbia researchers applied analytics to solve an issue that past experiments revealed — discussion forums used for online courses were not supporting productive engagement or discussion. They developed a Visual Discussion Forum in which students could visualize the structure and depth of the discussion, based on the number of threads extending from their posts. Learners in this study were also able to easily detect which topics needed more of their attention.
Learning Analytics in Practice
The following links provide examples of learning analytics in use in higher education settings:
Big Data in Education
Columbia University professors offer an online course for educators through Coursera to learn about the strengths and weaknesses of the various methods professors are currently using to mine and model the increasing amounts of learner data.
The competency map at Capella University helps students own their learning by constantly showing them where they are in each course, how much is ahead of them, and where they need to concentrate their efforts to be successful.
The University of Michigan uses Gradecraft, which encourages risk-taking and multiple pathways towards mastery as learners progress through course material.The analytics employed guide students throughout the process and inform instructors of their progress.
For Further Reading
The following articles and resources are recommended for those who wish to learn more about learning analytics:
Data Science: The Numbers of Our Lives
(Claire Cain Miller, The New York Times, 11 April 2013.)
According to a report by McKinsey Global Institute, there will be almost a half million jobs in data science in five years. Institutions are creating programs to train hybrid computer scientist/software engineer statisticians.
Learning to Adapt: A Case for Accelerating Adaptive Learning in Higher Education
(Adam Newman, Peter Stokes, Gates Bryant, Education Growth Advisors, 13 March 2013). A white paper funded by the Bill & Melinda Gates Foundation illustrates the current adoption of adaptive learning technologies in higher education, relevant obstacles, and the solutions being explored.
The Role of Learning Analytics in Improving Teaching and Learning (Video)
(George Siemens, Teaching and Learning with Technology Symposium, 16 March 2013.) Siemens reviews a number of case studies to show that when analytics are applied to education in a similar manner as companies
This article is a selected piece from the 2014 NMC Horizon Report For Higher Education