Insight and Action to Improve Learning from Data


What if teachers had detailed windows into each student’s past learning?

What if school leaders had these windows into learning and teaching for their schools?

What if these views of learning looked into the future as well as the past?

Education organizations of all kinds are struggling to do more with less. A cost they have taken on in recent years is the cost of maintaining immense and growing stores of data on their programs. Some data are collected purposefully in student and staff records, some data are mandated to be archived and protected, and some data are generated by systems such as learning and content systems. While education leaders know that their data can be used to student and school advantage, many education organizations lack the manpower, tools, and expertise to make use of the data.

These organizations include schools and systems at all levels from pre-K to adult, colleges, universities, governments, private training providers, education research groups, and NGOs that work in the education sector. To make use of data until now required expensive specialist analysis tools, custom-designed dashboards, analysts, and systems architects.

From the pattern of learners’ static data (for example, demographics; past attainment) and dynamic data (for example, pattern of online logins; quantity of discussion posts) schools can classify the trajectory that they are on (for example, ‘at risk’; ‘high achiever’; ‘social learner’), and hence make more timely interventions (for example, offering extra social and academic support; presenting more challenging tasks). Shum, S.B. (2012). Learning Analytics. UNESCO IITE Policy Brief, November 2012.

Now, with Power BI and Azure Machine Learning, education organizations have the powerful self-serve tools they need to view, analyze, and make predictions from data, thus turning the cost of data into advantage. In addition to reports on past performance viewed dynamically using Power BI, Azure Machine Learning includes data models that conduct predictive analytics that show likelihood of specific outcomes for students while there is time to make changes in school programs that improve outcomes. The analytics can recommend interventions and can calculate costs.

The advantages of data visualization, analysis and prediction benefit:

  • students who will get feedback on their pattern of performance in learning systems, and who may be assessed more frequently in ways that better guide progress and give them a personalized learning experience
  • parents who will get detailed reports on student progress
  • teachers who will get detailed reports on all students, as well as relative effectiveness of lessons and content, freeing them from low-order assessments to focus on more complex feedback
  • content designers and curriculum managers who will get data on content usage and relationships between content and learning
  • school leaders will get student progress data, teacher effectiveness data, and school level outcomes. Facility factors like busing, buildings, schedules, and activities can be factored into learning. Staff factors like professional learning and credentials can be analyzed
  • school system leaders who will get data across campuses year to year
  • policymakers who will get outcomes associated with different school and community conditions
  • education researchers who will get ongoing insights into impacts of practices and conditions at large scale

For more on the benefits, read A Call to Action for Research in Digital Learning: Learning without Limits of Time, Place, Path, Pace…or Evidence.

Comments (0)

Skip to main content