Enhancing teaching and learning through educational data mining and learning analytics
Posted on May 25, 2012
Posted on May 25, 2012
This brief, developed by the U.S. Department of Education’s Office of Educational Technology, describes how educational data mining and learning analytics can be applied in educational settings. The brief reveals the value and challenges associated with using data collected through online learning systems to make educational decisions; furthermore, it provides helpful recommendations on how to most effectively use data mining and learning analytics to improve student outcomes.
With the increasing number of students utilizing online learning, there are increasing opportunities to use technology to collect data that can be used to assess student knowledge and improve student learning experiences. The U.S. Department of Education’s National Education Technology Plan envisions “online learning systems collecting, aggregating, and analyzing large amounts of data and making the data available to many stakeholders. These online or adaptive learning systems will be able to exploit detailed learner activity not only to recommend what the next learning activity for a particular student should be, but also to predict how that student will perform with future learning content…” (p.3).
Here are a few suggested applications that I found beneficial:
1) User knowledge modeling customizes the system to the students’ specific needs. For example, the model may decide what problem to give a student by inferring what the student knows based on responses to questions, time spent practicing, and errors made.
2) Behavior modeling can be used to determine student engagement by measuring things like time on task and course completion.
3) User experience modeling measures student satisfaction. One way student satisfaction is measured is by collecting responses to follow up surveys and questionnaires. I was particularly impressed with the models that considered student engagement and satisfaction as necessary elements of data collection. I fear the student perspective is often overlooked; therefore, it was refreshing to see an emphasis placed on analyzing students’ thoughts and experiences with the online learning program.
4) Trend analysis collects data over time to identify trends and changes in student learning over time.
The brief is realistic and forthcoming about potential roadblocks to implementation.
Here are a few of the challenges that are considered:
1) Technical Challenges: Although online learning offers a plethora of opportunities for data usage, one challenge is having the technical resources needed to do data mining and learning analytics. The expense and storage required for implementation can be extensive.
2) Limitations in Institutional Capacity: Along with the expense of overcoming technical challenges, there is the need to expand the human resources required to prepare, process, and analyze the data.
3) Privacy and Ethics Issues: Privacy implications arise when collecting personal information about users to customize models to students’ specific needs.
I was most interested in the section that describes how to actually make data mining and learning analytics a reality in schools. Here are a few of the recommendations I found useful:
1) Develop a culture of using data for instructional decisions: Data should be useful and accessible to instructors and students so that it can be used to inform instructional practices.
2) Start small: Making small, ongoing, low cost changes can help build the culture for using data and can prepare districts for using more expensive, powerful systems over time.
3) Collaborate across sectors: System designers, researchers, and educators should collaborate to help build the capacity of schools to effectively use data mining and learning analytics.